Recollections of "Tucson Operations": The Millimeter-Wave Observatory of the National Radio Astronomy Observatory (Astrophysics and Space Science Library, 323) 1402032358, 9781402032356

A personal account of the evolution of millimeter-wave astronomy at the US National Radio Astronomy Observatory. The aut

118 108 8MB

English Pages 238 [228] Year 2005

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Recollections of "Tucson Operations": The Millimeter-Wave Observatory of the National Radio Astronomy Observatory (Astrophysics and Space Science Library, 323)
 1402032358, 9781402032356

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

AS SL

ASTROPHYSICS AND SPACE SCIENCE LIBRARY

RECOLLECTIONS OF "TUCSON OPERATIONS" The Millimeter-Wave Observatory of the National Radio Astronomy Observatory M.A. GORDON

RECOLLECTIONS OF “TUCSON OPERATIONS”

ASTROPHYSICS AND SPACE SCIENCE LIBRARY VOLUME 323

EDITORIAL BOARD Chairman W.B. BURTON, National Radio Astronomy Observatory, Charlottesville, Virginia, U.S.A. ([email protected]); University of Leiden, The Netherlands ([email protected]) Executive Committee J. M. E. KUIJPERS, Faculty of Science, Nijmegen, The Netherlands E. P. J. VAN DEN HEUVEL, Astronomical Institute, University of Amsterdam, The Netherlands H. VAN DER LAAN, Astronomical Institute, University of Utrecht, The Netherlands MEMBERS I. APPENZELLER, Landessternwarte Heidelberg-Königstuhl, Germany J. N. BAHCALL, The Institute for Advanced Study, Princeton, U.S.A. F. BERTOLA, Universitá di Padova, Italy J. P. CASSINELLI, University of Wisconsin, Madison, U.S.A. C. J. CESARSKY, Centre d'Etudes de Saclay, Gif-sur-Yvette Cedex, France O. ENGVOLD, Institute of Theoretical Astrophysics, University of Oslo, Norway R. McCRAY, University of Colorado, JILA, Boulder, U.S.A. P. G. MURDIN, Institute of Astronomy, Cambridge, U.K. F. PACINI, Istituto Astronomia Arcetri, Firenze, Italy V. RADHAKRISHNAN, Raman Research Institute, Bangalore, India K. SATO, School of Science, The University of Tokyo, Japan F. H. SHU, University of California, Berkeley, U.S.A. B. V. SOMOV, Astronomical Institute, Moscow State University, Russia R. A. SUNYAEV, Space Research Institute, Moscow, Russia Y. TANAKA, Institute of Space & Astronautical Science, Kanagawa, Japan S. TREMAINE, CITA, Princeton University, U.S.A. N. O. WEISS, University of Cambridge, U.K.

Recollections of “Tucson Operations” The Millimeter-Wave Observatory of the National Radio Astronomy Observatory

by M.A. GORDON National Radio Astronomy Observatory, Tucson, AZ, U.S.A.

A C.I.P. Catalogue record for this book is available from the Library of Congress.

ISBN 1-4020-3235-8 (HB) ISBN 1-4020-3236-6 (e-book)

Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands. Sold and distributed in North, Central and South America by Springer, 101 Philip Drive, Norwell, MA 02061, U.S.A. In all other countries, sold and distributed by Springer, P.O. Box 322, 3300 AH Dordrecht, The Netherlands.

Cover image: The 36-ft millimeter-wave telescope operated by the National Radio Astronomy Observatory on Kitt Peak, Arizona. Photo by J. Connors, Graphics Films, Inc., Los Angeles.

Printed on acid-free paper springeronline.com All Rights Reserved © 2005 Springer No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Printed in the Netherlands.

To Jim Warwick, who introduced me to radio astronomy.

Foreword The millimeter-wave radio telescope of the National Radio Astronomy Observatory (the NRAO) was one of the most successful telescopes ever built in the United States. Planned in 1962 and constructed on Kitt Peak, Arizona, this 36-ft parabolic antenna was not completed until 1967 because of the technical challenges associated with its novel design. Its early years produced few astronomically significant results. These observations consisted of continuum observations primarily at λ3 mm but also at λ1.2 mm. The problem was the low sensitivity of millimeter-wave radio receivers available at the time. What could be detected was just not interesting. In contrast, its later years were spectacularly productive. Starting in the 1970s, astronomers began detecting emission lines from interstellar molecules in these same wavelength bands. Unlike the earlier continuum observations, these results were completely new, creating a new way to explore the characteristics of cosmic objects. In time, they revolutionized our understanding of the nature of interstellar gas, chemistry at extremely low temperatures, and how stars form and galaxies evolve. The observations even provided a way to study the structure of the molecules themselves. For several years, the 36-ft telescope was in more demand than any other telescope in the United States, optical or radio. Apart from astronomy, the demands of millimeter-wave astronomy themselves stimulated developments in electronics and in computer software. These advances increased the sensitivity of the millimeter-wave telescope which, in turn, created new pressures for continued technical improvements and resulted in more astronomical discoveries. This symbiotic or, perhaps, “symtechnic” cycle is a hallmark of cutting-edge telescopes and continues today. With this success, the imperfections of the innovative 36-ft telescope became increasingly intolerable. Astronomers asked for improvements. Several attempts were made but most fell short of what was needed. To serve their vii

viii

FOREWORD

astronomer clientele better, the NRAO proposed as a replacement a 25-m telescope on the 13,792-ft Mauna Kea in Hawaii, but it was never funded. While awaiting a decision on the proposal for Mauna Kea, the NRAO replaced the venerable 36-ft surface of the Kitt Peak telescope with a low-cost but much more accurate 12-m surface as a stop-gap measure. Good as it was, this improvement fell short of what was needed. In the end, the results from these telescopes and similar ones elsewhere stimulated plans for an enormous synthesis array of 64 12-m paraboloid antennas at millimeter and submillimeter wavelengths, known as the Atacama Large Millimeter[-wave] Array or ALMA. The United States, Canada, Spain, and member countries of the European Southern Observatory (ESO) are funding this telescope jointly in cooperation with the Republic of Chile. Now under construction at a 16,500 ft site in northern Chile, this telescope is the child of the original 36-ft telescope and others and, of course, of the millimeterwave interferometers of the California Institute of Technology (Caltech), Berkeley–Illinois–Maryland (BIMA), and the Institut de Radio Astronomie Millimetrique ` (IRAM). Meanwhile, the 12-m millimeter-wave telescope has been turned over to the University of Arizona. Over the years, more than 240 NRAO employees have worked for Tucson Operations, including summer students. Based upon conversations with the original planners, assorted documents, and my personal involvement, this is the story of this innovative group. Many astronomers and NRAO employees read sections of the manuscript to verify the accuracy of my recollections. I have also discussed specific recollections with some of them. I do not cite their names individually because of the personal nature of this book but I am grateful to each of them. Editorially, I thank Ellen Bouton, Butler Burton, Colin Miller, and my wife, Julie, for reading the entire manuscript and giving me counsel regarding its content.

Contents Foreword 1 The 1.1 1.2 1.3 1.4

vii

Early Years The National Radio Astronomy A Millimeter-Wave Telescope . The 36-ft Telescope . . . . . . . The Kitt Peak Site . . . . . . .

2 Construction of the Telescope 2.1 The Rohr Corporation Design 2.2 Actual Construction . . . . . 2.3 Initial Performance . . . . . . 2.4 Initial Support for Operations

. . . .

Observatory . . . . . . . . . . . . . . . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

1 1 4 9 10

. . . .

13 13 15 19 27

3 Radio Lines from Molecules 29 3.1 The Gold Rush . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.2 Lingering Technical Problems . . . . . . . . . . . . . . . . . . 31 4 Dispatched to Tucson 35 4.1 Learning How the NRAO Functioned . . . . . . . . . . . . . . 35 4.2 My Research Interests Moving Toward Tucson . . . . . . . . 38 4.3 The Meeting . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 5 Expanding the Tucson Facilities 5.1 Physical Environment . . . . . 5.2 Relationship to KPNO . . . . . 5.3 Finding New Space . . . . . . . 5.4 New Mountain Laboratory . . . 5.5 New Operators’ Dormitory . . 5.6 The Sewage Crisis . . . . . . . ix

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

45 45 51 54 59 60 62

CONTENTS

x 5.7 5.8 5.9

The Fate of Our KPNO Office Space . . . . . . . . . . . . . . Moving to the University of Arizona . . . . . . . . . . . . . . New Astronomer Dormitories on Kitt Peak . . . . . . . . . .

64 64 69

6 Providing Adequate Electricity 73 6.1 External Power . . . . . . . . . . . . . . . . . . . . . . . . . . 73 6.2 Papago Tribal Utility . . . . . . . . . . . . . . . . . . . . . . . 74 6.3 Ground Currents . . . . . . . . . . . . . . . . . . . . . . . . . 76 7 Lightning and Kitt Peak

77

8 Software 8.1 The First Version . . . . . . . . . . . . . 8.2 Implementation of Green Bank Software 8.3 FORTH . . . . . . . . . . . . . . . . . . 8.4 The VAX Years . . . . . . . . . . . . . . 8.5 Moving to Unix . . . . . . . . . . . . . . 8.6 Off-Line Data Reduction . . . . . . . . . 8.6.1 Spectroscopy . . . . . . . . . . . 8.6.2 Continuum Mapping . . . . . . . 9 Millimeter-Wave Electronics 9.1 Local Oscillators . . . . . . . . . . 9.2 Quasi-Optical Techniques . . . . . 9.3 Receivers . . . . . . . . . . . . . . 9.4 Failures . . . . . . . . . . . . . . . 9.4.1 Millimeter-Wave Parametric 9.4.2 Millimeter-Wave Bolometer 9.4.3 The Hybrid Spectrometer . 10 Quantifying mm-Wave Astronomy

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . Amplifier . . . . . . . . . . . .

. . . . . . . .

. . . . . . .

. . . . . . . .

. . . . . . .

. . . . . . . .

. . . . . . .

. . . . . . . .

. . . . . . .

. . . . . . . .

. . . . . . .

. . . . . . . .

. . . . . . .

. . . . . . . .

. . . . . . .

. . . . . . . .

. . . . . . .

. . . . . . . .

79 79 80 80 83 85 86 86 87

. . . . . . .

91 91 94 97 99 99 101 103 107

11 Scheduling 111 11.1 The Initial Schedulers . . . . . . . . . . . . . . . . . . . . . . 111 11.2 Local Control . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 11.3 Paranoia and the Law of the Jungle . . . . . . . . . . . . . . 116 12 Improving Telescope Performance 119 12.1 “Foiling” the 36-ft Telescope . . . . . . . . . . . . . . . . . . 119 12.2 The Teepee . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

CONTENTS 13 The 13.1 13.2 13.3 13.4 13.5 13.6

xi

25-m Telescope What Should We Build? . . . . . . . . . . . Where Should We Build It? . . . . . . . . . Preparing the Formal Proposal . . . . . . . Negotiating for a Mauna Kea Site . . . . . . The Funding Process . . . . . . . . . . . . . A New 12-m Surface for the 36-ft Telescope

14 Odds and Ends 14.1 NSF Reorganization . . . . . . 14.2 Barry Goldwater’s Visit . . . . 14.3 The Chinese Visit . . . . . . . 14.4 AUI Board Meetings in Tucson 14.5 Changing NRAO Directors . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

125 . 125 . 127 . 129 . 130 . 140 . 147

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

161 161 163 164 166 169

15 The MMA and ALMA 173 15.1 The Millimeter-Wave Array . . . . . . . . . . . . . . . . . . . 173 15.2 ALMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 16 The 16.1 16.2 16.3

Twilight Years 185 Closing the 12-m Telescope—Part 1 . . . . . . . . . . . . . . 185 Closing the 12-m Telescope—Part 2 . . . . . . . . . . . . . . 188 Closing of the NRAO’s “Tucson Operations” . . . . . . . . . 189

A Time Line

191

B List of Tucson Employees

193

C Glossary

199

Name Index

203

Bibliography

207

List of Figures 1.1 1.2 1.3 1.4 1.5

Signing of the AUI-NSF Contract Dedication of the NRAO . . . . . Frank D. Drake . . . . . . . . . . Transmission of the Atmosphere . Isochasms of Cloud-free Days . .

. . . . .

. . . . .

. . . . .

4 5 6 8 11

2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10

Rohr Mill for the 36-ft Surface . . . . . . . . . . . . . . . The 36-ft Reflector En Route . . . . . . . . . . . . . . . The 36-ft Reflector on a Trailer . . . . . . . . . . . . . . Erection of the Astrodome Frame . . . . . . . . . . . . . Alignment of the Yoke and Alidade . . . . . . . . . . . . Back Structure of the 36-ft Telescope in 1967 . . . . . . The 36-ft Telescope Dome in Late 1967 . . . . . . . . . . The 36-ft Telescope in 1967 . . . . . . . . . . . . . . . . The Apex of the 36-ft Telescope with the First Receiver A Drive Wheel for the Astrodome . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

14 15 16 17 18 20 21 22 23 25

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12

The 36-ft Telescope in 1975 . . . . . . . The 36-ft Telescope in 1981 . . . . . . . The Nutating Subreflector in 1981 . . . The Control Room of the 36-ft Telescope Some of the Operations Staff in 1979 . . Some of the Technical Staff in 1979 . . . Entrance to the NRAO Offices on Forbes Aerial View of Forbes Industrial Park . . The “West Ridge” of Kitt Peak in 1972 New Operators’ Dorm . . . . . . . . . . 1984 Addition to Steward Observatory . The Back of “Trailer No. 2” . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

47 48 49 50 56 56 57 58 60 62 66 70

xiii

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . Dome in 1981 . . . . . . . . . . . . . . . . . . Boulevard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

LIST OF FIGURES

xiv 5.13 5.14

Entrances to Rooms 1 through 3 of the “New Dorm” . . . . The New Common Building . . . . . . . . . . . . . . . . . .

71 72

6.1

Electric Generators . . . . . . . . . . . . . . . . . . . . . . .

75

7.1

Summer Lightning on Kitt Peak . . . . . . . . . . . . . . . .

77

8.1

Mike Hollis and Maxine Thomas . . . . . . . . . . . . . . .

82

9.1 9.2 9.3 9.4

The Local Oscillator in 1972 . . . . . . Sketch of Quasi-Optic LO Injection . . SSB Receiver Temperatures Since 1971 Diagram of the New λ1 mm Bolometer

11.1 11.2

Dispersion of Referee’s Ratings . . . . . . . . . . . . . . . . 114 Selected Proposals . . . . . . . . . . . . . . . . . . . . . . . 115

12.1 12.2 12.3

Spherometer for Measuring Surface Figures . . . . . . . . . 120 Error Contours Projected onto the 36-ft Surface . . . . . . . 121 The Griffolyn “Teepee” on the 36-ft Telescope . . . . . . . . 123

13.1 13.2 13.3 13.4 13.5 13.6 13.7 13.8 13.9 13.10 13.11 13.12 13.13 13.14 13.15 13.16 13.17 13.18 13.19 13.20

The 25-m Millimeter-Wave Telescope . . . . . . . . . . . The Metal Astrodome of the 25-m Telescope . . . . . . . Two Spherometers on the Test Track . . . . . . . . . . . Artistic Rendering of the 25-m Telescope on Mauna Kea Island of Hawaii . . . . . . . . . . . . . . . . . . . . . . . Three-Dimensional Topo Map of Mauna Kea . . . . . . Access Road to Mauna Kea Summit . . . . . . . . . . . Photo of Access Road to Mauna Kea Summit . . . . . . Summit of Mauna Kea in 1977 . . . . . . . . . . . . . . Panoramic View of Mauna Kea Summit . . . . . . . . . Spherical Reflector Proposed by Frank Drake . . . . . . Collecting Areas of mm-Wave Single Telescopes in 2003 Removing the 36-ft Surface . . . . . . . . . . . . . . . . Installing the New 12-m Backstructure . . . . . . . . . . Attaching the Panels to the 12-m Backstructure . . . . . Mechanical Measurement Jig . . . . . . . . . . . . . . . Edge-on View of the 12-m Telescope . . . . . . . . . . . Back View of the 12-m Telescope . . . . . . . . . . . . . Performance of the 12-m Surface . . . . . . . . . . . . . Holographic Map of Damaged Reflector . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . . . . . . . . . . . . . . . . . .

. 93 . 95 . 100 . 101

. . . . . . . . . . . . . . . . . . . .

126 131 132 133 134 135 136 137 138 139 142 148 152 153 154 154 155 156 157 159

LIST OF FIGURES

xv

14.1

Transport to AUI’s “Wash Party” . . . . . . . . . . . . . . . 168

15.1 15.2 15.3 15.4

“Low” Portable infra-red Hygrometer . . . Composite View of Llano de Chajnantor . Fuzzy Logic Cost Estimates for the MMA Artistic Sketch of ALMA . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

177 178 179 184

List of Tables 11.1 Distribution of Time Scheduled on the 36-ft Telescope . . . . 113 13.1 Sites Considered for the 25-m Millimeter-Wave Telescope . . 128

xvii

Chapter 1

The Early Years “Radio astronomy” involves detecting and analyzing radio waves generated naturally by cosmic bodies. Like the optical waves our eyes detect (commonly called “light”), radio waves are part of the electromagnetic spectrum emitted by any warm material. Unlike light, radio waves require a different technology for detection. Tuned between broadcast stations, ordinary radio receivers detect this cosmic radio emission as a background signal called “static noise.” To astronomers, these radio waves carry astrophysical information that can help unravel the secrets of the universe in which we live, supplementing and complementing information carried by infra-red, optical, ultraviolet, Xray, and other parts of the cosmic electromagnetic spectrum. In practice, the term “radio” is imprecise. Officially, it refers to frequencies from 30 Hz (λ104 km)to 300 GHz (λ1 mm). Nowadays, radio astronomers include much higher frequencies, defined generally by the radio techniques used to detect them. Accordingly, the exact distinction between radio and far infra-red is somewhat blurred.

1.1

The National Radio Astronomy Observatory

World War II stimulated great improvements in the apparatus and techniques for detecting radio waves. Later, astronomers of many countries were quick to employ these new tools to pursue their studies of the cosmos. The results were exciting, producing discoveries impossible to make with traditional optical astronomy. In the early 1950s, the idea of a federally funded, national observatory for US radio astronomy attracted widening support [1, 2, 3, 4]. The initial idea was to create a large, centrally managed, general purpose radio telescope 1

2

CHAPTER 1. THE EARLY YEARS

that could be shared by astronomers from different institutions. According to Heeschen [3], to this end the National Science Foundation (NSF) established a committee of prominent astronomers and physicists (Bart Bok, Jesse Greenstein, John Hagen, John Kraus, Donald Menzel, Rudolf Minkowski, Ed Purcell, and Merle Tuve, chair) for advice on how best to promote the emerging science of radio astronomy. A similar process was underway for establishing a national optical astronomy observatory in the Southwest. In addition to the same “sharing” objective, dependably good weather was probably an additional consideration for the optical astronomers. Many of them belonged to institutions plagued by unpredictable, dismal weather. “In contrast to the optical astronomer’s smooth and businesslike progress toward their goal, however, the course of the radio group was rough and stormy” according to the NSF historian Merton England [2]. The problem lay in widely differing views on how a national radio observatory should operate. Donald Menzel wrote to Associated Universities, Inc., (AUI) on behalf of a Harvard-MIT group consisting of Bart Bok, Cecilia Payne-Gaposhkin, Julius Stratton, Fred Whipple, and Jerry Wiesner to see if AUI would develop a plan for such a facility [3]. AUI consisted of nine northeastern universities organized in 1948 to manage Brookhaven National Laboratory (BNL)—a high-energy physics laboratory on Long Island, NY—for qualified university physicists from anywhere in the US. Not only was this kind of operation what the Harvard-MIT group envisioned for radio astronomy, both Harvard and MIT were also members of AUI and so, in a sense, the new radio astronomy observatory would be an addition to the family. The president of AUI at that time, Lloyd Berkner, believed this would be a worthwhile venture for the consortium and promoted it aggressively. Merle Tuve promoted a contrary view. He felt that research flourished best in small settings like university departments or in ad hoc departments of large laboratories like the Naval Research Laboratory, where astronomers built and used their own telescopes. To him, a big-science, centralized observatory would not be in the best interest of a newly emerging branch of astronomy. Consequently, Tuve tried to block the formation of the big new observatory proposed by “a very small group of men” and managed by “selfapproving groups of ‘experts’ ” [2]. Furthermore, Berkner and Tuve not only held greatly different ideas of how research should be conducted but also did not like each other personally, which further widened the gulf between these two concepts for a national observatory. [3] In May, 1954, AUI hosted a conference at its New York City offices to

1.1. THE NATIONAL RADIO ASTRONOMY OBSERVATORY

3

consider the new observatory for radio astronomy. The participants recommended that AUI ask for funds from the NSF for a planning and feasibility study for a large radio astronomy facility, which they did. In response, the NSF committee “imposed a condition that the site be located within 300 miles of Washington, D.C. ... [probably] because all major optical telescopes were in the west and it was felt desirable to counterbalance that by putting the new radio observatory in the east.” [3] Of about 20 sites considered, Green Bank, West Virginia, was an easy choice because of its isolation from radio interference. The committee recommended funding, and in February, 1955, the NSF allocated $85k for a study. AUI submitted the study to the NSF in August, 1956, but Tuve continued to oppose its proposed scope. While Berkner envisioned the new observatory providing extensive equipment and services to facilitate observations by visiting astronomers, Tuve “wanted the observatory to consist of just a telescope, with minimal staff, facilities and services. Visitors would bring their own receiving equipment and staff, and even camp out in tents.” [3] The debates continued for months, with various scientists and government administrators expressing their individual, often conflicting views for how the new observatory should be operated. Nonetheless, the NSF negotiated an agreement with AUI for the new observatory and awarded a contract for $4M for construction in November, 1956. [2] (See Figure 1.1.) The Army Corps of Engineers acquired 2,700 acres (1,093 hectares) in Green Bank for the new observatory site over the next several years, while AUI began to construct the buildings and telescopes. The contract specified the construction of buildings and facilities and a 140-ft diameter radio telescope but did not mention other telescopes. The telescope would be designed primarily for centimeter wavelengths, that is, to operate at frequencies up to 30 GHz. “In fact, Alan Waterman, director of the NSF, specifically stated during a congressional hearing on the NSF budget that the new observatory would never need any other telescopes!” [3] Whatever the details of the process, the National Radio Astronomy Observatory (NRAO) had been finally created. Figure 1.2 shows the principals associated with the new observatory at the dedication of the NRAO on October 17, 1957. John Findlay and Dave Heeschen remained with the observatory for about four decades more.

4

CHAPTER 1. THE EARLY YEARS

Figure 1.1: The signing of the contract on November 17, 1956, between AUI and the NSF to construct and operate the National Radio Astronomy Observatory. Seated: Lloyd Berkner (AUI President) and Alan Waterman (NSF Director), Standing: left to right, Lee Anna Embrey (Assistant to the Director), Wilson F. Harwood (NSF Assistant Director for Administration), James M. Mitchell (Assistant to the Director), Franklin J. Callendar (Grants Administrator), C. E. Sunderlin (NSF Deputy Director), Charles B. Ruttenbery (Attorney), William J. Hoff (NSF General Counsel). [2]

1.2

A Millimeter-Wave Telescope

While the NRAO was being established on its radio-quiet site of Green Bank, West Virginia, some of its staff members [5] believed the new observatory should offer the astronomical community full coverage of the radio spectrum, that is, not only access to the short meter and long centimeter ranges where most of the “action” then existed in radio astronomy but also to the short millimeter wavelengths that lay near the infra-red band. If the new NRAO was to be an effective national radio astronomy observatory, it should cover all the “radio” wavelengths. At that time, other radio astronomical observatories were beginning to explore the millimeter-wave part of the astronomical spectrum and had con-

1.2. A MILLIMETER-WAVE TELESCOPE

5

Figure 1.2: Dedication of the National Radio Astronomy Observatory on October 17, 1957. From left: R. M. Emberson, L. V. Berkner, G. A. Nay, J. W. Findlay (seated), N. L. Ashton, D. S. Heeschen, and H. Hockenberry. A model of the 140-ft telescope is on the table in the background. NRAO photo. structed radio telescopes appropriate to this purpose. For example, in 1953 the world-famous P. N. Lebedev Physics Institute in Moscow had built a 22m dish whose surface had been manually adjusted by turning 40,000 bolts! It operated to wavelengths as short as λ8 mm. By 1959, an improved version of this telescope was being built in the Crimea to take advantage of better atmospheric transmission afforded by drier air. The NRAO did not have instrumentation to observe at millimeter wavelengths. A relatively new frontier, millimeter-wave astronomy used specialized equipment unknown to most practicing radio astronomers. At Texas Instruments, Frank Low had developed germanium bolometers cooled by liquid helium, which NRAO astronomer Frank Drake knew about. Flying to Dallas, Drake persuaded Low to leave his well-paid, comfortable position to move to Green Bank in 1961, where he could use his bolometer to detect cosmic radio emission [6]. With Low’s arrival in Green Bank, work on a millimeter-wave receiver

6

CHAPTER 1. THE EARLY YEARS

Figure 1.3: Frank D. Drake, the “father” of NRAO’s venture into millimeterwave astronomy, in Green Bank in 1962. The 300-ft radio telescope lies in the background. NRAO photograph.

1.2. A MILLIMETER-WAVE TELESCOPE

7

proceeded rapidly. Together, Drake and Low acquired a 5-ft diameter spuncast paraboloid and began to experiment with it at millimeter wavelengths. Drake recalls being amazed by the tiny microscopic feeds1 required for those wavelengths. In the winter of 1961-62, the new receiver was tested. While they did detect Jupiter, the atmospheric absorption was simply too great for useful work [7] even on clear, cold winter days when most of the atmospheric water vapor had frozen into crystals. The collecting area of the dish was too small and the amount of residual water vapor, too large, to conduct useful millimeter-wave astronomy from Green Bank. An effective millimeter-wave telescope would require a better site. “Better” meant fewer atmospheric absorbers along the line of sight, that is, much lower precipitable water vapor and, also, fewer oxygen molecules as is characteristic of a dry, high altitude site. Figure 1.4 illustrates the situation quantitatively. Observing prospects from the low level Green Bank site were not good except in winter—and not very good then either. Daily activity in the new high-tech observatory could be quite different than the agricultural life of the nearby villages. Here is an anecdote told to me by Dewey Ross. In 1960, Dewey was a single, 23-year old electronics technician at the NRAO. Later, in September, 1969 [8], after earning a degree in electrical engineering, he became one of the first NRAO engineers permanently assigned to Tucson. The new NRAO electronics laboratory in Green Bank used liquid helium (boiling temperature of 4 K) and liquid nitrogen (of 77 K) to cool sections of its ultra-sensitive radio receivers. To control the boil-off rate of these cryogenic liquids and to reduce the formation of ice plugs from atmospheric moisture, it was standard practice to install a perforated condom on the exit tubes of the dewar vessels—in effect, huge stainless-steel Thermos bottles— holding the liquids. Because of the extremely low temperatures, the latex condoms had a short lifetime and required frequent replacements. One of Dewey’s jobs was to buy condoms from the Marlinton, West Virginia, Rexall drug store every month—by the gross. The NRAO fiscal office was reluctant to order the condoms because they did not want these listed as entries in their federally audited books. Consequently, they asked Dewey to pay for them with his own money and to present the receipts for reimbursement as “miscellaneous supplies.” 1

“Feed” is the common name for the small, specially designed device at the focus of a parabolic reflector through which radio waves enter the radio receiver. This device is a small horn antenna.

8

CHAPTER 1. THE EARLY YEARS

Figure 1.4: The transmission of the Earth’s atmosphere as a function of frequency for three amounts of precipitable water vapor. The regions of high transmission are called “windows” because they allow cosmic radiation to reach the Earth’s surface. The legend indicates the approximate altitudes associated with these vapor levels for middle latitudes. Marked above the top ordinate are three rotational levels of carbon monoxide that might be observed from astronomical objects at millimeter and sub-millimeter wavelengths.

But Marlinton was a small village of about 500 people. Word regarding unusual activity quickly got around. By 1961, Dewey had become engaged to a Marlington girl—later, his wife—and was especially concerned that the villagers remain unaware of these massive purchases. According to his wife, Kay, [9], one day the local constable walked up to Dewey and said to him, “I know what you’ve been buying. I’m watching you carefully and you’d better behave yourself.” Dewey was horrified. Immediately, he pleaded with the NRAO fiscal office to place and receive these orders themselves. Evidently, Fiscal finessed this problem by finding a supplier of small balloons especially designed for use with cryogenic dewars. By the time Frank Low had arrived in Green Bank in late 1961, the NRAO lab no longer used condoms.

1.3. THE 36-FT TELESCOPE

1.3

9

The 36-ft Telescope

In 1962, the NRAO staff presented a proposal to the board of AUI to build a large millimeter-wave telescope on an acceptable site. It would cover a wavelength range of λ1 to λ7 mm. AUI approved the project and authorized the NRAO to proceed. Frank Drake recalls writing a part of the annual NRAO proposal to the NSF for $1.5M for construction, a justification that he remembered as consisting of a single paragraph [5]. He arrived at the estimate by scaling the cost of the 85-ft Tatel telescope at Green Bank by the factor (diameter)3 /(wavelength)1/2 —a calculation that shows the cost of a radio telescope to be about the price of hamburger per pound [6]. AUI submitted the proposal to the NSF in 1962, which funded the project. There were no peer reviews. Conceptually, the technical design of the mm-wave telescope had begun to gel. Frank Low heard that a naval shipyard in Newport News, Virginia, was machining ribs for the pressure hulls of nuclear submarines. Reportedly, the shipyard could machine pieces as large as 36 ft in diameter to accuracies of a few thousandths of an inch. The high accuracy of these circular ribs was needed to maximize the crush depths of the submarines. Frank remembered driving to the shipyard from Green Bank through a snow storm to inspect that facility. Indeed he found the shipyard to have that capability and that’s why the diameter of new mm-wave telescope was eventually chosen to be 36 feet [7]. Drake remembered the origin of the 36-ft diameter somewhat differently [6]. Wanting to build the largest possible millimeter-wave telescope, he thought $1.5M would be a good estimate for the cost of a first-class facility. Using the “hamburger” formula described above, he calculated the corresponding diameter to be about 36 ft. On the other hand, Drake noted that memories can be unreliable after more than 40 years! A little later, the NRAO awarded a contract to Rohr Corporation for a feasibility study. Rohr’s report stated that it should be possible to build a telescope of that size with an RMS surface accuracy of 0.002 in (50 µm), good enough to support useful observing at λ1 mm (300 GHz). In 1964, armed with the Rohr feasibility study, the NRAO filed a request for proposals for construction of the millimeter-wave telescope. There were several bidders, including Newport News Shipbuilding, North American, Philco-Ford, RCA, and Rohr. After reviewing all proposals, the NRAO awarded a contract to Rohr for $600k to build a 36-ft telescope on Kitt Peak. The telescope geometry itself was unusual. NRAO astronomer Peter

CHAPTER 1. THE EARLY YEARS

10

Mezger preferred a Cassegrain telescope2 but Low’s new bolometer required a prime focus instrument. The optics of the prime-focus mm-wave bolometer being developed at Green Bank also required a large focal ratio (focal length divided by diameter, or f/D). Frank Low wanted an f/D = 1 whereas most dish-type radio telescopes have f/D ≈ 0.4. An f/D = 0.8 was selected as a compromise. For comparison, remember that modern, film camera lenses have f/D ≈ 1.4 at maximum aperture. The surface accuracy was chosen to be as small as believed technically possible at the time for a 36-ft reflector, 0.002 in (50 µm) RMS, to facilitate efficient observing at λ1.2mm (230 GHz). If achieved, this would give a figure of merit of 216,000 (D/RMS accuracy)—better than any radio telescope had achieved at that time. From the beginning, the NRAO staff considered the new telescope to be a continuum instrument only. No capability for spectroscopy was even considered [7]. Later, in 1964, Peter Mezger looked into other receivers that might be appropriate for the new telescope [11] and, in 1967, NRAO engineer Rama Menon bench-tested a λ9 mm radiometer in Green Bank, but these were intended only for continuum observations [12].

1.4

The Kitt Peak Site

To be effective, the new telescope would require a much drier site than Green Bank. Figure 1.5 shows why the southwestern Unites States was attractive. Nick Mayall, then director of Kitt Peak National Observatory (KPNO), invited the NRAO to build the new telescope on Kitt Peak, a 6,900-ft arid mountain southwest of Tucson, Arizona, which served as the observing site of the new national optical observatory—the sister observatory to the NRAO. KPNO offered infrastructure that included a paved access road, electricity (to the summit, at least), telephones, dormitories, cafeteria, well-equipped repair shops, and a large on-site maintenance staff. Associated Universities for Research in Astronomy (AURA), a similar organization to AUI, managed KPNO. Some of the NRAO staff members suggested that the 9,200-ft summit of Mt. Lemmon near Tucson might be a better choice. At mid-latitudes, atmospheric water vapor has a scale height of about 1.6 km, so that a Mt. Lemmon site could have approximately 50% better transparency at millimeter wavelengths than the summit of Kitt Peak. 2

Guillaume Cassegrain was a French astronomer who invented this design in 1672 [10]

1.4. THE KITT PEAK SITE

11

Figure 1.5: Isochasms of the annual number of cloud-free days in the southwestern United States. Here a “clear day” is one with 0 → 0.30 mean sky cover. [13]. These data also roughly indicate the driest areas, that is, areas with the lowest annual rainfall. Tucson sites rank high in the data set.

Despite its higher altitude and correspondingly drier air, and equally good access via an all-weather paved highway, the Santa Catalina mountain range in which Mt. Lemmon is located has a political liability that may have been considered at the time. It was (and still is) the site of several powerful communications transmitters that might interfere with the operation of the new radio telescope. The co-existence of a new radio telescope with commercial transmitters would send the misleading message to U. S. and international frequency allocation officials that any radio telescope could cope with nearby radio transmitters. The special circumstances making millimeter-waves more difficult to interfere with compared to centimeter and longer wavelengths would surely be lost in the allocation processes. An unfortunate result could be that radio astronomers would weaken their claim of needing radio-quiet sites for their observatories as well as protected radio bands for observing. Also, the cost of developing a new site on Mt. Lemmon would be much

12

CHAPTER 1. THE EARLY YEARS

greater than on Kitt Peak. In 1962, the Mt. Lemmon site had no support facilities suitable for astronomy; the NRAO would have to provide everything themselves. Building this infrastructure could raise the cost significantly for what was to be, at that time, largely an experimental telescope to explore a new wavelength range of unproven value to radio astronomy. Both sites benefited from Tucson’s excellent access by commercial airlines, which would make it easier for visiting astronomers to get to the new telescope from their home institutions. In fact, travel to Kitt Peak would be much easier for most of the widely distributed US astronomers than to the more remote village of Green Bank. In the end, the NRAO chose a site on the southwest ridge (6,300 ft) of Kitt Peak mainly because of the infrastructure offered by KPNO [14]. Dave Heeschen and Nick Mayall worked out an agreement for support of the 36-ft telescope, which was subsequently approved by both the AUI and AURA boards. [15].

Chapter 2

Construction of the Telescope In the 1960s, the bidding and construction processes were more informal than they are today. The National Science Foundation was a much smaller agency. There were far fewer professional astronomers and, accordingly, requests for grant money were also fewer. It was quite normal for the buyer and the contractor to cooperate closely.

2.1

The Rohr Corporation Design

When the NSF authorization and allocation arrived, the NRAO director, David Heeschen, appointed the NRAO’s deputy director, John Findlay, to be the project manager. Unlike Low, Findlay believed the design and construction work was better suited to an aerospace firm than to a naval shipyard [7] and, in 1964, the Rohr Corporation of Chula Vista, California, won the contract–probably to its later regret. Rohr had just constructed a smaller mm-wave telescope for The Aerospace Corporation and, accordingly, had some experience with this kind of project [14]. The Rohr plan was to weld aluminum plates together and machine this “mirror blank” as one piece on a specially built lathe. A steel “backup” structure would support this reflector. The telescope mount would be altitudeazimuth, with a computer and clock converting the celestial coordinates of hour angle and declination to local altitude and elevation. Such a design was a new direction for the NRAO, which had long been wrestling with problems associated with the construction of the 140-ft equatorial telescope in Green Bank. 13

14

CHAPTER 2. CONSTRUCTION OF THE TELESCOPE

Figure 2.1: The milling machine of the Rohr Corporation used to cut the back structure (shown) and, eventually, the surface of the 36-ft millimeterwave telescope.

A large pipe-framed, fabric-covered dome would shield the telescope from wind to minimize pointing problems and from direct sunlight to minimize thermal distortions of the reflector surface. Unlike a radome, this dome would have a 40-ft openable slit to provide an unobstructed view of the heavens. It was to be a large version of the astrodomes used for optical telescopes. Its giant slit would track the azimuth of the radio telescope to provide a sheltered but unobstructed view of the sky. When the weather turned bad, the slit could be quickly closed to protect the telescope from wind, snow, or hail. Observing through the steel-framed dome would still be possible but with a great loss of sensitivity.

2.2. ACTUAL CONSTRUCTION

15

Figure 2.2: The 36-ft reflector near Winterhaven, California, en route from the Chula Vista factory to Kitt Peak on 16 February 1966. A yellow tarpaulin protected the surface. Hydraulic cylinders on the truck bed could tilt the reflector to ease difficult passages. John Hungerbuhler photo.

2.2

Actual Construction

The manufacturing plan for the mirror proved to be a disaster [16]. After the machining process had begun, the cutting tool cut a hole through the pieced aluminum surface. Rohr engineers welded the hole shut and began the machining process again. Unfortunately, the weld was harder than the original surface, and Rohr had to design special cutting bits to continue the cutting process. Ironically, even astronomy became an obstacle. The Rohr factory lay near the beach, which rhythmically swelled and contracted with the daily cycle of ocean tides. The axes of the lathe correspondingly changed with respect to the local gravity vector, creating surface errors thereby induced by the moon. The fix was to synchronize the machining process with the position of the moon and, hence, with the coastal tides. Ned Conklin [17] notes that Rohr built a better telescope than they realized. In the extended shutdown of the summer of 1971, “an engineer

16

CHAPTER 2. CONSTRUCTION OF THE TELESCOPE

Figure 2.3: The newly milled 36-ft reflector stored on a trailer at the base of Kitt Peak in 1967, awaiting completion of the support tower before it could be installed. KPNO photo 3959 [Bill Horne] from Green Bank made the very serious mistake of disconnecting both sets [actually, one set [18]] of guy wires from the large and heavy (2,500 lbs ??) bipod that supported the receivers at the prime focus. The bipod fell over, struck the dish heavily on one side and somersaulted onto the floor. Nobody was hurt, but the rim of the dish was dented and we all thought its figure had been destroyed. In an incredibly short time (a week or so) the bipod and associated wiring were repaired and we made a long series of measurements at 3 mm. There was no detectable degradation in aperture efficiency.” To begin the preparation of the site on the southwest ridge of Kitt Peak and construction of the astrodome, John Findlay and an NRAO civil engineer, Sidney Smith, traveled from Green Bank to Tucson in October, 1968 [18]. John drove a stake in the ground on the Kitt Peak ridge and said, “Put the antenna here.” The KPNO mountain manager told Findlay that he could route the access road through the KPNO picnic area. Returning to Kitt Peak on October 21, Sidney laid out the site in detail. He made about for eight trips between then and June 24, 1966. On one of

2.2. ACTUAL CONSTRUCTION

17

Figure 2.4: Erection of the astrodome for the 36-ft telescope in 1967. The workman on the top of the structure illustrates the scale. KPNO photo 7954 those trips, he learned that KPNO had changed its mind regarding routing the access road through the picnic area, and the access route was changed to the present one—at an extra cost to the project. Eventually, the NRAO engineering supervisor, Hans “John” Hungerbuhler, arrived for a six-month stay from which he never returned to Green Bank. Hungerbuhler was an extraordinary individual. Swiss-born and educated, he was an excellent linguist, quick of mind, and possessed a salesman’s persuasive tongue. It must have been a turbulent time in his life. Instead of overseeing the telescope project, he spent most of his time playing music in Tucson bars and cabarets. Consequently, the telescope project proceeded more or less unattended [19, 20]. In 1965, project manager Findlay left to become the site director of the recently completed (in early 1964) 1,000-ft radio/radar telescope in Arecibo, Puerto Rico. Responsibility for completing the troubled 36-ft telescope

18

CHAPTER 2. CONSTRUCTION OF THE TELESCOPE

Figure 2.5: Alignment of the yoke and alidade structure of the 36-ft telescope in 1967. Note the theodolite at the bottom of the yoke. KPNO photo 3955

project then fell to Hein Hvatum, the NRAO’s associate director for technical services in Charlottesville. Meanwhile, Frank Low had moved to Tucson in anticipation of the telescope’s completion. During the long delay, his interest had turned to even shorter wavelengths, infra-red astronomy. He had joined the University of Arizona astronomy faculty to pursue this new research area. Figure 2.2 shows the reflector en route from the Rohr factory at Chula Vista, California, to Kitt Peak. Hydraulic cylinders mounted on the truck bed supported the reflector. These could tilt the reflector to ease difficult passages. The trip began on February 14, 1966 and took three days, with overnight stays at rest areas along the route. Eventually, on February 17, 1966–rather late compared with the original estimates–the completed telescope arrived and was stored on a trailer at the base of Kitt Peak, guarded by Pinkerton watchmen, as shown in Figure 2.3.

2.3. INITIAL PERFORMANCE

19

It could not be installed on its site because the dome and telescope mount had not been completed. When these components were completed in 1966, the telescope was installed at the site. The first astronomical observation was a temperature profile of the moon during an eclipse in October, 1967. The receiver was a λ9.5 mm, prime-focus receiver. Neil Albaugh, Don Cardarella, George Grove, Hein Hvatum, Emily Kitchen, Don Logan, and Bill Terrell (chief telescope operator) were present [21]. The observations used the 12-in optical telescope to guide the 36-ft telescope. An Esterline-Angus chart recorder recorded the observations.

2.3

Initial Performance

The telescope went into regular service in 1968. Tests showed it achieved almost none of its original design goals. The surface accuracy was poor and, because of its one-piece construction, could not be adjusted. It was good enough to support observations at λ3 mm but disappointing at the design wavelength of λ1.2 mm, where the aperture efficiency was only 1015%. Nonetheless, Frank Low did observe regularly at λ1.2 mm with his bolometer [17]. Ned Conklin also pointed out that sensitive receivers weren’t available at these short wavelengths so that the lower surface accuracy was not the only impediment to successful observing at that time. The combination of aluminum surface and steel back structure caused the mirror to distort like a bimetallic strip. By mounting temperature sensors in suitable locations, Ned Conklin developed an algorithm predicting the focal length from the temperature difference between the front and back of the dish surface and, correspondingly, repositioning the prime focus receiver to the correct focus. There were big problems with the tracking computer. The 36-ft telescope, an alt-azimuth design, required a computer program to convert the altitude-azimuth coordinates to the astronomical ones of declination and hour angle. A functioning pointing program was as essential to this telescope as the structural components. There was no way the telescope could point at astronomical objects without it. For this purpose, the telescope control system employed a Honeywell DDP-116 computer. The Computer Control Company in Massachusetts had developed this computer but the company was purchased by the Honeywell Corporation. According to Johann Schraml [22], a German astronomer working in Green Bank at the time, Wally Weller—an astronomer—wrote the first control program, which was so complicated that it could not even perform the

20

CHAPTER 2. CONSTRUCTION OF THE TELESCOPE

Figure 2.6: A view of the back structure of the 36-ft telescope in 1967. The slit of the dome is closed. The top of the steel tower beneath the yoke shows a section of the azimuth torque motor incorporated into the telescope mount. Also shown is the optical telescope used for determining the gross pointing of the telescope. The adjustable lead counterweights are mounted on the semi-circular truss beneath the back structure. In the distance on the right is the electrical control panel for the motors that rotated the dome and opened the slit. At left is the original small control room. Behind the control room is the “yellow peril,” the name given to the electrical ladder affording access to the prime focus and to the top of the dome.

2.3. INITIAL PERFORMANCE

21

Figure 2.7: The 36-ft telescope in 1967. The astrodome protected the telescope from direct sunlight and winds.

fundamental task of pointing the telescope. To fix this problem, Hein Hvatum hired Emily Kitchen to write a simpler program from scratch, which did work. The matching of the complexity of a real-time computer program to the computer’s ability is not a trivial skill, especially for an astronomer rather than a trained software engineer. Often, the code is written at the same time as the selection and purchase of the computer. Usually, the time demands of the calculations are not known until the computer is actually connected to the device it is intended to control. The most spectacular case I know about is Robert Hjellming’s attempt to write the data processing program (named “CANDID”) for the Very Large Array telescope (VLA) while the telescope was under construction. Hjellming, a brilliant astronomer and skilled programmer, spent a few years leading this project but underestimated the timing requirements of the VLA. The DEC 10 computer and

22

CHAPTER 2. CONSTRUCTION OF THE TELESCOPE

Figure 2.8: The 36-ft millimeter-wave telescope near the end of 1967 within its fabric-covered astrodome, with guy wires stabilizing the bipod receiver mount. The pipe carried cooling air to the prime-focus receiver, which was bolted to the adjustable Stirling mount—a smaller version of the ones used in other NRAO radio telescopes.

2.3. INITIAL PERFORMANCE

23

Figure 2.9: The first receiver of the 36-ft telescope, mounted at prime focus. The pipe carries cooled air. The motors of the Stirling mount are clearly seen, as are the guy wires that stabilize the bipod mount. Mechanically tuning the klystron used for the local oscillator involved reaching into the receiver enclosure from the top and turning adjustment bolts.

24

CHAPTER 2. CONSTRUCTION OF THE TELESCOPE

its program (written in the SAIL computer language) could not keep up with the data flow, even from the initial 6 of the planned 27 antennas. At the eleventh hour, Barry Clark, Dave Ehnebuske, and Jerry Hudson (aided by Carl Bignell and Nancy Vandenberg) solved the problem by hastily designing a simple, ad hoc program that did work but omitted many of the sophistications embodied in Hjellming’s original computer code [23]. Ned Conklin [17] remembers: “The problem with Emily’s program was that it was all in assembly code. It ran fine, but changes were just about impossible to make, requiring a very skilled programmer conversant with the DDP-116 instruction set and lots of time (more than a single maintenance day). I remember that it had a crude continuum data-taking subroutine which we never used, because it typed out the data in neat columns on the teletype console, laboriously spacing over the page and taking many minutes to print even a very short run, during which time you could do nothing else. And modifying the print out routine would have been a major task in assembly code.” After the program was fixed, the staff discovered that the servo system could point the telescope adequately but had a lot of problems with its transient performance [17]. In addition, the azimuth and elevation axes weren’t quite perpendicular, requiring elaborate corrections to the pointing. Ned recalled that the situation improved to an RMS pointing error of about 10-15 arcseconds RMS by the early 1970s. Fixing the pointing proved to be a long-term project for the Tucson staff, lasting through the 1990s when Phil Jewell and Jeff Kingsley were able to solve the problem decisively with the help of the stable new 12-m surface and with new inductosyn shaft-angle encoders engineered by Bob Freund. The specially built prime-focus receivers had high noise temperatures and could only detect planets in the solar system, observations that quickly reached a sensitivity limit. By 1969, receivers had improved sufficiently to support routine observations of stronger continuum sources such as QSOs, Seyferts, BL Lac objects, and E/S0 galaxies, etc [17]. By 1970, the telescope was used every night except during summer months when the elevated humidity sharply reduced atmospheric transparency at mm-wavelengths. Daytime observing was less common because direct sunlight distorted the reflector. The dome would not track the telescope position until an ingenious system of infra-red detectors was installed on it, which signaled which direction the dome motors had to turn. This system consisted of a wide-angle infrared transmitter mounted on the telescope. Variable-speed motors replaced the off-on ones that had been delivered with the telescope (see Figure 2.10).

2.3. INITIAL PERFORMANCE

25

Figure 2.10: One of the four truck wheels that drove the dome. Attached to the movable dome, the variable-speed motor drove an 8-ply truck tire (within the protective steel cage) along the circular concrete stem wall, which turned the dome in azimuth. Orange insulating curtains hanging from the dome (here peeled away for inspection) covered the electrical slide conduits to protect personnel. Author photo

When the telescope turned, the changing infra-red signal detected by the two dome sensors activated the variable-speed dome motors. The relative amplitudes detected by the dome sensors determined the direction to drive the dome, and a servo loop accelerated and de-accelerated the dome until both sensors detected a signal of equal strength. This simple, reliable system synchronized the dome azimuth to that of the telescope. Some of the problems stemmed from innovative engineering. One example is the electric torque motor for the azimuth drive. The torque produced by this type of motor is linearly proportional to the applied drive current, facilitating excellent servo control—in theory. The entire top of the azimuth mount was an integral part of this motor; no gear reduction was used. But this innovative motor carried an unexpectedly high price. To make the azimuth mount as stiff as possible, Rohr selected a large diameter

26

CHAPTER 2. CONSTRUCTION OF THE TELESCOPE

tapered-roller bearing made by the Kadon Corporation, originally designed to support the turrets of army tanks. The NRAO soon discovered that the integral torque motor caused a temperature differential between the outer and inner races of the bearing, causing the bearing torque to vary with temperature, making a reliable tight servo loop impossible to implement, and consequently destroying any possibility of accurate pointing in azimuth. Sometimes, the bearing actually seized. The solution was to install temperature-controlled heating strips on both sides of the azimuth bearing. In this way, the torque required to drive the azimuth bearing stayed constant. The geometry of the azimuth torque motor required the azimuth braking to consist of a large diameter (4 or 5 ft) brake rotor bolted to the azimuth bearing and equipped with horizontally mounted hydraulic brake calipers. These particular aircraft brake calipers were never designed to operate in a horizontal position. Consequently, they often leaked, spreading hydraulic fluid on the disk rotor and destroying the braking action. On more than one occasion, a burst of wind blew through the slit of the astrodome, impacted on the dish surface as if it were a sail, and rotated the telescope well away from its commanded position. The slick brake rotor could not stop the winddriven slewing, and the telescope rotation ripped out the cables connecting the receiver to the ground. The fix was to replace the original 1960s brake calibers with improved ones that were designed to operate in a horizontal position and to install a huge sump for brake fluid just in case the calibers did leak, a solution implemented in the mid-1970s by Neil Albaugh. In passing, I note that Albaugh had a penchant for using parts stripped from disassembled military aircraft, which were abundant in Tucson because the storage “boneyard” of Davis-Monthan Air Force Base. Albaugh’s design for the brake system used an aircraft hydraulic sump and hundreds of rivets. I suspect the brake system never needed service again. In retrospect, these failures could sometimes be funny—at least to the support staff. In December, 1973, during one observing run over a weekend, Lew Snyder (University of Illinois) and Dave Buhl (NRAO) were using the 36-ft telescope to search for molecules, when a burst of wind caught the telescope and slewed it rapidly in azimuth because the fluid-slick brake rotor could not stop the rotation. The uncontrolled slew ripped out most of the cabling, rendering the telescope completely useless. Our crew responded instantly, driving the 50+ miles from their Tucson homes to the telescope to make repairs. All except John Payne and me, who were having dinner at my house. It took us a little longer to get there because we decided to finish our meal, knowing the crew was already on its

2.4. INITIAL SUPPORT FOR OPERATIONS

27

way. When we finally arrived, we discovered that Lew had chalked on the blackboard the names of the repair crew in the order they arrived on the scene. We restored the telescope to full operation in a few hours—an unexpectedly short time for such a disaster. When his research paper appeared in The Astrophysical Journal [24], I discovered that Lew had acknowledged the NRAO rescue by listing these names in order of arrival. John and I were listed dead last among the repair crew, just ahead of the two telescope operators.

2.4

Initial Support for Operations

Despite the problems with performance, the NRAO remained optimistic regarding the eventual usefulness of the new telescope. It staffed the telescope to assist visiting astronomers, but minimally. George Grove, an experienced telescope operator moved from Green Bank to Tucson. Other permanent employees were Bill Terrell, Don Cardarella, and Ralph Burhans. Neil Albaugh and Don Logan came to the site frequently to cable the receiver systems for the first observations, which occurred in October 1967 [21]. By 1970, the local staff had increased to six. While Nick Mayall was director, KPNO went out of its way to accommodate the new NRAO observatory. By agreement worked out between observatory directors Nick Mayall and Dave Heeschen and ratified by the boards of AURA and AUI [15], the KPNO mountain staff would be available for major repairs when needed and when available, on a charge-back cost basis. Moreover, the NSF increased their annual allocation to KPNO by $25k per year to pay for the additional overhead (telephones, copiers, cleaning services, pencils and pens, etc.) of supporting the NRAO telescope— approximately the salary of an experienced engineer, a significant sum in the mid-1960s. To accommodate visitors, the NRAO transferred money to KPNO to pay for two rooms in a new dormitory (there were others) being constructed on the summit. To contribute to the Kitt Peak infrastructure as well as to provide power to the 36-ft telescope, the NRAO built a high capacity high-voltage power line from the summit to the southwest ridge where the 36-ft had been sited. This power line was sized to supply additional telescopes on that ridge, if they were to be constructed. Title was transferred to KPNO and, later, to the electric company providing power to Kitt Peak. Some office space was also needed in Tucson. Initially, the KPNO build-

28

CHAPTER 2. CONSTRUCTION OF THE TELESCOPE

ing on Cherry Avenue had no extra space so the NRAO operated from a “store front” in the Campbell Plaza Shopping Center on North Campbell Avenue, sharing an area with KPNO’s Space Sciences Division [21] . KPNO expanded their Tucson headquarters in late 1967, and the NRAO Tucson group moved in. The NRAO paid KPNO $60k for 1,200 ft2 of basement area in the new addition to be used for offices. The high cost (at that time) of $60/ft2 came from the design of the building. Although a building with only a ground floor and a basement, the architects designed it to support additional floors if KPNO chose to expand in later years.

Chapter 3

Radio Lines from Molecules The huge demand for the telescope came later, in 1970, when radio emission from molecules was being detected from the interstellar gas. Optically detectable molecules from stars and interstellar gas had long been known. Most of these were seen in absorption in the atmospheres of stars. While clearly identified, these spectral lines were usually well broadened by the atmospheric pressure of the star. One absorption complex, known as the interstellar absorption bands, was ubiquitous in spiral galaxies but could not be identified with any particular molecule or molecules. Consequently, astronomers found these lines interesting but difficult to use as general tools for the exploration of galaxies. In 1944, at a Dutch astronomy meeting held during World War II, Henk van de Hulst [25, 26] predicted the spin-flip spectral line of atomic hydrogen at λ21 cm could be detected in astronomical sources. It was later detected in 1951, [27, 28] providing a great deal of new information regarding the structure and kinematics of galactic gas ever since. The 1964 detection1 of radio recombination lines (RRLs) from atomic hydrogen had also proved to be a powerful tool to investigate the interstellar gas, this time, the ionized rather than the neutral part component of the gas. The small optical depth of these lines allowed astronomers to probe the full extent of sight lines through the diffuse interstellar gas. Similarly, if spectral lines from multi-atom compounds—molecules—in the interstellar gas could also be detected, they would not only complement the astronomical information given by the two types of atomic spectral lines described above, but also give new information about atoms other than 1 see §1. .2 of Gordon astronomical s.

orochen o [2 ] for the complicated story of the detection o f

29

CHAPTER 3. RADIO LINES FROM MOLECULES

30

hydrogen and about the chemistry of the interstellar gases.

3.1

The Gold Rush

Astronomers had realized for many years that the detection of molecular lines in the radio spectrum might provide an excellent astronomical tool. In 1957, Charles Townes even prepared a list of likely candidates [30]. From the point of view of abundance, the best candidates for detection were the rotational transitions of ground vibrational states of simple (diatomic) molecules or ions containing cosmically abundant light elements, such as hydrogenic molecules. Their smaller moments of rotational inertia positioned these lines at millimeter wavelengths. Unfortunately, the sensitivity of radio telescope systems at these wavelengths wasn’t very good in 1957. Partly because of dramatic increases in sensitivity of radio telescopes at centimeter wavelengths, the first searches for radio molecular lines in astronomical sources took place there. In 1963, Weinreb et al. [31] discovered emission from the hydroxide radical (OH) with the Millstone radio telescope of MIT Lincoln Laboratory at λ18 cm (1.6 GHz). In 1968, Cheung et al. at UC Berkeley discovered astronomical radio emission from ammonia (NH3 ) at λ1.25 cm (23.6 GHz) [32] and, in 1969, water (H2 O) at λ1.35 cm (22.2 GHz) [33]. In 1969, Snyder and colleagues [34] discovered emission from interstellar formaldehyde (H2 CO) at λ6 cm (5 GHz) with the 140-ft telescope of the NRAO at Green Bank, West Virginia. While astronomers were looking for molecules in the centimeter wavelength range, sensitivities of radio telescopes also improved at millimeter wavelengths, facilitating searches for the putatively more abundant lighter molecules. In 1970, Robert Wilson, Keith Jefferts, and Arno Penzias announced [35] that, with the NRAO 36-ft telescope, they had detected carbon monoxide (CO) from the cold gases of the Milky Way–our own galaxy. Ned Conklin remembers the signal being so strong that no integration was necessary to see the spectrum [17]. They quickly rigged up a remote real-time spectrum display at the telescope operator’s console and used the display to peak up the telescope position on the CO signature directly. That announcement quickly transformed the under-utilized 36-ft telescope into one of the most sought-after telescopes in the United States, optical or radio. This discovery, coupled with the earlier radio astronomical detections of OH, H2 O, NH3 , and H2 CO, promised a new method of studying the interstellar gas2 . Such molecules formed and existed in the 2

Lew Snyder once suggested that the decision of the Bell Labs group to search for CO

3.2. LINGERING TECHNICAL PROBLEMS

31

cold gas from which stars form, but which the very same coldness made almost impossible to observe in emission in the optical regime. Also exciting was the opportunity for studying chemical reactions in conditions that could not be easily created in terrestrial laboratories, such as low-pressure and low-temperature gas-phase chemistry. At that time, many young astronomers were entering the job market. Those who chose radio astronomy looked for new fields in which to establish themselves. They were, in many ways, the academic crop sown by the race to space. These young adventurers were joined by a few others more experienced in astronomy but looking for a new venue. What could be more exciting than the new field of molecular astronomy? Consequently, these detections whetted their appetites, and new proposals to look for molecules poured into the NRAO headquarters in Charlottesville. The bulk of these requested observing time on the 36-ft telescope.

3.2

Lingering Technical Problems

Despite the renewed interest in millimeter-wave astronomy, the 36-ft telescope still had functional problems. In addition to the low efficiency and poor pointing of the telescope itself, the computer program to reduce spectroscopy data was inadequate. The initial system was modeled after the one in use in Green Bank. The data recorder dumped the filter bank output onto punched paper tape. Every day, the paper (later, magnetic) tapes were taken to KPNO in Tucson for processing on their Control Data Corporation (CDC) mainframe computer. The computer printouts were then returned to the telescope, affording the astronomers their first look at observations taken the previous day. While perhaps adequate for a rigidly preplanned observing program, it was too restrictive for astronomers needing to know instantly whether to continue observations on one astronomical object or frequency, or to change to another object or frequency. During this period, the Tucson crew expanded. The NRAO hired more telescope operators to facilitate round-the-clock observing as was done at the 140-ft and 300-ft telescopes in Green Bank. Ned Conklin arrived as manager in September 1969, after completing a Ph.D. from the electrical engineering (EE) department at Stanford, where he had worked in the field of cosmic background emission. Elizabeth (Bess) Rather and astronomer might have stemmed from the 1969 detection at Green Bank of formaldehyde, of which CO is a component. This is incorrect, because CO was on Townes’ 1957 list [30] that most radio astronomers knew of.

32

CHAPTER 3. RADIO LINES FROM MOLECULES

husband John Rather, a new Ph.D. from Berkeley, also arrived. Bess was a wizard with programming even though her degrees were in Medieval History from UC Berkeley—with a Phi Beta Kappa key. Bobby Ulich arrived with a Ph.D. in EE from the University of Texas—along with a tendency to work 20 hours of every day and a visceral understanding of parabolic antennas. Meanwhile, George Grove had returned to Green Bank, and the NRAO had sent another engineer, Dewey Ross, to assist. In the summer of 1973, Mike Hollis arrived to join the programming group. After spending seven years as a nuclear submarine officer, Mike had left the Navy and was finishing a Ph.D. dissertation in astronomy at the University of Virginia. It was a competent, dedicated, hard-working group. In spite of this talent and devotion to long days and hard work, the 36-ft telescope was a tough beast to tame. The most productive use of the telescope was radio spectroscopy, which was technically difficult. Compared to spectroscopy, continuum observing is easy—in my opinion. Seldom did continuum observers have problems other than calibrating the telescope sensitivity in one data channel and correcting the telescope pointing, which usually was a challenging task with the 36-ft telescope. In spectroscopy, many channels are involved, and the observer must be concerned about the intensity accuracy of data arriving in each of them, not just one. Initially, the most successful spectroscopic observers were those with a flare for hardware. The Bell Laboratories group usually brought their own receivers and, occasionally, computer but used the NRAO filter banks [17]. Bob Wilson worked with the software, Keith Jefferts built and maintained their receiver, and Arno Penzias focused upon observing strategy. In the team of Lew Snyder and Dave Buhl, Dave had an undergraduate EE degree and was skilled with electronics. With Phil Schwartz and Bill Wilson, Bill had an undergraduate degree in electrical engineering. Other less hardwareoriented astronomers often teamed up with those who were technically able. In hindsight, such observing teams were characteristic of those observing with university telescopes then and now, where the students and staff were required to solve their own technical problems, if not actually to construct the hardware. In contrast, the NRAO—being a visitor-oriented observatory—generally could not allow guest astronomers to tinker with hardware. Its obligation was (and still is) to maintain its telescopes in “perfect” working order for each new observing team, a state which could easily be spoiled by a technically ambitious but ham-fisted group preceding them. The 36-ft telescope was an anomaly. It was the only NRAO telescope that allowed astronomers to modify the electronics. The NRAO management considered it to be an experimental instrument compared to, say, the 140-ft

3.2. LINGERING TECHNICAL PROBLEMS

33

telescope because that was the only way astronomers could make the telescope productive.

Chapter 4

Dispatched to Tucson My personal route from the Charlottesville headquarters of the NRAO to Tucson involved a number of events, unintentionally preparing me for a critical meeting with Dave Heeschen. Here are some of them.

4.1

Learning How the NRAO Functioned

Unlike many staff astronomers, I had not used NRAO telescopes for graduate work and was unfamiliar with the NRAO “system.” I had joined the NRAO in September, 1969, to work with Peter Mezger and Bob Hjellming in the research field of RRLs. At that time RRLs were a new tool for radio astronomy, discovered in 1964 and offering insights into the nature of H II interstellar gas1 [29]. I knew how to make spectroscopic observations but did not have a clue how the NRAO itself operated. In addition to my research, I found myself assigned as “friend of the 140-ft telescope,” presumably because of my earlier technical experience with the 120-ft Haystack antenna of MIT’s Lincoln Laboratory and because I was a frequent user of the 140-ft. The principal duties of a friend were to assist visiting astronomers. One needed to be familiar with operation of the electronics as well as with the observing techniques appropriate to the telescope to which you were assigned. In short, the telescope friend was an in-house technical advisor. My predecessor was Ivan Pauliny-Toth, who had made substantive contributions to correcting the errant pointing of the 140-ft telescope. By then, Ivan wanted to return full-time to research. Also, 1 The symbol H II is spectroscopic shorthand for ionized hydrogen, that is, for neutral hydrogen—symbolized by H I—after it has been broken into electrons and protons.

35

36

CHAPTER 4. DISPATCHED TO TUCSON

he was more expert with continuum rather than spectroscopic observing, which was taking an increasing share of time of the 140-ft telescope. Since arriving at the NRAO, I had made two contributions to the 140-ft telescope system. First, I installed the “total power” observing technique that I had invented at Haystack in 1967 with Joe Carter [36, 37]. It involves taking a spectrum of the total power on the radio source and nearby the radio source, and combining the two as (On-Off)/Off for each frequency point in the spectrum. In that sense, the spectrum is “position-switched” on the time scale required (typically several tens of seconds) to move the telescope beam between the on and off positions. This observing technique eliminated the troublesome ripples in the spectroscopic baselines produced by the older “load-switching” or “frequencyswitching” observing procedures. These ripples resulted from standing waves arising from impedance mismatches between the telescope feed and the telescope’s metal structure. They rather than the duration of the observation limited the sensitivity of spectral observations. Eliminating these ripples made the baselines flat, allowing much longer integrations of spectroscopic data and greatly increasing the sensitivity of the observations. Soon after I arrived at the NRAO, Joe Greenhalgh coded the on-line acquisition software. Allen Farris coded the off-line reduction software. It is now the standard method of spectroscopy used by radio astronomers. Installing this system on the 140-ft telescope taught me about the technical interfaces in Green Bank. The total-power observing technique required an extremely stable electrical connection between the receiver at the antenna apex and the signal processor in the control room. This connection used RG8 coaxial cables at the joints of rigid co-axial cable mounted on the telescope structure. These 6 to 10-ft flexible cable sections are known as jumpers and had decayed with use such that the RF impedance changed as the telescope slewed. The jumpers had to be replaced to facilitate total-power observing. Replacing these was an adventure. While the receiver itself was the acknowledged responsibility of the Electronics Division, the 140-ft telescope was the responsibility of its chief telescope operator, Howard Brown. Howard took his job very seriously. He barred astronomers and electronics personnel from fiddling with the (“his”) telescope without his permission. New to the NRAO and coming from a do-it-yourself institution, I did not know about these management domains and, accordingly, could not understand why Mike Balister, the Green Bank Electronics Division head at that time, refused to change the jumper cables. Accordingly, I made new ones and, on a maintenance day, installed them myself, using a time-domain reflectometer to ensure that the cable impedance no longer changed with telescope motion.

4.1. LEARNING HOW THE NRAO FUNCTIONED

37

Howard was furious. Not only had I fiddled in his domain without his permission, but I had made the measurements in metric units, meters. “If you want to use foreign units, go to Europe,” was his comment. I never made that mistake again! But I got the job done, and Howard and I were able to work well together now that I clearly recognized his purview. By this time, the NRAO at Green Bank had evolved to an observatory where “professionals” designed, built, and operated the telescopes and electronics. It did not operate like the do-it-yourself university observatories or like the Arecibo and the Haystack radio telescopes, with which I had a great deal of hands-on experience. With few exceptions, astronomers were not to interfere. Hein Hvatum, an engineer and the NRAO associate director for technical services for many years, often told me, “The NRAO would be a much better place to work if it weren’t for the astronomers.” There may have been some justification for his remark. Few radio astronomers were as technically capable as the members of the NRAO electronics staff. Well-meaning astronomers tinkering with the hardware could easily cause problems not only for their own observing runs but also for the astronomers scheduled to follow them. Second, I became involved with replacing the 140-ft’s old analogue pilot drive by a digital computer, principally by designing a new operators’ control panel. The new panel used visible, mechanical digit switches that were much easier to read and change than paging through a touch-sensitive computer screen that had been suggested. This new layout would allow the telescope operator to see all the settings at a single glance and, when necessary, to change settings quickly. During the installation of the computer, the Green Bank staff planned to measure and reset the surface of the 140-ft telescope. Coordination of both tasks in the shortest amount of time required us to use a “critical-path planning program.” Hein Hvatum used PERT (Program Evaluation and Review Technique) critical-path planning software to minimize the telescope down-time. After the work has been broken into discrete tasks with a logical relationship between them, this kind of software organizes the project tasks into the most efficient sequence and identifies those “critical” to completing the project on time. It was the first time the NRAO had used “WBS” (Work Breakdown Structure, as it’s now known) software. Decades later the NRAO senior staff rediscovered this technique in the late 1990s for planning the Atacama Large Millimeter-wave Array (ALMA) telescope2 . 2 Actually, ALMA is the acronym for Atacama Large Millimeter Array, which makes no sense.

CHAPTER 4. DISPATCHED TO TUCSON

38

The critical-path plan worked exceedingly well. Unfortunately, the surface resetting fell short of expectations. Later, we learned that the surface flexed so much as a function of alt-azimuth position such that a simple reset would never have worked3 . But the digital computer and the digit-switch control panel were huge successes and continued to be used for many years afterward.

4.2

My Research Interests Moving Toward Tucson

During this period, my research interests evolved from investigating the characteristics of specific H II regions to exploring the interstellar gas in the Milky Way in a general way. In a long-shot experiment, Steve Gottesman and I had detected RRLs from ionized gas between discrete H II regions in our own galaxy, the Milky Way, [38]. These were broad, extremely weak lines made detectable only with integration times of days, made possible by the “total power” observing technique I had previously installed. Not only were we able to detect α-type RRLs, but we were later able to detect the much weaker β-type lines [39]. The observationally derived quantity  2 T 1.5 d, where N is the electron density; T , the absolute was p e e e path Ne /T temperature of the gas; and d, the path length element through the gas. Interpreting these observations was difficult. Each measurement was a ratio of two unknown quantities over a finite path length. We could de termine the path length d from the velocity width of each spectral line and the known rotation characteristics of the Galaxy4 , however. But was the characteristic temperature of this diffuse ionized gas 10,000K or 100K or 10K—or some other value? Choosing a temperature would then give a corresponding density and vice versa. George Field and colleagues [40] had just published a theoretical model of a two-phase interstellar medium but, without actual temperatures, we could not be certain how to interpret our observations to be useful to this model. Using their adopted temperatures produced discordantly large electron densities for both phases. There were other problems. The weakness of these lines made them impractical as general-purpose probes of the interstellar gases. Was the emitting gas in thermodynamic equilibrium (LTE)? And how much of the detected line emission came from emission from discrete H II regions entering 3

After the 140-ft telescope was changed from a prime focus to a Cassegrain telescope, Sebastian von Hoerner and John Payne calculated and installed a deformable subreflector to compensate for the surface flexing. This solution worked very well. 4 To astronomers, the noun “Galaxy”, with the initial consonant capitalized, refers to our own galaxy, commonly called The Milky Way.

4.2. MY RESEARCH INTERESTS MOVING TOWARD TUCSON

39

the sidelobes of the antenna beam? I remember giving an NRAO colloquium about these observations. In the question period, Barry Clark wisely asked me if I could be sure that the telescope sidelobes weren’t picking up RRL emission from nearby H II regions. At that time, I could not, although we had tried hard to avoid this problem. A couple of years later, Jay Lockman answered all of these questions when he selected this phenomenon for his Ph.D. dissertation. Formally, I was Jay’s thesis advisor but, because I had moved to Tucson, Bob Brown worked most closely with him. He found the diffuse RRL emission to be quite real and, later, to come from hugely extended H II regions ionized by UV emission escaping from the discrete, high-density H II regions of our galaxy, like the Orion Nebula. There had to be a better way to explore the characteristics of the interstellar gas further, to supplement the extensive observations of H I, which were available. “Plan B” was to explore the cold Galactic gas by observing the absorption of the relic background emission by Galactic H2 CO at 5 GHz. This made use of a curious phenomenon in which this molecule absorbed the 3K Big Bang radiation, giving an absorption spectral line with the radial velocity of the molecule. By comparing this velocity with the overall rotation pattern of the Galaxy, an astronomer could locate the molecule and thereby trace the location of the cold interstellar gas in which it was embedded. Subsequently, Mort Roberts and I tried to use the 140-ft telescope for this observation but without success [41]. Good as they were, the NRAO receivers were not sufficiently sensitive for these observations5 . To improve the experiment, Hein Hvatum arranged for Olof Rydbeck to invite me to use the much more sensitive maser receiver at the Onsala Space Observatory in Sweden. Mort elected to remain home in Charlottesville. Dave Heeschen kindly had the NRAO underwrite this extended trip. This trip was an especially interesting one for me. Despite the earlier arrangements, Rydbeck would not grant me observing time. One day, visiting students from Lund University appeared at the Onsala observatory. Rydbeck asked me to give them a talk about the physics and observations of RRLs. Immediately after the talk, the observatory Docent, Bertil H¨ o¨glund, came to our cottage, told me that Professor Rydbeck was impressed with my talk, and asked if my wife and I could join Professor Rydbeck for dinner that evening. Evidently, I had passed the test! From then on, I got all the 5 The diameter of a parabolic radio telescope does not increase sensitivity for observations of spatially extended sources. It is the “system temperature” that counts.

40

CHAPTER 4. DISPATCHED TO TUCSON

telescope time I wanted. Even so, we could not detect extended H2 CO absorption there either, although we had used the world’s best receivers. Plan B was a failure. However, we were able to detect additional RRLs from the diffuse interstellar gas [42]. It was now time for “Plan C.” Returning to Charlottesville from Sweden, I considered that extensive observations of newly detected interstellar CO emission [35] might work. These could be made with the NRAO 36-ft telescope near Tucson. Best of all, Butler Burton had recently joined the scientific staff of the NRAO in Charlottesville from Leiden University, where he had completed a dissertation using λ21 cm hydrogen emission (H I) to probe the large-scale structure of the Galaxy. While visiting, Marc Kutner advised us what was technically possible with the 36-ft telescope and what was not. Consequently, Butler and I prepared a proposal, along with graduate students Tom Bania and Jay Lockman. The referees approved, and Bill Howard scheduled the first phase of the project on the 36-ft telescope in the spring of 1973. The first observing session proved to be a disaster. With respect to projects done routinely with other NRAO telescopes like the 140-ft and 300-ft telescopes in Green Bank, it was exceedingly difficult to use the 36-ft telescope for a believable survey. I remember two problems in particular. There were many others. First, the system for marking Signal (on source) and Reference (off source) spectra was unreliable. A rotating absorber pad mounted in front of the (prime focus) feed generated identification pulses by cutting across a photocell. Blocking the photocell identified the incoming RF signal as Reference; unblocking, as Signal. If the digital system skipped a pulse, then the subsequent Signal became identified as Reference, and vice versa. The result was that a CO spectrum formed in the usual way by the standard (SignalReference)/Reference instead became (Reference-Signal)/Signal. The resulting spectrum would grow when the count was correct but then diminish after the computer lost the identification of the Signal and Reference spectra. Unfortunately, the system lost count often. Second, something was generating a pseudo-random series of digital pulses that moved through the system. Mike Balister referred to this phenomenon as the “marching Chinese,” after the pattern of blips in the baseband signal that could be seen in an oscilloscope. The result was that the spectral noise would not integrate down, thereby limiting the peak-to-peak background noise and the sensitivity of the observations. The four of us realized that our straightforward program of mapping CO in the Milky Way was going to be much more difficult than we had thought.

4.3. THE MEETING

41

Technically, the 36-ft telescope was just not up to the job, at least in the state we found it. Ultimately, this observing program proved to be a great success. These CO observations revolutionized our understanding of large-scale star formation in spiral galaxies by discovering the star-forming annulus in the Milky Way galaxy [43, 44], revealing the location of cold molecular gas, identifying the places where stars form in the galactic plane, and ruling out the then accepted theory in which galactic atomic hydrogen converted into stars. Three of the resulting papers that dealt with the quantitative aspects of these observations [45, 46, 47], have been cited hundreds of times in the astronomical literature. And Scientific American invited us to write a popular article on this topic, which we did [48].

4.3

The Meeting

The morning after our return to Charlottesville from our first attempt to use the 36-ft telescope, the NRAO director, Dave Heeschen, stopped at my office to ask how things went in Tucson. I told him as briefly as I could. He asked me to attend a meeting in his office after lunch. The meeting consisted of Sandy Weinreb (Assistant Director for Electronics), Hein Hvatum (Associate Director for Technical Services), Dave Heeschen, and me. Mike Balister may also have been there but I don’t recall his presence with confidence. Dave asked me to repeat what I had told him earlier. There followed a brief discussion of what should or could be done, and the meeting ended. What I later heard was that our disappointing experience had not been unique. Several prominent astronomers such as Alan Barrett (MIT) had complained about the performance of the 36-ft telescope. While technically skilled astronomers were able to make the telescope and electronics perform, others often ran into insurmountable problems. In the years preceding the detection of molecular lines, the demand for the telescope was not overwhelming and, consequently, unreliable performance was not such a problem. To illustrate the low demand for telescope time, I quote from a 1969 letter from Dave Heeschen to Arno Penzias regarding their proposal to search for CO emission: “I think that at this early date I can’t guarantee the full eight weeks that you have requested. However, I can schedule you now for four weeks next spring, with the understanding that another four weeks will probably be forthcoming [49].” In contrast, by the mid-70s, an astronomer

42

CHAPTER 4. DISPATCHED TO TUCSON

would be lucky to get four days of observing. Later, he would be lucky to get two days. Now, the telescope was in extremely high demand, and the NRAO had the responsibility to make it work well. To deal with the unreliability, Heeschen decided to change the administrative environment of the 36-ft telescope. Previously, the responsibility for the telescope had lain with the NRAO Technical Services Division in Charlottesville because the telescope had been considered an experimental instrument. The site manager in Tucson—Ned Conklin—reported to Hein Hvatum rather than to Dave Heeschen. Being an engineer, Hein had difficulty relating to the “astronomer personalities” that were complaining about the poor performance of that telescope. Hein was very angry about the meeting. That afternoon, he accused me of making a mountain out of a molehill. He told me that he had arbitrarily assigned “credibility factors” to the astronomers’ complaints and had reacted accordingly. Specifically, he said that while he assigned a 1.0 to anything that Bob Wilson told him, he assigned factors as low as 0.01 to complaints made by some others. I don’t recall the index Hein had given me but it was surely not as high as Bob’s. He also told me he was angry that I had not come directly to him immediately after returning from Tucson. But this was unreasonable. At that time I was not involved in the NRAO management and had known nothing about a history of complaints about the 36-ft telescope. Furthermore, Dave Heeschen had come to my office door first. I had not approached him. It was time for drastic changes. The ultimate management of the 36-ft telescope had to become empathetic to its users. Therefore, based upon the operating problems and the greatly increased demand for telescope time, Heeschen created “Tucson Operations” to be a major operations center for the NRAO, parallel to Green Bank Operations, and headed by an assistant director who would report directly to him. Technical Services would no longer be responsible for the telescope—except for its electronics. Possibly because of my long-term CO observing project and my experience with the 140-ft telescope in Green Bank (I’ll never know), Dave Heeschen asked me to move to Tucson to become its first assistant director. He also told me that he would supply whatever extra money and staff might be required, but my job would be to make the telescope reliable. Being ambitious (and inexperienced) and confident that I could turn things around, I accepted and agreed to move to Tucson by September 1, 1973. My wife was much less enthusiastic about a move away from Charlottesville. Nevertheless, she agreed to give it a try.

4.3. THE MEETING

43

My predecessor, Ned Conklin, was offered an Assistant Scientist position at the NRAO’s offices in Charlottesville, which he declined. Instead, he accepted an appointment with the Arecibo radio telescope in Puerto Rico. Nonetheless, Ned arranged to stay long enough in Tucson to ensure a smooth transition. Frankly, I felt badly about the situation. A very bright guy, Ned had worked extremely hard to make the 36-ft a success and, in fact, had solved a lot of its problems. Even more ironic was that, in a sense, we should have become buddies. Not only had the two of us been Yale undergraduates a year apart but our fathers had been Yale classmates. After my appointment, his father wrote my father a note wishing me well in Tucson.

Chapter 5

Expanding the Tucson Facilities Because it was an experimental facility, the size of the support facilities in Tucson were modest. One of the first items on my list was to expand them.

5.1

Physical Environment

In September, 1973, the NRAO offices occupied about 1,300 ft2 in the basement of the KPNO building on Cherry Avenue in Tucson. KPNO had provided an extra 300 ft2 to the 1,000 ft2 that the NRAO had earlier purchased. Nonetheless, this space was already inadequate. The complexity of receivers being shipped to Tucson was increasing, requiring even more space for maintaining them. The administrative staff had increased to three people, including Bill Gust for shipping and receiving, Maxine Thomas for secretarial support, and Ned who doubled as an engineer/scientist. There was a full complement of telescope operators (Don Cardarella, Dave Myers, Paul Rhodes, Werner Scharlach, and Cal Sparks) working on a Flexible Work Week schedule of 7 days on and 7 days off and a telescope mechanic (Marty Tester). Elizabeth Rather and Mike Hollis were supporting the computer. And there were two engineers (Bobby Ulich and Dewey Ross) and several electronics technicians (Neil Albaugh, Chuck Lipscomb, John McBrian, and Armand Sperduti). John Rather was building a continuum bolometer as well as pursuing his own astronomical research. From time to time, technical people from Charlottesville visited the Tucson group, further increasing the demands on the Tucson space. Although much of the work took place 45

46

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

at the telescope, a great deal still occurred in the Tucson offices. Working with Ned revealed information about his working environment generally unknown in Charlottesville. Ned had indeed requested more financial assistance for the telescope from time to time but Hein had not always fully agreed to his requests. For example, a huge, new cryogenically cooled mixer receiver was under construction in Charlottesville and was scheduled to arrive in Tucson in early 1974. See Figure 5.1. This would require a robust “cherry picker” to lift the heavy receiver and place it in the vertex of the reflector. Previous receivers were uncooled and much lighter, installed at the apex (prime focus) of the telescope. When asked how much one would cost, Ned told Hein that it might be as much as $30k but he would have to investigate. When he called Hein to tell him that the lowest price was $40k, Hein said that he would authorize only the original estimate—$30k. Consequently, Ned had to shorten the cherry picker to meet the budget—which meant that it could not reach the top of the astrodome and would be much less useful over its lifetime. Learning the facts regarding work in Tucson, I became convinced that Charlottesville had not appreciated the enormity of problems in Tucson as well as how hard the Tucson staff had been working just to cope. Finally, Ned himself suggested that he was unfamiliar with the NRAO culture of supporting visiting astronomers at any costs [17]. This story illustrates the conflicting priorities experienced by the NRAO employees associated with the 36-ft telescope. Located in Tucson, Ned had responsibility for accommodating visiting astronomers and installing new equipment. Located in Charlottesville, Hein had responsibility to see that observatory money was being spent effectively. Even though he had been intimately involved with the construction of the telescope, he had become unaware of the detailed requirements of the Tucson environment. The Tucson staff did not have adequate tools on site to maintain the telescope. When something more than, say, a simple hand tool was needed, the NRAO employee often drove from the 36-ft telescope to the Kitt Peak summit facilities to borrow appropriate tools from the KPNO shops. This practice was consistent with the way the telescope had originally been set up, when the NSF increased the KPNO allocation by $25k per year to pay for these services. However, the situation had changed a great deal since 1967. The telescope was now deluged with observing proposals, astronomer expectations had risen a great deal, and the priority for reliable, continuous operation had become high. Adequate on-site tools were now essential. We began spending lots of money equipping Tucson Operations with adequate

5.1. PHYSICAL ENVIRONMENT

47

Figure 5.1: The 36-ft telescope in 1967 with the first cryogenically cooled receiver, which operated at λ3 mm. The cryogenically cooled receiver is now mounted at the antenna vertex. The hole through the reflector on the left is for an optical pointing telescope. Plastic sheeting protected the subreflector mount from rain.

48

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

Figure 5.2: The 36-ft telescope in 1981 with Cassegrain optics shortly before being replaced by a paneled, 12-m surface. Highly evolved from the Cassegrain receiver shown in Figure 5.1, this receiver uses a complicated quasi-optics system to inject the local oscillator, contained in the heavy metal box on top on the receiver. A quadrupod supports the prime focus equipment. Insulation covers each leg for thermal stability. The buildings within the dome include an expansion to the original control building, an astronomer’s office, and a small welding shed. At left is the scaffold for servicing the prime focus.

5.1. PHYSICAL ENVIRONMENT

49

Figure 5.3: The underside of the nutating subflector associated with the Cassegrain optics of the 36-ft telescope. In its center is a noise generating tube for continuum calibration. The “squawk box” at right allows technicians to communcate with the control room while servicing the unit. NRAO photo, GB81-19394-TUC tools, a pattern that continued through the closing of the telescope. Over the years we bought so many handtools that the dome floor could have been covered with them to a depth of a foot! During this period, I was learning about administration by trial and error. During the few years I had worked for the NRAO, I had recognized that the observatory style was to be “lean and mean.” I tried to put this into practice in Tucson. To reduce copying, KPNO had installed card-controlled controllers on their copying machines. This proved awkward for our group. Although assigned one of these access cards, we did not always know who had it and had no effective system to track this, considering the hectic nature of our operations. So, I leased an inexpensive “wet-process” copier and had it installed in our office area. After only a day, Armand Sperduti, one of our new technicians, marched into my office to tell me that this parsimony was a hardship on the already over-worked staff, and insisted that I get a dry-

50

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

Figure 5.4: The control room of 36-ft telescope in 1981. The racks behind the chief operator Paul Rhodes contain modules controlling the telescope focus and pointing offsets, a weather station, and a continuum receiver. By this time, the telescope, electronics, and computers had become sophisticated compared to 1967. NRAO photo GB81-19419-TUC process copier even if it cost a lot more. He was absolutely correct, and we returned the cheap copier and replaced it with an adequately sized Xerox copier at three times the lease rate of the other one. The lesson: one can be penny-wise and pound-foolish. In December, the observatory allotted me money to buy Christmas presents for children of employees. Children’s Christmas parties were a tradition at the NRAO sites. However, for the Tucson group, December 1973 was overwhelming. The observing load was high, the receivers were still unreliable, and our group was small. We seemed to be working all the time. There was no way we could find time to buy gifts, wrap them, and hold a Christmas party. Thinking about this, I decided to divide the money by the number of children, get some new dollar bills and gift envelopes from a nearby bank, and give the filled envelopes to their parents who, in turn, would give them to their children. The more I thought of the idea, I liked it more and more and proceeded to do it. Shortly afterward, at one of the

5.2. RELATIONSHIP TO KPNO

51

quarterly meetings of the NRAO management, I described what I had done. Both Ted Riffe (Associate Director for Administration) and Dave Heeschen (NRAO Director) exclaimed together, “You did what? You never, never give cash to employees. It’s a taxable wage.” I saw the point immediately and felt badly about a convenient but stupid move. But Heeschen then grinned and said, “Under those conditions, I might have done the same thing. ” Typical. He was a wonderful guy to work for.

5.2

Relationship to KPNO

In the summer of 1972, KPNO had acquired new senior management. Leo Goldberg had left a professorship at Harvard College Observatory in Cambridge, Massachusetts, to become KPNO director. He brought along a business official, Harry Albers, from the adjoining Smithsonian Center for Astrophysics (CFA) to become the new KPNO business manager. Neither was aware of the detailed arrangements made earlier between the NRAO, KPNO, and the NSF for support of the 36-ft telescope. This was a consequence of their being new to KPNO and to the often oral nature of the agreements. Nonetheless, Leo Goldberg was enthusiastic about the astronomy associated with the 36-ft telescope. This was his nature. Even while department chairman at Harvard, Leo had not only actively participated in his own research but always had one or two Ph.D. students he worked with. Furthermore, he knew a great deal about the NRAO, having been involved with its formation and having served on the AUI board. He was a scientist first, an administrator second. Furthermore, he and I had worked together on RRLs before I came to the NRAO; we had a personal relationship. But there were limits to what he could do for us. Although sympathetic to our need for additional space, there was none to be had in the KPNO building. The lack of space in the building and the different styles of the NRAO and KPNO led to some conflict. Unlike KPNO, the NRAO had geographically separated sites, often requiring long-distance telephone calls. All of our long-distance calls had to go through the KPNO switchboard, which had established procedures for controlling similar calls originated by the KPNO staff. The Tucson NRAO group used KPNO purchasing on occasion, but there were forms to be filled out for purchase requests. The use of KPNO cars required forms. One reason for this formalism might have been that, unlike the NRAO, KPNO received money from more than one federal agency and was required to account for spending under each source. Despite these regulations, most KPNO officials bent over backwards to assist the NRAO

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

52

group, especially their assistant director Beverly Lynds, their Purchasing Manager, Jerry Rabb, their Mountain Superintendent, Dick Doane, and the mountain Mechanical Supervisor, Max Galey. Still, KPNO employees raised their eyebrows from time to time about our perceived free-wheeling nature. In contrast, the NRAO operated much less formally, perhaps because it received all its money from one agency—the NSF. Both Heeschen and his associate director for administration, Ted Riffe, expected their managers to perform their jobs with minimum paper work—as long as they did not violate the Federal Procurement Regulations—and with initiative and common sense. To facilitate this fluidity, Heeschen delegated a great deal of authority to his managers, discouraged superfluous paperwork or unnecessarily complex procedures, and made sure your leash was a long one. For example, each assistant director had spending authority up to $25k1 on all observatory accounts, even those assigned to other assistant directors. There were no controls on long distance telephone calls or copy machines as in KPNO. The NRAO telephone book never used “Dr.” in their listings whereas the KPNO one always did (and still does). The NRAO at that time generally eschewed organization charts—although one did exist for required presentations to the NSF. Accountability came from the close relationships between supervisors and supervisees; it was a climate of trust. To make certain you knew the rules, though, Ted Riffe urged each of the assistant and associate directors to read through the NSF-AUI contract annually. I actually did that. Furthermore, Jay Marymor, an attorney and the NRAO purchasing manager, did not hesitate to ask penetrating questions when he saw something he did not like. An example of this simplicity is a discussion I stimulated very early after my appointment as an assistant director. In an NRAO director’s meeting in Charlottesville, I asked Ted Riffe whether the observatory should institute a dental plan for employees. “We already have the best possible plan,” replied Ted. “We do?” I said. “Yes, it’s called a salary,” said Ted. “Every family has the opportunity to put a portion of their salary into a bank account earmarked for dental work. In fact, they can put more or less away, according to how much protection they want. Best of all, this plan has no administrative cost to the observatory.” At the time, I thought Ted’s answer was hopelessly conservative. But looking back, I think he was absolutely correct. Unlike medical crises, dentistry rarely involved bankruptcy risks for employees. 1

A good annual salary at that time.

5.2. RELATIONSHIP TO KPNO

53

Heeschen also encouraged management to be as local as possible. I remember when he decided to create a new employee classification of “System Scientist.” He telephoned to advise me about the forthcoming announcement, noted that Tucson Operations had one Ph.D, Bobby Ulich, currently classified an “Electronic Engineer I” who might prefer to be classified as a “System Scientist.” II, not Heeschen, was to ask Bobby’s preference and report back. At that time, Bobby preferred to remain classified as an EE I. Ironically, several years later, Mort Robert’s refusal to give him the previously offered System Scientist title was a major factor in Bobby’s decision to leave the NRAO to join the Multiple Mirror Telescope group (MMT) of the University of Arizona. There were other substantive differences as well between the NRAO and KPNO. The NRAO staff astronomers had to compete directly with guest astronomers for access to the NRAO telescopes. In this respect, there was no advantage to being a member of the observatory staff. In contrast, KNPO astronomers had access to a separate block of telescope time set aside for them, in recognition of their service to their observatory. They did not always compete with guest astronomers for time on the optical telescopes. With these differences in style, it was inevitable that each organization chafed a little at the other’s style in spite of good personal relationships and in spite of their similar objectives as national observatories. Many of the KPNO staff probably saw the NRAO Tucson group as an organization of “cowboys” compared to their more ordered observatory. Certainly, the NRAO group viewed the KPNO group as unnecessarily formal. Over the last decade another factor had appeared. The funding climate had changed. Operating money had grown scarcer for both observatories, which impacted the hitherto casual financial relationship between them. One day, the new KPNO business manager, Harry Albers, summoned me to his office to present me with a list of “free” services that the KPNO was providing to the NRAO group in Tucson. The incredibly detailed list included grass-cutting, charges for long distance phone calls, access to the library, parking space, security, copiers, postage, key duplications, janitorial support, consultations with KPNO personnel, etc. He asked that the NRAO pay $50k annually for these services. He also suggested we might be happier if we found quarters outside of KPNO. This was a shock to me. At the time, I knew nothing of the arrangements made during previous years between the two observatories. I quickly telephoned Heeschen, who soon came to Tucson. He described the earlier financial arrangements to me. Rather than get involved himself and risk undercutting my local authority, he advised me to tell Harry that, if

54

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

we moved from the building, the NRAO would have less money and KPNO would have more money—a situation surely unintended by the NSF in earlier arrangements. Consequently, we would ask for compensation for the loss of office space and for part of the annual allocation that KPNO had been receiving from the NSF to support us. Later, I duly passed along this recommendation informally to Harry (in the men’s room, actually). Dave Heeschen had assessed the situation perfectly. I heard nothing more from KPNO regarding fees. In any event, the NRAO staff had indeed grown too large for its space. Equally pressed for space, KPNO could not allocate additional space in their building. Even if they miraculously found extra space, continued coexistence would not have been in the best interests of either observatory. Something had to be done.

5.3

Finding New Space

There were two possible solutions to the space problem. My first idea was to construct office and laboratory facilities on Kitt Peak and bus everyone to and from the mountain on a daily basis. The second was to lease or build new facilities in Tucson and continue to send only essential people to and from the 36-ft telescope. Dave Heeschen told me he was willing to fund either choice. Since late 1972, Ned Conklin had been looking for a Tucson building to house the Tucson offices. He had found a small building with an inner courtyard but it required extensive and expensive renovation. By the time I arrived in September 1973, this building was no longer available. However, it might not have been large enough to house the laboratory and workshop space we now required. It was now the spring of 1974, and we needed to find something quickly. As a temporary measure, we had rented an office trailer and moved it adjacent to the KPNO building on Cherry Avenue. While it indeed alleviated the space problem, the NRAO was still located on KPNO turf and, in part, operating through its aegis. Dale Webb had arrived as Tucson business manager in May 1974 and joined me in looking for appropriate space. By this time, I had decided that it would be crazy to transport everyone to and from Kitt Peak. Many employees such as our secretary, Maxine Thomas, and our administrative assistant, Bill Gust, needed to be in Tucson to interact with visiting astronomers and to procure supplies for the

5.3. FINDING NEW SPACE

55

telescope. Furthermore, it was important for some of our engineers and technicians to remain separated from the often frenetic environment at the telescope. They needed a calmer environment to design and build long-term solutions for the technical problems. For them, it would only be necessary to go to Kitt Peak on maintenance day (every Wednesday) or when observing equipment failed. Daily commutes to the telescope would waste a lot of time for which the NRAO would have still to pay salaries. Finally, my plan was to improve the telescope reliability so much that only the operating staff would be needed at the telescope. Why not initiate this operating mode as soon as possible? Dale and I looked at several existing buildings but none seemed to be right for us. We decided to lease about 4,000 ft2 of space in a new tilt-slab building under construction in an industrial park on the west side of Tucson, known as the Forbes Interstate Industrial Park and located on the southwest corner of Grant Road and Interstate 10. This option allowed us to design the interior space to meet our needs. We planned adequate office space, laboratory space, and an area for storage and minor repairs to our vehicle fleet. Being on the west side of Tucson, this location afforded easy access to the Ajo road that led to Kitt Peak, eliminating the necessity of driving through the “stop and go” city traffic and shortening the driving time to the telescope. And it was cheap—about $40k per year or $10/ft2 per year, I recall, with lots of parking spaces. I insisted upon a shower room because several of us had begun bicycling from home to the office. This was a “first” for any NRAO facility and it was well-used. By this time, there were nineteen NRAO employees in Tucson. Sandy Weinreb had transferred John Payne from Charlottesville to Tucson to head the electronics group. John had been reluctant to move but knew he had no choice when he learned that Sandy had reassigned his Charlottesville office and desk to another engineer. Jack Cochran and Jesse Davis, also of the Charlottesville group, came with the new cooled receiver. In August, Bob Freund arrived from graduate school at MIT. Meanwhile, Ned Conklin had gone to the Arecibo observatory, and Elizabeth Rather and Chuck Moore had left the NRAO to form FORTH, Inc. The new quarters proved to be a success—so much so, that we leased an additional 1,600 ft2 of adjoining space when it became available. Figure 5.7 shows the entrance and Figure 5.8, the entire complex. Also of importance was that the new space gave us an identity. The name on the front door was now “National Radio Astronomy Observatory” rather than KPNO, and we had full control over our administrative operations such as long distance telephone calls, postage, key access, leased cars, and parking. The only

56

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

Figure 5.5: Some of the Operations Staff in 1974. Left to right: Dale Webb (business manager), Maxine Thomas (secretary), Paul Rhodes (chief telescope operator), Betty Stobie (programmer), Dan Bass (electronics technician), Jack Cochran (electronics technician), Mark Gordon (site manager), and Stan Sullivan (mechanic). How young we were!

Figure 5.6: Some of the Technical Staff in 1974. Left to right: Ruben “Bud” Hill (electronics technician), Ron Silver (electronics technician), Robert Freund (engineer), John Payne (head, electronics division), Terry White (electronics technician), Mike Routt (electronics technician),Jesse Davis (engineer), and Dewey Ross (engineer). Again, how young we were!

5.3. FINDING NEW SPACE

57

Figure 5.7: The entrance to the new NRAO offices on Forbes Boulevard, in a tilt-slab building leased in the Forbes Interstate Industrial Park in 1974. Standing under the portico are Paul Rhodes (chief telescope operator), left, and Bill Gust (administrative assistant). drawback was a new isolation from the Tucson astronomical community but, at that time, we were completely focused on getting the 36-ft telescope to function reliably. There was no time for any of us to do astronomical research. The expansion of the Forbes facilities unintentionally carried a financial boon. When we attached the adjacent space, we expanded our electrical circuits into the new area. In the early 1980s, the local power company, Tucson Electric Power, discovered that only two of the three phases supplying these offices were being metered. The electrician had made an error when he connected the expansion circuits. Evidentally, the NRAO had been paying for only two-thirds of the power we had used since 1975. At first, TEP wanted us to pay at least $150k for the unmetered electricity we had used. However, it proved difficult to estimate how much this was. TEP finally settled the claim for about $3k. It seems that every two steps forward were accompanied by one step

58

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

Figure 5.8: An aerial photograph showing the Forbes Interstate Industrial Park, looking northeasterly. An arrow identifies the building housing the NRAO offices. The black strip running approximately horizontally across the picture is the Southern Pacific railway, running along the east side of Interstate 10. Below the buildings of Forbes Park is a meander of an ephemeral river, the Santa Cruz. Such meanders work their way downstream (leftward) during large flows, as we discovered in October, 1983, when waters washed away some of the NRAO parking lot. backward. In the early 1970s, the nation was experiencing a gasoline shortage. “Conservation” was the solution recommended by the US government. Accordingly, in late 1974 Aaron Asrael of the NSF sent me a letter requiring us to reduce our energy consumption by reducing the miles driven on our leased cars by a significant percentage. Moreover, we were to generate reports proving our compliance. It was impossible to comply and still maintain full, reliable operation of our telescope. I wrote him a letter asking for an exemption and received no answer. So, we continued to operate as required to support telescope operations. Obviously, this was an executive order passed to all government agencies which, in turn, passed it down. There was no provision for enforcement. Furthermore, the letter set limits on the interior temperatures of gov-

5.4. NEW MOUNTAIN LABORATORY

59

ernment buildings, a category that included our leased space. I think the NSF ordered winter temperatures to be no higher than 68F, and summer temperatures no lower than 78F—but I don’t recall exactly. To comply, we installed locked boxes on all thermostats. It did not take long for our clever employees to find a way around the restrictions. Neil Albaugh routinely sprayed electronics refrigerant through ventilation holes of the locked boxes onto the temperature sensors, which consequently gave falsely low readings and ignited the furnaces. I suspect he used a heat gun in the summer. The NSF never anticipated the level of resistance they were dealing with!

5.4

New Mountain Laboratory

There were still space problems on Kitt Peak, however. The original control room within the astrodome was too small to support normal operations and, in 1971, had been augmented by a second room within the dome known as the Dewey Ross lab. This space fell short of what was really needed to maintain the large, new cryogenic receivers that facilitated the cutting edge of our research. Unlike the older, smaller prime focus receivers, the massive Cassegrain receivers were difficult to haul to and from the mountain. Accordingly, in 1975 we planned a new, 2,500 ft2 laboratory next to the astrodome and Dave Hogg, the NRAO associate director for operations, wrote a letter asking KPNO, our Kitt Peak landlord, for their approval [50]. KPNO was unhappy with our proposed design [51]. The NSF leased the Kitt Peak observatory site from the Papago—now known as Tohono O’odham—tribe. As the primary representative of the lessee, KPNO was obligated to oversee the amount and style of construction within the observatory grounds. First, we had planned a metal, pre-fabricated building to keep costs low. KPNO felt that such a design would be inconsistent with the existing architecture on Kitt Peak. Second, we planned to elevate the building so that its floor level would match that of the dome, facilitating a way to roll the heavy receivers between the lab and the astrodome. It was a conflict in points of view: we were thinking about an effective, cheap solution that could be implemented immediately; KPNO, the longer term visual impact of the observatory. Accordingly, we changed to a masonry construction on a site to the East of the dome, selected a neutral color (“desert beige”) that blended with the environment, and agreed to a lower floor level than planned earlier. Harry Albers approved the plans in December 1975 and the Pace Construction company began work immediately.

60

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

Figure 5.9: An aerial view of the “West Ridge” of Kitt Peak, looking south, as it looked from approximately 1970 to 1972. Both living trailers are in place. The small backup generator is seen on the north side of the dome. Site testing equipment of KPNO is shown near the top of the photograph. KPNO photo 9660F9

5.5

New Operators’ Dormitory

Our telescope operators were having difficulties getting a sound sleep. They had been using the two dormitory rooms on the Kitt Peak summit, which the NRAO had paid for some years earlier. Our visiting astronomers had been using two inexpensive, dilapidated house trailers. One had originally been installed on the summit for visiting NRAO observers and was moved down to the 36-ft site in 1971 [21]. The second came from a German aeronomy research group who were observing ionospheric winds by means of the fluorescence of sodium released from a high-altitude rocket, from a location near the 36-ft site. After this experi-

5.5. NEW OPERATORS’ DORMITORY

61

ment was completed, Ned Conklin acquired their trailer for a song in 1972. For astronomers, the proximity of the trailers to the telescope more than offset their rusticity, especially for the brief two or three days they were using the telescope. As long as the linen was clean, the furnace worked, and hot showers were available, the astronomers were happy. Architectural beauty was not important to them. The situation for our telescope operators was very different. Their working shift was a 13-hour day shift or an 11-hour night shift for a seven-day period of a two-week cycle. It was essential that they be able to withdraw from their work and get a sound sleep during the “off” hours of their shift. This was not easy because the KPNO night assistants with whom they shared a mountain dormitory worked only nights, had working hours that varied with the season (short in summer, long in winter), and often played loud stereos after they awoke [52]. There seemed to be no way to reconcile the different working hours and, consequently, the NRAO telescope operators working nights usually did not get adequate sleep during the day. Being a light sleeper, Dave Myers had the most problems with the ambient noise. There was an additional complication. To ensure that optical astronomers had consistent observing throughout the night, KPNO paid over-time to the night assistants at a rate of 1.5 times their hourly pay calculated on a daily (8 hours) basis—more than required by U. S. Labor Law. The NRAO operators, however, were paid overtime on a “fluctuating work week” basis (80 hours over two weeks), which usually resulted in significantly less pay. Not only did the KPNO night assistants prevent our telescope operators from getting a sound sleep in dormitory rooms that the NRAO had paid for, they also took home more pay for essentially the same work. The combination was terrible for morale. The only solution seemed to be to find new, separate, accommodations for our operators. Accordingly, I asked Leo Goldberg in 1976 to allow the NRAO to erect a high-quality, two-bedroom pre-fabricated cottage near our telescope for the exclusive use of our telescope operators [53]. The deal— worked out with Harry Albers—was that the NRAO would pay $8.5k plus their financial interest in the mountain-top dormitories in return for KPNO purchasing and erecting this mini-dorm. While the building was to be included in the KPNO inventory, the NRAO would have exclusive use of it as long as needed. The NSF approved this arrangement, and the Bullock pre-fab cottage was scheduled for installation in mid-December 1976. Seen in Figure 5.10, it proved to be a big success with the NRAO telescope operators. Associated with the mountain housing is an example of the detail that

62

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

Figure 5.10: The prefabricated operators’ dorm on its site near the 36-ft telescope. It was installed in 1976. consumed our small Tucson group. When we installed 7-ft beds in the trailers to accommodate tall astronomers, we bought appropriately sized sheets, longer than the standard sized sheets used by KPNO. Yet, for economy, both observatories sent their combined laundry to the same commercial laundry. We needed a way to identify our longer sheets. To facilitate separation, Dale Webb and I divided all of the new white linen, each bringing half of it to our homes and personally dying it with blue RIT dye in our home washing machines. It is hard for me to imagine that happening today.

5.6

The Sewage Crisis

While not one of the high-profile items in the history of Tucson Operations, this story illustrates our relationship to KPNO and to the NRAO in Charlottesville. To be responsible members of the Kitt Peak mountain community required ignoring some NRAO internal bureaucracy. In March of 1976, Dick Doane, then the mountain superintendent for KPNO, called to advise me that their maintenance people had learned that raw sewage was leaking from our septic system on the west ridge. A stickler for conforming to health and safety rules, Dick told me that if we did not fix

5.6. THE SEWAGE CRISIS

63

the problem immediately, that the KPNO management would shut us down. He was also concerned that a story about leaking sewage did not come to the attention of the Papago tribe or to Tucson newspapers. As principal lessee of Kitt Peak and accountable to both the NSF and the Papago tribe, there was no question that KPNO had the authority to do this. We had to fix this problem immediately. As I recall, both Dale Webb and I drove to our telescope to see for ourselves. The situation was exactly as Dick had described. Our old sewage system was no longer functioning correctly. I asked Dale to find some company to fix it immediately. Meanwhile, I telephoned Jay Marymor, our contracts manager/attorney in Charlottesville, and asked him to expedite approval from the NSF, which we required. Jay informed me that this was not a crisis, that KPNO did not have authority to shut us down, and that approval from the NSF could take some time. He advised me to be patient and to carry on. Dale had found the Griffin Construction Company in Tucson that would install a new system for $11k—a lot of money for us in 1976. This was a “find” because, at that time, few companies would take their heavy equipment from Tucson to Kitt Peak for a smallish job. This overhead would eat up their profit. Armed with the cost information, I called Jay again and told him that we had been lucky enough to find a willing contractor and needed to move quickly. Again, he said I had to wait. Thinking more about this, I decided that being an assistant director meant I was responsible for solving problems like this one. So, I told Jay that I was proceeding to spend the money and that he should find a way to square my decision with the NSF. From the roar over the telephone, I thought he’d had a stroke. Even today, I think it was the correct decision. Furthermore, it signaled to KPNO that the NRAO wanted to be a responsible member of their mountain community. The work proved to be a disaster for our contractor, however. All the gravel for the percolation medium of the leaching field had to be hauled up the constant-grade Kitt Peak highway to our site. On one of the trips, the dump truck stalled, lost the air in its braking system, and began to roll backwards. The driver jumped out, striking his neck on the rear-view mirror and breaking a couple of vertebrae, and the dump truck rolled off the road and tumbled down the flank of the mountain.

64

5.7

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

The Fate of Our KPNO Office Space

After we had moved to the industrial park, it was essential to keep an office in the KPNO building on Cherry Avenue. Our people frequently used their CDC mainframe computer as well as their excellent library. When the NRAO left, we traded the 1,000 ft2 that the NRAO had purchased for $60k for a small, inner office in the building, a swap endorsed by Leo Goldberg. The office could be located anywhere in the building, as long as we had sole use of it. In the early 1978, Geoffrey Burbidge replaced Leo as director of KPNO. He was a talker rather than a listener, who often rode roughshod over anyone he dealt with. Ignoring an agreement reached earlier with Leo Goldberg, Burbidge told me that the NRAO could continue to have an office in the KPNO building only as long as we used it four or five days a week. This was impossible for us; our principal offices were in an industrial park five miles away. I managed to convince him that we could use this as a shared office, that is, we would be willing to share this office with a visitor to KPNO as long as we could retain our right to use it when needed. (I wrote two research papers using that office.) Six months later, his secretary called to inform me that a moving van was en route to our Forbes offices with the contents of that office. This arbitrary action, breaking an agreement and ignoring the earlier NRAO financial arrangements, broke the last physical link that the NRAO group in Tucson had with the academic astronomy community in Tucson, the University of Arizona, and KPNO.

5.8

Moving to the University of Arizona

I had become increasingly concerned that the scientific connection between Tucson Operations and academia had gone. The KPNO expulsion was the last straw. With few exceptions, fewer and fewer of our senior people were publishing research papers in the standard, refereed journals. Almost all our focus was on maintaining telescope operations: repairing failures whenever they occurred, improving existing systems, maintaining the physical plant, and implementing advances in technology regarding local oscillators, receivers, feeds, and software. We had become more like an “auto repair shop” than an astronomical observatory. In my mind, the most desirable move would be one to the close vicinity of the University of Arizona. I had had several conversations with Peter

5.8. MOVING TO THE UNIVERSITY OF ARIZONA

65

Strittmatter, director of Steward Observatory, to ask if space for us on the campus might become available. Although empathetic to our situation, he had known of none. Late in 1982, Peter called me to tell me that the Regents had approved a new addition to the astronomy building. There would be an opportunity to add an extra floor to house our group. The new addition would be available in mid-1984. If we were interested we needed to act quickly. The next day, one of the architects visited us at the Forbes facility to ascertain our needs. Dale and I told him that we liked our present layout: a ring of small offices surrounding or adjacent to a centralized laboratory area. We also told him that we needed to lock up the entire area to safeguard federal property. The standard university configuration of individual locked offices opening onto open corridors would not work for us. As I recall, he provided a sketch the very next day. Very soon afterward, I met with Rodger Thompson, who was acting director of Steward, to review our decision and to ask about options. Peter Strittmatter was away on sabbatical leave in Bonn. When I told him we needed a shower for our bicycling commuters, Rodger was aghast. He told me that this issue had arisen among the Steward staff, which the management had successfully resisted. To Rodger’s credit, he said we should do what we needed to do but he foresaw troubles from the Steward staff when the word got out that the NRAO would be including a shower facility. We disguised it as an enclosed emergency shower, and no one from Steward ever complained as far as I know. Bicycling seemed to be a passion for many Tucson employees. The shower was heavily used. Easily, the most frequent user was the NRAO astronomer Jeff Mangum, who bicycled the 18-mile round trip from his home every workday, winter and summer, rain and sunshine, for an annual mileage of about 4,500 miles over the five years he lived in the Tucson foothills! Engineer Bob Freund took a close second for annual mileage, riding almost every workday over only an 8-mile round trip but for a 30-year period, thereby winning the prize for the grand total. There were a few “distant thirds.” Technician John Fitzner was undoubtedly the strongest rider with his love of cycling to mountain summits in Arizona and Colorado, including Tucson’s Kitt Peak and Mt. Lemmon. Engineer John Payne commuted by bicycle almost every day he was in Tucson and once rode from Tucson to the VLA in summer—an unbelievable grueling trip because of the 100+F temperatures and unrelenting sunshine. I commuted only two or three days a week over a 20-mile round trip in non-winter months and, hence, was a wimp in comparison with the others. But Hein Hvatum and I cycled from Mis-

66

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

Figure 5.11: Southwestward view of a new addition to Steward Observatory, completed in 1984. The University of Arizona added the top floor specifically to house the NRAO’s Tucson Operations. I took the photograph from the roof of KPNO’s Tucson office on Cherry Avenue.

soula, Montana, to Pueblo, Colorado, in June, 1976, over a 1,300-mile route that connected the high mountain passes along the way. And there were other employees who commuted by bicycle sporadically over much shorter distances, depending upon the weather. I believe that all of this cycling went a long way to maintain our collective sanity. Maintaining the fragile millimeter-wave telescope and electronics for visiting astronomers with high expectations for the system reliability could be emotionally taxing. On the other hand, some might allege that this work ordeal drove many of us sufficiently crazy to engage in all of the cycling. To negotiate the details of the deal, Dale and I met with the Assistant Vice President for University Facilities, George Cunningham, on January 3, 1983. I first proposed that the NRAO pay the construction cost of the entire additional floor, give title to the university, retain the right to occupy it as long as necessary, and pay all maintenance costs. This was the arrangement

5.8. MOVING TO THE UNIVERSITY OF ARIZONA

67

the NRAO had made with with the University of Virginia in Charlottesville with respect to their main office, a building known to the University as “Stone Hall.” This idea did not appeal to him. We agreed that Arizona would build the floor to our specifications and that the NRAO would enter into a renewable, five-year lease for the new quarters. Principally because parking spaces were in short supply at the university, I asked George if our employees could have parking privileges as well as discounts for books, athletic tickets, and low-cost access to university classes—just as provided to university employees. He agreed. For this the NRAO would pay $70k per year, which was what we would have paid at our existing quarters in the industrial park. Dale Webb remembers this $70k figure to have been the approximately $50k that we would have paid for continuing our lease in the industrial park plus an increment for the additional space available in the new building. I was ecstatic about the new arrangements. This ecstacy was short-lived. The NRAO staff in Charlottesville gave us problems. First, having roughed out the terms of the lease, I called the NRAO contracts manager and attorney, Jay Marymor, in Charlottesville to ask him to work with the University attorney to finalize the lease. He told me that my commitments were irresponsible, that I could have done better, and that he wanted to renegotiate the deal. With great difficulty, I persuaded Jay to stick with what I had agreed to; I did not want to lose this marvelous opportunity that, time-wise, was a small window. Second, our capable NRAO administrator, Ted Riffe, opposed the deal because he suspected that Strittmatter was using us to get the University to build additional space that the astronomy department would take over as soon as our five-year lease expired. I got around that one by telling Peter Strittmatter exactly why Ted Riffe objected to the deal. Peter gave me his word that the space would be ours as long as we needed it and, consequently, Ted Riffe allowed us to proceed. As time has shown, Peter Strittmatter kept his word. There was one last hurdle to be cleared. At the eleventh hour, the Regents of the University of Arizona sent word they were reluctant to approve our lease until they were satisfied that the lease rate was comparable to similar commercial space in that area of the city. Here was a possible dealbreaker. We had estimated the $70k figure from what we would have paid to continue in the industrial park rather than from a survey of comparable rates for office space. When we actually considered commercial rates for equivalent property, we discovered that we should be paying a lot more for the university space. To accommodate the Regents, we declared the corridors, rest rooms, and the conference room of our floor to be “public areas” and subtracted this area from our “leased” space. Using a realistic rate for

68

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

comparable office space in the area, we were then able to multiply our usable office space by this rate to demonstrate that $70k was a reasonable annual rent. Considering the fluidity of university space in terms of space used for rest rooms, corridors, and meeting rooms, I believed this was a justifiable approach. The Regents accepted our proposal. We moved our offices to the new university space in the summer of 1984. Figure 5.11 shows the new building. By then, Bob Brown had replaced me as site director, so I never got to use the office I had designed! Associated with this move is my favorite story illustrating Peter Strittmatter’s great sense of humor. The initial bids for the addition to the astronomy building exceeded the university budget for the project. To reduce costs, Strittmatter decided to leave the stairwells and corridors unfinished. He eliminated vinyl-tiled floors in favor of painted floors, as well as the suspended acoustical ceilings in the hallways and in the labs. In contrast, because the university had initially told us that we would have to provide our own janitorial services, we had elected to install industrial carpet everywhere except in our labs. Carpet is the least expensive commercial flooring to care for. You just vacuum it—no labor-intensive mopping or waxing is needed. After we moved into the new addition, Strittmatter delighted taking visitors to our floor, calling it the “Penthouse,” showing them the carpet, and announcing in a loud voice that only national observatories could afford such luxurious flooring—knowing we had chosen it to save money. The new quarters were wonderful. The NRAO in Tucson had re-entered the academic community. We had quick access to the university’s extensive library system, to their precision machine shops and cryogenic facilities, to colleagues at Steward Observatory in the floors below us, to the Max-PlanckInstitut group that was building a submillimeter-wave telescope on Mount Graham, and to KPNO whose offices lay directly across the street. The NRAO librarian, Ellen Bouton, visited us from Charlottesville and set up a small but convenient library. It became easy for employees to take courses at the university. A few earned undergraduate and graduate degrees this way. Even lunches were better, being taken at the Student Union or nearby restaurants that catered to the university community rather than obtained from the mobile “roach coach” that came to the Forbes facilities at noon. It was easier to attend astronomy colloquia, and I even gave several of them. I’m still grateful to Peter Strittmatter for making the move possible.

5.9. NEW ASTRONOMER DORMITORIES ON KITT PEAK

5.9

69

New Astronomer Dormitories on Kitt Peak

The low-cost house trailers Ned had acquired in 1971-72 had served NRAO visitors well. But with time, they became increasingly run down and their maintenance cost began to soar. Because they had been built prior to the 1973 trailer safety law, they had built-in problems. From time to time, we experienced problems with their aluminum wiring. Figure 5.12 shows the rear side of the younger of the two trailers. The propane furnaces were always problematical. One winter, John Rather almost blew himself up trying to relight one of them, charring a new Harris tweed sports jacket. In those days, astronomers occasionally wore sports jackets while observing. Always persuasive, John managed to persuade the NRAO business office in Charlottesville to buy him a new one. Furthermore, the trailers always reeked of skunk spray and decaying rodent carcasses. Skunks loved the shelter beneath the trailers. In 1975, a French group was scheduled for observations of planets. Because the observing run was scheduled over Christmas holidays, they brought their spouses and children with them. You guessed it! All of them were sprayed by the skunks—a cute but pungent animal evidently little known in Europe. Their holiday in the American West must have been an enormous disappointment, although certainly memorable! By 1991, the trailers were truly in terrible condition. Even the astronomers began to complain, and they usually put up with almost anything. Darrel Emerson was the site manager at that time, and I volunteered to find replacement dormitories. His deputy, Phil Jewell, had envisioned replacements consisting of motel-style dormitories and a separate common building suitable for reading, conversation, and watching television, whereas I preferred a single, family-style building. But what kind of dormitory did the astronomers want? To find out, I emailed many on the NRAO users list to determine whether they would prefer a “house” model in which separate, lockable rooms, kitchen, and living room facilities would be contained in a single building—or a “motel” style building with a separate common building as Phil had envisioned. The respondents overwhelmingly preferred the motel concept. Knowing that money was hard to get, Dale and I approached several manufacturers of pre-fabricated buildings in our area. Our plan of extreme simplicity was based upon a great deal of experience with the trailers: no closed cabinets below sinks and no closets where astronomers could forget personal items. We wanted excellent acoustical isolation between rooms, good heating and cooling, durability, light-proof window shades, and im-

70

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

Figure 5.12: Dale Webb inspecting the back side of “Trailer No. 2.” Note the ad hoc duct work on the roof to distribute cool air from the “swamp cooler,” the corrugated skirt along the base of the trailer to discourage rodents, and sundry cables connecting a TV antenna to a TV in the lounge of the trailer. In the left background is the dome of the millimeter-wave telescope. This trailer is the one from which the resident skunks sprayed the French visitors. penetrable rodent barriers. The quoted prices for such a design ranged from $58k to $125k, with a corresponding range in quality. When I reported this to the NRAO director, Paul Vanden Bout, he agreed to look for funding and, ultimately, asked AUI to lend us the money. The AUI president, Bob Hughes, generously decided to simply give us the money for the least expensive option, which would be a huge improvement over the trailers. We placed a contract with the lowest bidder, Marlette/Schult Homes, of Buckeye, Arizona, on August 15, 1991 [54]. KPNO lent a bulldozer and grader to prepare the sites, and the NRAO staff installed the electric, data, and telephone lines themselves in large-diameter plastic pipes connecting the buildings and buried in the ground. Somehow we managed to find sufficient money to furnish the rooms at “bargain-basement” prices. Wanting to equip each room with a good, reclining reading chair, we discovered that Tucson retail outlets wanted at

5.9. NEW ASTRONOMER DORMITORIES ON KITT PEAK

71

Figure 5.13: The entrances to rooms 1 through 3 of the “new dormitory.” Acquired at a astonishingly low cost, they provided a comfortable environment to visiting astronomers. least $300 for a small, basic La-Z-Boy recliner—way above our budget. Resourcefully, Dale Webb telephoned their manufacturer directly and obtained a price of $150 per chair if we ordered nine of them. We also ordered good quality but inexpensive 7-ft long, wooden beds from the “General Services Adminstration (GSA),” and Hunter “Original” overhead fans directly from Hunter manufacturing, saving a great deal of money in the process. Telescope operator Duane Clark and I personally installed the Hunter ceiling fans in every room. Perhaps most important was that the new dormitories had a bedroom that was accessible to handicapped astronomers, unlike the old trailers. Figure 5.13 shows the entrances of rooms 1 through 3. Note that each room had its own climate control, separate entrance, and sound-resistant solid-wood door. The surprisingly durable light-green exterior siding was made of pressed cardboard! These dorm rooms proved a huge success with visiting astronomers accustomed to the old cramped rooms of the trailers, always reeking of skunk odor and decaying rodent carcasses. However, such perceived plushness may have damaged the “lean and mean” image that I

72

CHAPTER 5. EXPANDING THE TUCSON FACILITIES

Figure 5.14: The common building or “Library” under construction. (and, Darrel Emerson as well) had long cultivated for Tucson Operations. The only disappointment was that the rooms were not sufficiently isolated acoustically from each other. Although we knew the appropriate solution, we could not afford it. To truly isolate the rooms would require a dense masonry wall between each room, and a separate masonry support for the floor of each room. This was just not possible in the prefabricated, trailertype design compatible with our limited budget. Nonetheless, astronomers seemed considerate of their neighbors. Moreover, the shared schedule of the telescope usually meant that one group slept while the other observed. And the common-area building, shown in Figure 5.14, provided space for astronomers out of the dormitory rooms when not sleeping and not observing. The US Forest Service (USFS) happily took possession of the old trailers. While unacceptable for our use, evidently they were considered good quality by the USFS compared to other facilities they had.

Chapter 6

Providing Adequate Electricity From the beginning, reliable electric power was a problem for telescope operations.

6.1

External Power

The 36-ft telescope weathered storm winds by shutting the slit of the astrodome. This protection was only as reliable as the electric power supplied to the slit motors. We soon found that the Tri-County Electric Cooperative (Trico), a part of the Rural Electric Authority (REA) in southern Arizona, had problems maintaining electric service to Kitt Peak. Their fleet of trucks was too small for the enormous area Trico served. And the small dieselelectric generator purchased in 1972-3 could not carry the full load of the millimeter-wave telescope—a consequence of inadequate financial support from the NRAO. Searching through the excess property lists in February, 1977, [55] Bobby Ulich and Dale Webb found a pair of huge, two-stroke, military diesel generators manufactured in the 1940s by the Detroit Diesel Corporation. Available as excess government property, I decided to take them. This was my first education about the economies of surplus equipment. Such acquisitions may look cheap but, often, they’re not. When we received the generators, we sent them to the Detroit Diesel service agency in Tucson (Neil’s Detroit Diesel) who discovered one had a cracked head. Furthermore, we had to redo one of the weatherproof enclosures. $5,700 later, we installed one of them that had been fixed by cannibalizing the other. The remaining 73

74

CHAPTER 6. PROVIDING ADEQUATE ELECTRICITY

carcass was sent to “the Meadow,” the storage boneyard on Kitt Peak in case we needed additional spare parts. Moreover, we found the actual generator was configured to produce power only in the “delta” configuration, the opposite of the “wye” configuration in which the telescope power system had been wired. The output could not be changed by a simple reconnection of the electric taps. More expense. Finally, we were told the importance of periodically bringing the generators to an 80% load capacity to exercise them. Unfortunately, these generators were so large that connecting all of the station load could not exceed 50% of their capacity. Although there was now sufficient emergency power to close the dome and run the cryogenic compressors to keep the receivers cool, the solution was more expensive than we had anticipated, and was not sized correctly. It was another example of the pitfalls of doing things too cheaply. Eventually, we found the money to order and install a new properly sized and wired, 155 kVA Caterpillar diesel generator. We kept the Detroit Diesel as a backup for the backup! Figure 6.1 shows the arrangement in 1989. Later, in November, 1992, as the connected load grew, the Caterpillar became the backup to a larger capacity (200 kVA) Onan diesel generator, replacing the old Detroit Diesel.

6.2

Papago Tribal Utility

In the late 1970s, the observatories on Kitt Peak learned that our landlords, the Papago Tribe (now Tohono O’odham), had purchased the electric distribution system on their reservation from Trico. This system supplied power to Kitt Peak. Not subject to regulation by the Arizona Utility Commission, the tribe immediately hiked the rates for electric power. As their largest account, KPNO decided to generate their own power and, they hoped, to persuade the Papago Tribal Utility Authority (PTUA) to reduce their rates. Not a chance! The cost of generating power was significantly larger than the new cost of PTUA power, and the tribe knew it. KPNO even investigated building a special transmission line from Tucson to Kitt Peak to allow them to buy power from Tucson Electric Power Corporation. After a month or so of generating their own power, KPNO gave up and returned to buying their electricity from the PTUA. Soon after this buyout, one of the transformers at the end of “our” 3phase 480 v power line failed. The convention is that the responsible electric

6.2. PAPAGO TRIBAL UTILITY

75

Figure 6.1: Generator backing up a generator near the 12-m telescope in 1989. Far left, a new 4-stroke 155 kVA Caterpillar diesel generator; middle left, an inter-connection panel; center, the refurbished WWII 2-stroke Detroit Diesel generator; right, a metal storage shed. The view is to the south of the dome. The weather station may be seen in the background.

power company makes the repairs at their expense. In this case, the NSF had paid for the power line and title had been turned over to Trico Electric Cooperative. With the purchase of this distribution system from Trico, the PTUA now held title—or so we thought. However, the US Government had inserted a clause in the transfer agreement after the line was built [56]. This clause stated that, in case of a sale of the REA provider, title to the power distribution system would revert to the US Government via KPNO [57]. Therefore, the PTUA did not hold title to the failed transformer and was not obligated to replace it. Accordingly, the NRAO bought three transformers and gave them to PTUA for the repair of all three phases. To their credit, we soon found that the PTUA gave us better service than Trico. Their service area was far smaller than Trico’s, and it took less time for their truck to get to our telescope when trouble occurred.

76

6.3

CHAPTER 6. PROVIDING ADEQUATE ELECTRICITY

Ground Currents

Since observing began in 1967, more and more electrical equipment had been added to the 36-ft telescope. Much of this had been done by electricians on an ad hoc basis, with minimal attention being paid to ground currents. The problem stemmed from not knowing where to bond neutral lines to ground in the now complicated network of interconnections. Consequently, different devices often had small potentials with respect to each other, causing small currents to flow from equipment rack to equipment rack. This was unacceptable for reliable operation of the telescope and its extremely sensitive receivers. Fortunately, the ideal solution was at hand. Bob Stevens had been an electrician for KPNO for many years. Since then, he had received an EE degree in power engineering, had passed the Professional Engineer (PE) exam, and had started his own electrical engineering company. Bob not only knew engineering theory but also understood the requirements of astronomical observatories. In 1977, for $30k, he designed a new electric distribution system for the telescope and its dome-enclosed laboratories, and rewired this space that summer. The NRAO chief engineer, Buck Perry, also a EE power engineer with a PE certificate, approved (and admired) Bob’s design. Our ground loop problems were solved.

Chapter 7

Lightning and Kitt Peak

Figure 7.1: Lightning strikes on Kitt Peak, Arizona, with respect to the domes of the optical telescopes. The NRAO 36-ft telescope lies unseen c 1972 Gary Ladd. Used with behind the photographer, at left. Copyright  permission. Lightning is a severe problem for telescopes (or any metal structure) located on Southwestern mountain tops. Figure 7.1 is a photograph of strikes typically experienced from thunderstorms. In Tucson, lightning storms are 77

78

CHAPTER 7. LIGHTNING AND KITT PEAK

common in the “monsoon” season of late July, August, and early September. Furthermore, the natural conductivity of the granite soil of which Kitt Peak is composed is poor. For the 36-ft telescope, the builders created a master electrical ground by burying a large grid of copper wire and connecting all grounds to it. To improve this when the sewage system was repaired, we installed a huge grid of large diameter copper wire, a 20 by 20 array with each row 2 ft apart, in the leaching field itself, hoping that the moisture would improve the connection of the ground to the mountain soil. (See § 5.6) Nonetheless, the astrodome of the 36-ft telescope was much higher than the surrounding terrain. Although two conventional lightning rods were mounted on the top of the door arches, lightning often struck the dome. Unlike earlier optical telescopes, a radio telescope consists entirely of sensitive electronic circuits. Lightning strikes could cause enormous damage. For example, Bob Freund remembers filling a large styrofoam cup with dead integrated circuit chips (ICs) after just one particular damaging strike. Recovering from this one strike took several days. There were others—usually, one bad one per season. The solution seemed to come by trial and error. In 1977, we installed a surge protection system to protect particularly vital parts of the electronics and computers. The first Transtector system was installed in the early 1980s to protect critical equipment only. In 1984 when the telescope reflector was changed to 12 m, the optical shaft encoders were replaced with rotary transformers (Inductosyns), and their signals were coupled optically to the control electronics. In addition, Bob Freund gradually replaced most of the signal connections with optical fiber over the years. This optical coupling helped shield the telescope. In the late 1990s, we installed a Transtech protection system on all power systems.

Chapter 8

Software The 36-ft telescope came into operation at a time when computers were becoming increasingly used. An altitude-azimuth telescope, unlike the equatorially mounted 140-ft, the 36-ft telescope had been planned from its inception to employ a computer to convert the astronomical coordinates of hour angle and declination to altitude and azimuth.

8.1

The First Version

The first computer was a Honeywell DDP-116, which was considered a marvel of power and compactness at the time. It came in three, six-foot high, relay racks and had 48k of memory [17]. Wally Weller wrote the first telescope-control program for this computer, but the program was so complex that, when tested, it could not point the telescope. Emily Kitchen started again from scratch and wrote a simpler program that worked well, with modifications made from time to time as bugs became apparent. The DDP-116 was crude compared to modern computers. Booting it required typing a few words and then loading a punched-paper tape. Astronomical coordinates were entered via punch cards, and pointing corrections via hand keys. Astronomers complained. Johann Schraml [22], an NRAO astronomer/programmer now at the Max-Planck observatory (MPIfR) in Bonn, remembers a wonderful story about the early pointing system. After improvements to the pointing program, it happened that there was an AUI meeting in Tucson in which the Trustees visited the 36-ft telescope. On Kitt Peak, as he passed the telescope, someone asked Dave Heeschen how the pointing was coming. Dave said he would show them. He went in, turned on the power generator, 79

CHAPTER 8. SOFTWARE

80

opened the dome, turned on the computer, keyed in the computer loader, read in the program, turned on the receiver, read in coordinates for Orion A, and pressed “go.” Off went the telescope to the astronomical source, and the paper chart recorder indicated a signal. Inserting pointing offsets by the thumb wheels diminished the signal, indicating that the telescope really was pointing correctly. No one asked if the pointing of the 36-ft telescope was operational after that. What they did not know was that Dave had anticipated the question and had practiced the demonstration the day before.

8.2

Implementation of Green Bank Software

Initially, no one envisioned spectroscopy for the 36-ft telescope and there was no provision for it in the software. The 36-ft was designed to be a continuum instrument only. Eventually, the NRAO had realized that the 36-ft should have a spectroscopic capability. Sandy Weinreb had selected a Honeywell 316 to be the standard telescope computer for all the NRAO telescopes. The 36-ft was not to be an exception. The larger 316 could not only point the telescope but also could process spectra. The spectral-line reduction program was essentially a clone of what was used at the 140-ft telescope in Green Bank. The observations were taped at the telescope, and the tapes sent to Tucson to be processed on the KPNO CDC mainframe computer. The problem was that astronomers could not see the results of their observing until 24 hours later. This was an unacceptable situation for hunters of molecular lines, where the next astronomical target and frequency depended upon the outcome of previous observations.

8.3

FORTH

To accommodate the Bell Laboratory search for CO emission in 1970, Bob Wilson brought a computer with them to acquire the observations from a filter bank and process them. Unlike the NRAO system, this allowed realtime assessment of the spectra. This situation was enormously symbiotic. In the early 1970s, Chuck Moore had arrived at the NRAO offices in Charlottesville from Mohasco Industries, Amsterdam, New York, where he had developed an efficient, stackoriented programming language called FORTH [58, 59] for an IBM 1130. The need for a stand-alone program for controlling the 36-ft telescope and

8.3. FORTH

81

processing the data in real time was a perfect match for Chuck. Soon after arriving, Chuck had installed the language on one of the NRAO’s new Honeywell 316 computers, as well as on an IBM 360-50 mainframe computer. By the summer of 1972 (maybe, 1971), the FORTH-equipped Honeywell 316 had been shipped to Tucson. The DDP-116 pointed the telescope and the new 316 processed the data using FORTH [17]. Real-time spectra were now available via a Tektronix vector storage scope and a hard-copy printer. By the next summer, Chuck, Bess Rather, and Mike Hollis had improved the pointing and data reduction system to run on a single DEC PDP-11 computer at the 36-ft telescope. The display system remained the same. FORTH is conceptually different from other languages: “speaking of a FORTH program is sloppy, for FORTH is the program” [60]. It is the operating system as well as the programming language—more like a sophisticated calculator than a compiler. It is extremely efficient in its use of computer memory, which was expensive in those days when memory consisted of a network of magnetic beads. It was excellent for controlling hardware. However, it was not particularly portable. Being also an operating system, it needed to be tailored for the hardware interface unique to the computer it ran on. Nonetheless, it was an enormous advance in computer language in terms of its computing ability with respect to the memory required. When I arrived in Tucson, FORTH was having teething problems— despite its huge improvement over the earlier computer programs. It was new, a programming state where perfection never occurs. The fixes implemented during telescope maintenance were later checked out by astronomers after the programmers had returned to Tucson. If one caused a bug, the astronomers had to live with the consequences until the programming staff could return. To minimize these problems, I asked that two versions be made: the standard, dependable version, and the beta version. Only the NRAO staff would be allowed to use the beta version until it proved to be reliable, when it would then become the standard version. To support “off-line” development of the program, we bought a second DEC PDP11-40 for the Tucson offices to facilitate development without risking astronomical observing. The dedication of Mike Hollis, the duplicate downtown PDP-11, and the inherent power of FORTH finally won over the astronomers. For reliability, I had asked Mike never to make program changes unless our own staff had the opportunity to check them out first. Years later, Mike told me that, in fact, he had to do that occasionally to support some observing programs but took great care that these changes were so reliable that I would never find out about them! For this reason alone, I was sure that Mike—shown in

82

CHAPTER 8. SOFTWARE

Figure 8.1: Mike Hollis, astronomer and FORTH programmer, and Max Thomas, secretary, in the lobby of the NRAO offices in the Forbes industrial park in Tucson circa 1976. Author photo.

Figure 8.1 with Maxine Thomas—had been an excellent engineering officer when he served aboard nuclear submarines. FORTH proved to be a great success. Our sister observatory KPNO adopted it for use at their optical telescopes, as did many other observatories in later years. Bess Rather and Chuck Moore left the NRAO at the end of 1973 to form FORTH, Inc., which is still in business today. After a stint at the Arecibo radio telescope, Ned Conklin joined them. Since then, it has been implemented on myriads of computer types. My personal favorite was the hand-held SuperTracker carried by every FedEx delivery agent. There are many, many more applications. As testimony to its usefulness, FORTH—in the form with the original kernel, actually—remained in use as the control computer program at the 36-ft telescope until about 1990, when it was replaced by a Sun workstation and Unix—a remarkable life span of almost 19 years in an era of enormous technological growth.

8.4. THE VAX YEARS

8.4

83

The VAX Years

As the observing capability of the 36-ft telescope grew, astronomers asked for more and more data processing ability. Tom Cram, in Charlottesville, had developed a FORTRAN-based analysis system used with the 140-ft and 300-ft telescopes in Green Bank. Based upon the People-Oriented Processing System (POPS) computer language invented by Jerry Hudson of the NRAO Charlottesville, these FORTRAN programs were not only extremely flexible but allowed astronomers to write and implement their own routines by sequentially stringing together “verbs” that would call individual analytic programs. Furthermore, many astronomers used the Green Bank telescopes as well as the 36-ft telescope and wanted the data processing software to function in the same way at all three telescopes. In 1979, Mike Hollis announced he was leaving the NRAO for an astronomer position with NASA. His departure was a reaction to a conversation with Mort Roberts, the NRAO director at that time. Now equipped with an astronomy Ph.D. and with a record of published research papers, Mike had asked Mort to change his official title from Programming Analyst I to System Scientist, which Mike considered more representative of the work he was actually doing. Mort had refused. It would be a big loss for Tucson Operations. Mike was not only a self-motivated, tireless, first-rate programmer, he also knew astronomy. It was not only necessary to find a replacement for Mike but also a time to re-assess the direction of Tucson computing. Modern computers provided much more compute power, and astronomers had become accustomed to having lots of bells and whistles at their disposal. Luckily, Betty Stobie joined us from the NRAO computing staff in Charlottesville in November, 1979. I wanted to install the POPS data reduction system at the 36-ft telescope. Furthermore, a failure of the single computer at the telescope meant losing data as well, even though we wrote new data to tape as soon as possible after acquiring a “scan.” Betty had had a lot of experience maintaining that software in Charlottesville and Green Bank, and was the obvious choice. Happily for us, she and her family agreed to move to Tucson. FORTH would be new to her but she would be able to learn it quickly. The first thing she did was to install the back-East POPS reduction system in a second DEC PDP 11-40 that we had in the Forbes office [61]. The operating system was RT-11. This computer was then moved to the telescope, and we bought a PDP 11-44 for the Forbes staff, running RXS-11 as its operating system. After installation, the original PDP 11 controlled

84

CHAPTER 8. SOFTWARE

the telescope, performed the observations, and ported the data to the second PDP 11. This arrangement made the data reduction system operate exactly like the system used by the Green Bank telescopes but on different computers. However, all was not perfect. A few astronomers preferred the old FORTH data reduction system. To accommodate them, Betty wrote a FORTRAN program for the second PDP to emulate the entire FORTH reduction system in all respects, including the function of the single character commands [61]. This was a lot of work. The Tucson staff often went to great effort to keep the visiting astronomers happy. This situation worked very well for several years until we needed a larger and faster computer to process data. Our plan was to buy two VAX 780s. One would be installed at the telescope; the other, in the Forbes office complex to facilitate development and general computing. At the telescope, the FORTH system would continue to control the telescope and the datataking but would pipe the actual data to the reduction system running on the VAX 780. Those wanting to continue using the FORTH reduction software could do so. Again, dealing with Charlottesville was not easy. Green Bank, Charlottesville, and the VLA in New Mexico had fallen in love with Modcomp computers, which provided a great deal of computing power for a moderate price. In comparison, the VAX computers were expensive, especially when one included the license fee for their admittedly excellent software. We wanted VAXes because they were directly compatible with Unibus, a proprietary DEC data bus to which all of our hardware was designed to connect. Changing to Modcomps would have required rebuilding all of this proven hardware. Nonetheless, the head of the Charlottesville Computer Division, Bob Burns, was insisting that we choose Modcomps. When we hesitated, Hein Hvatum embargoed any computer purchase for us for six months until the Charlottesville Computer Division and Tucson reached agreement. In the end, Bob Burns supported our preference of VAXes but would only buy the smaller 750 models as a compromise to their higher costs. Parenthetically, I recall that Betty was distinguished not only by her willingness to work hard but by her incredible memory. Looking through her code, I was always astonished by the absence of comment lines. Absolutely necessary to me because of my poor memory, comment lines were unnecessary to Betty because she could remember the detailed function of her code lines years after she had written them. From time to time we discussed it, but I could never persuade her to use them liberally. Evidently, they took up too much valuable space in the computer!

8.5. MOVING TO UNIX

8.5

85

Moving to Unix

By 1986, Darrel Emerson had arrived as site manager and, unlike me, was especially skilled with computers and software. He immediately saw the limitations imposed by the now-ancient FORTH control system and decided to replace it. This was a process that took at least two years. According to Darrel [62], there were several plans. Think of them as paths through a dense forest to a distant clearing. No one knew which path would prove best. Plan 1 was to transfer everything to the VAX 750, which had been used only to analyze data transferred to it from the FORTH system. All “real-time” tasks were to be done in powerful mini-computers, each using an operating system know as VxWorks designed especially for this kind of work. The VAX would control each of them. Unfortunately, the VAX 750 was not powerful enough and was soon replaced by a new MicroVAX computer borrowed from the Charlottesville computer group. Exploring this path involved a great deal of learning how to separate the real-time tasks which had been combined in the old FORTH system from each other. New special-purpose hardware had to be designed and constructed. The first stage was to make a special purpose “tracker-servo” computer that would take over the functions of FORTH. This was implemented by November 1989, running in parallel with the FORTH control computer. Furthermore, our staff had to build a new “interface bus,” a digital language (protocol) that allowed the system components to talk to each other. The first one they chose, Hewlett Packard’s General Purpose Instrument Bus (GPIB), did not work well enough, and they soon adopted the newer Ethernet bus. These jobs took a long time. Assembling this system soon showed that it would not be fast enough to control the telescope, acquire the data, and allow the astronomer to process it. The work could not disrupt observing. This meant that most changes had to be implemented and tested in the summer when the telescope was not used. There were circumstances where the computer people were scheduled on the telescope for testing, but Darrel kept this to a minimum. The work involved sociological changes as well. Traditionally, digital engineers tend to solve problems with digital hardware; analogue engineers, with analogue circuits; programmers, with software. There was not much overlap between specialties. Specialists were comfortable with the techniques they had successfully used for years. Rather, most complicated projects had interfaces where different technologies met but did not overlap. This project was the best of all worlds. While the hardware people had to

CHAPTER 8. SOFTWARE

86

learn about software and vice versa, the allocation of specific tasks to minicomputers established “realms” managed by specific engineers. Although the technical competence of the engineers widened considerably, they were still able to be master of the specific task assigned to them. Plan 2 was to replace the MicroVAX with a Sun workstation running Unix. Ultimately, this path worked and is the system in use today. Obviously, this route was not a completely new start. It made use of what had been learned earlier in Plan 1 and the new hardware. After completing the project, Darrel Emerson, the Tucson manager, told me about his experience with TimeLine, a critical path planner that I had tried to persuade him to use to track the software project. After entering all of the tasks involved in the project, he found that TimeLine predicted the job would take at least two years. He rejected this projection as a spurious result, believing that the job could be completed in a few months. Actually, TimeLine was correct. The project did take two years!

8.6 8.6.1

Off-Line Data Reduction Spectroscopy

All of the earlier spectral-line reduction software at the 36-ft and, later, the 12-m telescope was not particularly portable. Although copies existed on NRAO computers in Tucson and Charlottesville, most astronomers tried hard to reduce their data before leaving the telescope. The FORTH and Unipops (FORTRAN) programs required a high level of skill to install on university computers, and so “home” processing was not an option for most of them. To fill this need, an NRAO astronomer in Charlottesville, Harvey Liszt wrote an entirely new analysis program that ran on early models of “Personal Computers” (IBM-type PCs). The first version took him six months to write. Written in Pascal for an early AT&T (American Telephone and Telegraph) PC using an Intel 8086 processor, this program, called DrawSpec, would do almost anything the millimeter-wave telescope software could do, except for producing fancy, annotated, journal-quality spectral plots. Remarkably compact, DrawSpec would run on any PC and, hence, was extremely portable. It was a “God Send” to many millimeter-wave astronomers. Even to this day, I’m not sure that the NRAO management is aware how important this software was. Many years later, Institut de Radio Astronomie Millimetrique ` (IRAM) astronomers converted their powerful analysis software, “CLASS” [63], into a FORTRAN version that would run on any IBM

8.6. OFF-LINE DATA REDUCTION

87

PC. It really did not do much more than the older DrawSpec but was identical to the Unix version that was by then commonly used by most millimeterwave telescopes to reduce the radio spectra. It was now very easy to process spectra observations at home and to produce journal-quality illustrations of the results.

8.6.2

Continuum Mapping

Observations of continuum emission from “point sources” is an easy business with a single-dish telescope, even at millimeter-wavelengths. A point source is a source of radio waves with angular dimensions small compared to the telescope beam. The astronomer calibrates the telescope, checks the pointing accuracy, measures the intensity of the source, and corrects for the atmospheric extinction. No software more sophisticated than a pocket calculator is required; a child could do it. The challenge lies in the interpretation of the results. In contrast, mapping the continuum emission of spatially extended sources is difficult. First, extended objects often fill or exceed the width of the telescope beam, thereby radiating not only into the main part of the beam but also into its sidelobes. Because most parabolic telescopes aren’t perfect geometrical figures, the sidelobes are not easy to predict. They have to be measured. Second, finding a good calibration source for this kind of measurement can be difficult. Planets are the best choices because their angular size is known and their emission is usually circularly symmetric. But the planetary emission is not perfectly known as a function of wavelength, thereby fundamentally limiting the absolute accuracy of the calibration [64]. Three radio astronomers provided the rescue. Beginning at the University of Manchester, Glyn Haslam developed a collection of FORTRAN subroutines, known generally as NOD2 [65], to analyze spatially extended structures. He and colleagues extended these greatly during his subsequent career at the MPIfR in Bonn. While there, Darrel Emerson, Uli Klein, and Glyn [66] created a technique for mapping extended sources by rapidly switching the telescope beam in azimuth, driving this now double beam (one positive, one negative) across the extended source at a sequence of elevations, and thereby creating a set of raster scans of the extended source convolved with the switched beam. Subsequent deconvolution of these scans produced an intensity map that could then be analyzed with the Haslam software. Together, the switched-beam observing technique and NOD2 formed an effective system for accurately mapping the continuum emission of extended radio sources with a single-dish telescope. One calibrates the system by

88

CHAPTER 8. SOFTWARE

mapping a planet of known intensity, and forcing the deconvolved planetary map to produce the “correct” integrated flux at that wavelength. When astronomer Chris Salter arrived in Tucson in 1984, he installed the system on the VAX computers of NRAO’s 12-m telescope. It was a powerful addition to the NRAO observing software; a huge step forward in the capability of the NRAO telescope. Darrel Emerson added to the Tucson installation of NOD2 after he arrived in Tucson in 1986. I (with Phil Jewell, Chris Salter, and May Kassim) personally made heavy use of the system by mapping almost every extended dust cloud I could detect with the 12-m telescope to determine their continuum spectra. Chris Salter is the most unusual radio astronomer I have met. I admire him greatly. As an English bachelor, he seemed to roam from radio observatory to radio observatory, spending a year or two at each, and carrying a backpack containing everything he owned. I suspect this consisted of ten or so logo T-shirts, several pairs of jeans, and perhaps a book or two. In a sense, he was like a wandering troubadour of the Middle Ages, going from castle to castle, spending a little time and doing some good at each stop. Eventually, his situation changed greatly. Several years after his Tucson stay, he married another radio astronomer, is now a father, and became well ensconced at the giant Arecibo radio telescope in Puerto Rico. But nothing in radio astronomy is simple. Glyn Haslam had taken great care to write his subroutines in standardized FORTRAN so they would compile on any computer. While at IRAM, before joining the NRAO, Darrel and others adapted the NOD2 system to the VAX computers purchased for the new 30-m telescope. He also added additional subroutines. To improve NOD2 by making the system execute more rapidly and use computer memory more efficiently, he had incorporated the faster specialized, non-standard FORTRAN functions allowed by the VAX compiler. These changes improved NOD2 significantly. But after these changes, his improved NOD2 system would only run on a VAX computer; it was no longer compilerindependent software. This was also the version Chris Salter had installed at the Tucson telescope. A few years later, when the VAX-equipped United Kingdom Infrared Telescope (UKIRT) telescope on Hawaii was being adapted for mapping extended sources in the millimeter and submillimeter wavelength ranges, Darrel sent them this VAX-compatible NOD2 system. Glyn Haslam learned of this by accident and was furious. First, he had copyrighted the code to guarantee that the code would remain standardized FORTRAN and felt that Darrel had no right to distribute “his” NOD2 system, no matter how well-intentioned the gift. Second, he was very annoyed that the VAX version

8.6. OFF-LINE DATA REDUCTION

89

wasn’t the “official” version which he alone wanted to be able to distribute to ensure that the NOD2 system was identical at every telescope. The result was that Glyn Haslam “cut off” the NRAO from any improvements to the code. He even threatened to sue us if we gave anyone else a copy. Even in 1988 through 1991 when I intermittently worked at the MPIfR in Bonn analyzing dust cloud observations taken with the NOD2 system at the IRAM 30-meter telescope, Glyn would only allow me access to the compiled NOD2 code. Under no circumstances would he let me see the FORTRAN itself, to prevent me from bringing the improved code to Tucson. There were lots of bad feelings in spite of good intentions on either side. It really was quite a rift between two old friends!

Chapter 9

Millimeter-Wave Electronics Building sensitive receivers for millimeter wavelengths was a challenge from the beginning. The first receivers were room temperature mixers installed within the Stirling mount at prime focus.

9.1

Local Oscillators

A big challenge was the local oscillator (LO). Figure 9.1 is a sketch taken from one of my 1972 observing notebooks—which is slightly incorrect because the phase-locking is what ultimately stabilizes the local oscillator. The LO consisted of the output from a klystron oscillator, phase-locked to a reference signal. The locking device, known as the Weinreb lockbox, was an early hallmark of the 36-ft telescope. Designed by Sandy Weinreb, it generated an appropriate signal of, for example, 115 GHz from a klystron to mix with the incoming astronomical signals to produce an intermediate frequency (IF) of 1,390 MHz. Understanding the design was essential to using it. Using a variable oscillator to produce a clean frequency signal in the range of 100 MHz, the astronomer calculated which harmonic (n ≈ 20) of a 100-MHz signal was needed to produce a phase-locked 2-GHz reference signal that was ±400 MHz on either side of what was needed at the frontend mixer. Mixing this with the klystron output produced a 400-MHz signal that, in turn, was mixed with a variable reference signal used to steer the receiver. The error signal between these two 400-MHz signals, ∆f , went to a frequency-to-voltage converter that generated a small DC voltage to adjust slightly the high voltage, and thereby the frequency tuning, of the klystron. When all was correctly set up, the astronomer steered the receiver 91

92

CHAPTER 9. MILLIMETER-WAVE ELECTRONICS

frequency by adjusting either the 400-MHz signal or the reference oscillator. When the klystron was sufficiently close to the reference frequency, a phaselocking device took over and locked the phase and frequency precisely to the correct values. In 1972, this process was a tour de force! With this design for the LO, any one of four LO frequencies could be used to convert the “sky frequency” into the IF amplifier frequency: Fsky ± 1, 390 ± 400 MHz. If the offsets were all negative, the LO selection was called “Quad 1”; all positive, “Quad 4.” “Quad 2” referred to the situation where the LO frequency lay at −1, 390+400 MHz and, “Quad 3,” the reverse. While this arrangement could lead to errors (especially when you were tired), it also provided a great deal of flexibility in tuning the receiver. FORTH was able to calculate frequencies for each selection, taking into account the velocity of the telescope with respect to the Local Standard of Rest. Even if the astronomer understood the process, there were lots of mistakes to avoid. First, the klystron had a screwdriver adjustment for the resistance and capacitance in addition to the electronic tuning. This circuit “floated” at several thousand volts. It was often necessary to climb to the prime focus of the telescope, insert the screwdriver into the high-voltage locking system to adjust the klystron mechanically while avoiding a shock. Occasionally, you had to do this after going without sleep for a day or more. Second, it was easy to make a calculation mistake, which would tune the mixer to the wrong sideband. The lifetime of the klystrons could be short, sometimes less than 100 hours, even though they cost several thousands of dollars because each was handmade. While warranties usually were for a thousand hours (42 days when used continually), replacement could take months. Extending klystron lifetimes quickly became a major objective of the NRAO Electronics Division. The ultimate solution was to place the klystrons in a water-jacket to prevent them from overheating. Most of the klystrons operated at lower frequencies than needed for millimeter-wave mixers. The NRAO built frequency doublers and triplers and used their outputs as LOs. Initially, Varian supplied most of the klystrons. Soon, the OKI corporation of Japan entered the field with klystrons of a physically different design that proved more durable but did not cover as high frequencies or produce as much power as the Varian design. The new design required us to change how klystrons were mounted, which involved lots of machine work. What evolved was an interchangeable mixer plate; each plate containing either a Varian or OKI klystron with hardware and connections appropriate to that particular design. To cover all the wavelength bands, the NRAO had to keep a large num-

9.1. LOCAL OSCILLATORS

93

Figure 9.1: A sketch of the phase-locked local oscillator system in use in early 1972, taken from my observing notebook. The mixer was a “harmonic mixer,” one of whose inputs was the Nth harmonic of the reference oscillator. A note at the top right indicates the klystron was a Varian 2123A7, serial number 70246.

94

CHAPTER 9. MILLIMETER-WAVE ELECTRONICS

ber of klystrons at the telescope. Each had a narrow tuning range. This tied up a lot of money. As the receivers moved to higher and higher frequencies, the klystrons became physically smaller, had smaller percentage tuning ranges, and were more expensive. For example, a Varian 144- to 152-GHz tunable klystron would cost about $10k and might last only 100 hours. Even with high prices, Varian Associates could not make a profit and eventually stopped making them. Varian Associates had an interesting history. Russell Varian and his brother Sigurd invented the klystron oscillator tube that became the basis of microwave radar and helped the Allies win World War II. Sigurd was a pilot for Pan American World Airways. Russell was the more practical of the pair, finding ways to transform Sigurd’s imaginative ideas into workable devices. With colleagues, the two founded Varian Associates. Many people still refer to the two of them as simply “the Varian brothers.” As the years went by, a Berkeley graduate student rescued us from the klystron problem. Using inexpensive Gunn diode oscillators, in 1984, John Carlstrom (now an astronomy professor at the University of Chicago) mounted them into tunable cavities. The only exit from each cavity was a waveguide selected to transmit the desired range of LO frequencies; the waveguide served as a filter. A combination of mechanical and electronic tuning adjusted the oscillator to produce appropriate harmonics with adequate power, which would then leave the cavity via the waveguide. Not only were the Gunn diodes much cheaper than the klystrons, they had very long lifetimes.

9.2

Quasi-Optical Techniques

One of the most important refinements to Tucson receivers came from a new technology called “quasi-optics.” Early Tucson receivers tended to be millimeter-wave versions of those used at lower frequencies, like those in use in Green Bank. The mixer diode was mounted in a block, and a separate waveguide attached to that block brought in the LO signal. As John Payne wrote in 1979 [67], “The problem is that the higher the frequency the smaller the waveguide becomes. At a frequency of 150 GHz, for instance, the waveguide we use measures only 0.051 x 0.0225 inches and the problem of machining inside [a] waveguide of this size is pretty obvious.” The term quasi-optics recognizes that optical techniques can be used at millimeter and sub-millimeter wavelengths where the actual wavelength is smaller than the physical dimensions of the device. For example, Figure 9.2

9.2. QUASI-OPTICAL TECHNIQUES

95

Figure 9.2: Sketch [67] showing how quasi-optical techniques inject the local oscillator into a millimeter-wave mixer on the 36-ft telescope. The mixer lies within a vacuum chamber chamber and is cooled to 17K. See the text for an explanation.

96

CHAPTER 9. MILLIMETER-WAVE ELECTRONICS

sketches a typical situation where quasi-optics are used. The location is the vertex of the telescope, where the wavefront reflected downward from the subreflector enters the receiver. On its way to the mixer, the incoming wavefront first strikes two mirrors arranged like those of a Michelson interferometer. The lefthand mirror is made of fine wire grids that allow the LO to pass into the main axis of the radiation. An essential feature of these mirrors is that the wires lie at 45 deg forming the half-reflecting mirror to the plane of the paper. At Port 1, the incoming radiation splits. One half of the beam is reflected to the right, the other proceeding through the half-reflecting mirror at Port 2. The reflected and transmitted waves are in phase quadrature. As the polarized beam makes its way from reflecting surface to reflecting surface of the fully reflecting mirror on the right, the linear polarization changes at each reflection. Arriving at Port 4, the polarization has now changed by 90 deg, such that it now has the same orientation as the other linear polarization arriving by passing straight through the half-reflecting mirrors. Now, however, the two halves of the RF signal have a phase shift between them because of the extra path length of the righthand part. Changing the spacing between the corner mirrors changes the relative phase such that both RF halves arrive in phase at Port 4. The teflon lenses L1 and L2 shape the incoming LO and signal wavefronts to the circular wavefront needed by the mixer horn. This system of introducing the LO into the mixer works only because the wavelengths are smaller than the dimensions of the mirrors, in other words, because the environment is quasi-optical when sized by wavelength. This LO injection system can also minimize ripple in spectral baselines. The recombination of the RF signal split at Port 1 and recombined at Port 2 creates an amplitude sinusoid in the RF signal as a function of frequency. If a motor changes the distance between the two mirrors in the direction indicated by “Tune,” the frequencies at which the amplitude peaks change. In this way, the natural ripple occurring from multiple reflections between the reflector surface and other parts of the telescope will be smeared out in the time integration, leaving a “flat” spectroscopic baseline for astronomical observations. It was a brilliant solution to injecting LOs into millimeter-wave and submillimeter-wave mixer receivers. These techniques can be expanded further to process each polarization separately—and, indeed, were.

9.3. RECEIVERS

9.3

97

Receivers

The first receiver I used at the 36-ft telescope was the λ3-mm mixer with which we observed CO emission from the Milky Way. The date was late Spring, 1973. This receiver was a room-temperature version and was mounted at prime focus. The system temperature was about 1,500K at 115 GHz, single sideband (SSB). Used as a continuum receiver where both sidebands were detected, the system temperature was half that. Developing receivers was a major part of the electronics work. In the beginning, receivers were built in Green Bank and, then, in Charlottesville and shipped to Tucson fully assembled. The logic was that the Green Bank and Charlottesville electronics groups not only were experienced building receivers but they had the appropriate tools to do so. The drawback was that the Tucson group had neither of these assets. When a receiver failed, it was difficult to repair in Tucson. Furthermore, these eastern groups were unfamiliar with the Arizona climate and the peculiar demands it placed on the equipment. As always, the principal objective was lowering the noise temperature of the receivers. Emile Blum of Observatoire de Paris, then a member of the AUI Visiting Committee, suggested to Sandy Weinreb, then head of the NRAO Electronics Division, that the Schottky diodes could operate well as mixers when cooled to liquid nitrogen temperatures. In a test conducted in Charlottesville, Sandy found that the cooled diodes performed well indeed. The result was the construction of the first “cooled receiver,” designed by both Weinreb and Jesse Davis. In 1974, the NRAO Electronics Division delivered the first cooled receiver for the 36-ft telescope. It was huge and had to be mounted at the vertex of the telescope because of its weight. Figure 5.2 shows how it looked in 1981. This was the prototype for almost all subsequent receivers. This new receiver had an SSB system temperature of approximately 400K. Because sensitivity varies as the system temperature and the inverse square of the integration time, the new receiver took 14 times less time to reach the same peak-to-peak noise levels—a huge improvement. To make this system viable required a parallel development in cryogenic systems. The large demand for observing time with the 36-ft telescope required cryogenic refrigerators rather than dewars filled with liquid nitrogen. No one wanted to take the time to refill a dewar. Consequently, the NRAO technicians in Green Bank and in Tucson developed durable cooling units (evaporators) and nitrogen compressors by modifying units produced commercially.

98

CHAPTER 9. MILLIMETER-WAVE ELECTRONICS

As you would expect, operating at these temperatures also required developing appropriate mechanical components such as tuning motors and gear boxes. Retuning the receiver to change from one observing program to another had to be done while the receiver was still cold. As the years rolled by, receiver temperatures got better and better. In 1984, semiconductor-Insulator-semiconduction (SIS) diodes replaced the Schottky diodes as mixer elements. These generated less receiver noise, as can be seen in Figure 9.3, because their quantum mechanical nature produced higher conversion efficiencies, that is, the process of transferring the intensity envelope of the incoming RF signals to the IF signal that carried information to the detector and spectrometer. And being quantum mechanical devices with smaller energy gaps than Schottky diodes, they intrinsically generated less noise. The journey from ambient temperature to cryogenic receivers was a stepby-step process over many years. For example, unlike Schottky diodes, SIS mixers would not operate at room temperatures. In fact, they required “helium” temperatures of 4K and lower. The NRAO’s development of cryogenics systems made it possible to employ the SIS junctions. The first cryogenic system for an SIS λ3-mm receiver was a “batch” system, a dewar filled with liquid helium and with lowered pressure to reduce the boiling point of the helium below 4K. On a good day, the system would remain cold for about 24 hours—long enough for productive observing but not long enough to accommodate the demand from astronomers using the telescope. Closed-cycle refrigerators eventually arrived, making use of the JouleThompson (JT) effect used by most modern refrigeration systems. The expansion of compressed gas immediately after flowing through a small nozzle lowers the gas temperature enormously. While the first 4K JT closed loop system had been used for the λ3-mm Schottky mixer, it was developed further for SIS junctions and first applied to a λ1.3-mm mixer receiver. The results were spectacular, sometimes attaining temperatures as low as 3.5K or −270C. Other aspects of SIS junctions made them especially appropriate for millimeter wave and sub-millimeter wave receivers. First, they required much less LO power than Schottky diodes. This relaxed the power specifications for LOs. For example, the output of Gunn-diode oscillators was too puny for Schottky mixers. Second, they were much more appropriate for high frequency use, facilitating low-noise receivers in the 150-GHz and 230-GHz observing bands. Within the NRAO, this technical evolution resulted from a relation-

9.4. FAILURES

99

ship that had evolved between technical groups at Charlottesville (mixer blocks and junctions), Green Bank (cryogenics), and Tucson (quasi-optics and receiver construction). Each group had unique “know-how.” Precision machining was available in all three locations. Not all was well, however. The pressure for receiver reliability was enormous. People worked very hard. When something failed, there was occasional finger-pointing. However, the results showed that all three groups worked very well with each other, in the main. All of this progress was not due solely to the NRAO engineering skills, of course. Other groups at the California Institute of Technology (Caltech), Jet Propulsion Laboratory (JPL), University of California at Berkeley, University of Massachusetts, Bell Laboratories, and IRAM, had been also developing low-noise receivers for millimeter wave and sub-millimeter wave receivers. Not only were there ample opportunities for developers to exchange information at meetings but, occasionally, engineers from one group spent extended visits with other another group. The result was a highly productive technical symbiosis enjoyed by millimeter-wave radio astronomers everywhere. For example, physicist James Lamb joined Tucson Operations and developed some of our most sensitive receivers, then moved to IRAM in Grenoble, France, for several years, and is now developing sensitive electronics for the millimeter-wave interferometer of Caltech in the Owens Valley of California.

9.4

Failures

The NRAO electronics group had its share of failures. These are inevitable when one tries something new. And a national observatory must try new things if it is to remain at the cutting edge of technology. Not all of these initiatives will be successful, but many of them will turn out remarkably well. Actually, there weren’t very many failures, which is a testament to the great competence of the electronics group. But those that did occur were expensive.

9.4.1

Millimeter-Wave Parametric Amplifier

The first one I remember was a “49 GHz” parametric amplifier that began development in Charlottesville. The idea was to build a cooled parametric amplifier for the 36-ft telescope, which would be suitable for detecting spectral lines. Joachim Edrich, a Ph.D. engineer in Charlottesville’s Central Development Laboratory, undertook the job.

CHAPTER 9. MILLIMETER-WAVE ELECTRONICS

100 2000

SSB Receiver Temperature (K)

NRAO 36-ft & 12-m Telescopes 1500

90 GHz 150 GHz

1000

250 GHz

500

0 1970

1975

1980

1985

1990

1995

2000

Year

Figure 9.3: A plot of the single-sideband (SSB) receiver temperatures over the years. The big decreases were associated with changes from cooled to uncooled Schottky diodes and, later, from these to cooled SIS mixers. There was no 150 GHz receiver from 1982 through 1991.

It proved to be a challenging assignment. Edrich worked on it for years, even taking the project with him when he left the NRAO for a faculty position at the University of Denver. To his credit, he finally delivered the receiver. But although a technical triumph, it was an astronomical failure and was never used. To get the parametric amplifier functioning, Edrich had to make a number of compromises. One was using a narrow frequency band that could not be tuned very well. Another was shifting the frequency upward from the design specification. This bandwidth did not contain any spectral lines of interest to astronomers. It was not useful for spectroscopy. Furthermore, its narrow bandwidth made it useless to continuum astronomers. Counting manpower and expenditures, I estimate that the NRAO may have spent at least $500k on this project.

9.4. FAILURES

101

Figure 9.4: Diagram of the new λ1 mm bolometer. The bolometer module lies in liquid helium 4 (4 He). The module interior is cooled by helium 3 (3 He) to a temperature of about 0.3K, reached by pumping evaporated 3 He gas to lower the surface pressure and evaporation rate (and hence, temperature) of the liquid. The detector is a doped Germanium bolometer.

9.4.2

Millimeter-Wave Bolometer

Perhaps our most expensive failure was the attempt to get a working bolometer1 on the 36-ft telescope. Originally, Frank Low had proposed the design of the 36-ft telescope to accommodate his λ1 mm bolometer. Ironically, twenty years later, the modern Tucson staff was unable to build one that worked on the same telescope. Technically, the new NRAO design was a masterpiece. Figure 9.4 shows the details. The bolometer’s detector was packaged within a module cooled by liquid helium 3 isotope (3 He) to a temperature of 0.3K, reached by low1 In astronomy, a detector extremely sensitive to impinging radiation. From the Greek word for beam of light.

102

CHAPTER 9. MILLIMETER-WAVE ELECTRONICS

ering the surface pressure of the 3 He with a vacuum pump. This entire module was immersed in liquid helium 4 (4 He) at a temperature of 1.3K, also obtained by pumping to lower the normal boiling point of 4K. Radiation entered the bolometer through a 200- to 300-GHz band-pass filter, was focused by a teflon lens, and passed through a quartz window into the cryogenic environment. The combination of a wide-band filter matched to the entire atmospheric window and a detector maintained at the extraordinarily low temperature of 0.3K produced a supremely sensitive instrument—in theory, at least. Bolometers offer astronomers two extreme possibilities for use. One is the sensitivity to detect nearly anything in the radio sky, achieved by removing the bandpass filter. In fact, astronomers from the French Planetary Institute had successfully operated a bolometer in this way on the 36-ft telescope in the late 1970s. In this case, only the atmospheric windows themselves limit the bandwidth. (See the transmission curves in Figure 1.4.) This observing technique leads to problems in interpreting detections astrophysically. Such a wide bandwidth makes it difficult to determine the detailed spectral distribution radiated by the cosmic source. The most intense radiation of such a signal could lie anywhere within the acceptance bandwidth of the bolometer, that is, from 90 through 280 GHz. For planetary observations at high frequencies, this may not be a problem because the planetary emission is more or less black-body over a wide frequency range. For other types of radio sources, it is a big problem. The other approach is to restrict the incoming radiation to a well-defined range, that is, to use a bandpass filter. In this way, a detection offers the possibility for some astrophysical interpretation but at the price of significantly reduced sensitivity. One would then know the intensity of the detected radiation as a function of frequency, that is, be able to distinguish between the radiation intensity in the individual λ3 mm, 2 mm, and 1.3 mm bands passed by the earth’s atmosphere. For objects outside the solar system such as the interstellar dust clouds in which stars form, this kind of spectral information is essential to understanding the astrophysical environment of the detected radiation. Development of these important band-pass filters initially occurred at Queen Mary College (QMC) of the University of London. Their physics group had specialized in bolometer observations and, in fact, were using one to observe from the United Kingdom astronomical observatory on the summit of Tenerife in the Canary Islands and another on the summit of Mauna Kea in Hawaii. Peter Ade, a graduate student at QMC at the time, had developed technology to make wideband filters with high transmissivity.

9.4. FAILURES

103

Completing his Ph.D., Peter joined the Tucson group specifically to develop a new detector for the bolometer. The NRAO’s Jesse Davis worked with him and, in later years, perfected these filters so well that his company became the principal supplier. Remarkably, the transmissivity of these filters approached 70% at the centers of their bands! The entire bolometer group was not small. To make this new bolometer a success, the NRAO enlisted the skills of other specialized groups. The physics department of the University of Oregon (especially, Ira G. Nolt and Jim Radostitz) provided their expertise regarding cryogenic refrigerators and detectors. From time to time, the various members of the QMC group also contributed. In spite of all this talent, the bolometer really never performed well. No one understood exactly why it did not. The problem may have been microphonics within the detector or problems coupling the bolometer to the telescope. Darrel Emerson[62] told me that one of his first executive decisions as site manager was to kill this project. It was soaking up too many resources of the small Tucson staff without producing astronomically useful results. I estimate that the NRAO spent well over $1M on the project. Humbling for us, there were other observatories that had successfully built millimeter- and submillimeter-wave bolometers. For example, Ernst Kreysa of the MPIfR routinely produced bolometers of incomparable sensitivity for IRAM’s much larger 30-m millimeter-wave telescope. Was he brighter than the NRAO crew? Capable as he was, I suspect that bolometerbuilding in those years was as much about technique (I’m reluctant to say “art”) as technology. Furthermore, the NRAO design worked very well on the UKIRT infra-red telescope on Mauna Kea. Therefore, our design was intrinsically OK. The 36-ft telescope and, later, the 12-m replacement, had to be the culprit.

9.4.3

The Hybrid Spectrometer

Finally, there was the hybrid-spectrometer project. In my opinion, this was doomed from the start, which I told Sandy Weinreb at the outset. Spectrometers are devices to measure spectral lines. Traditionally, they consisted of an array of narrowband filters, each tuned to a slightly different frequency and arranged contiguously like the tines of a comb. Plotting the voltage output of each filter against frequency produced a spectrum of the signal applied to the filter bank. The problem with these simple, analogue devices was the amplification of each filter could vary with time, such that the resulting spectrum was not always a reliable representation of

104

CHAPTER 9. MILLIMETER-WAVE ELECTRONICS

the frequency distribution of the input signal. One had to calibrate them often. The digital autocorrelator produces a spectrum by means of a different technique. Because frequency and time are directly related, analysis of the amplitude variations of the incoming signal as a function of time can also produce a frequency spectrum. This technique consists of restricting the bandwidth of the incoming signal by passing it through a filter and converting the now band-limited signal to a stream of digital 1’s and 0’s—a process known as “sampling.” Correlating the stream against itself (autocorrelating) produces an “autocorrelation function” that can be converted into an ordinary frequency spectrum through a mathematical process called Fourier Transformation. Radio astronomers usually credit Sandy Weinreb as the father of this type of spectrometer digital autocorrelator. Actually, Dick Goldstein simultaneously developed a digital autocorrelator at JPL. But Sandy’s design became better known to radio astronomers because, after building one, he used it to search for an astrophysically important but elusive spectral line as part of his doctoral research [68]. This line was the 327 MHz (λ92 cm) spinflip line of cosmic deuterium, 2 H, a naturally occurring isotope of hydrogen (1 H) but 6,600 times less abundant. It differs from common hydrogen by containing a neutron in its nucleus in addition to the proton. The spectral line itself is the deuterium equivalent of the famous 1,421 MHz (λ21 cm) of ordinary hydrogen, proven to be a powerful tool for astronomers. The search itself was unsuccessful but the digital spectrometer performed extraordinarily well. In his observations, Weinreb demonstrated that his digital spectrometer continued to reduce peak-to-peak radio noise correctly even up to 107 seconds (115 days) of integration. In contrast, conventional, analogue resonant-filter spectrometers failed to reduce peak-to-peak noise after only 103 seconds (about 17 minutes) because of instabilities in their amplifiers. Consequently, the digital autocorrelator soon became the standard spectrometer used by astronomers working at centimeter wavelengths. The situation was vastly different at millimeter wavelengths. The autocorrelator requires sampling the incoming signal at a rate that is twice the frequency bandwidth you want to analyze. At millimeter wavelengths, spectral lines are at least ten times wider than at centimeter wavelengths, and the digital sampler has to work at least ten times faster. This required a sampler working at 0.5 to 1 GHz—sampling speeds technically impossible in the 1970s and 1980s. For millimeter-wave astronomy, Weinreb cleverly proposed construction of a hybrid of an analogue filter bank and a digital autocorrelator. First the

9.4. FAILURES

105

analogue filters would break down the spectrum into eight contiguous broad frequency bands, each of which would then be processed by an autocorrelator. The output of this “hybrid spectrometer” would consist of eight digital spectra plotted next to each other. This spectrometer never functioned adequately—unfortunately, just as I had predicted. The individual analogue spectral sections would float up and down with respect to each other, a fault called “platforming” by the NRAO staff. Moreover, the baseline slopes of each segment were usually different. The overall result was a spectral baseline that was impossible for astronomers to interpret. Consequently, most astronomers at the 12-m telescope retreated to the old, reliable analogue filter banks. I estimate that the NRAO spent over $1M in parts and manpower trying to make the hybrid spectrometer work. By the mid-90s, extremely fast samplers became available, the generally useless hybrid spectrometer was replaced by a fully digital autocorrelator, and the spectral sensitivity of the 12-m telescope accordingly increased manyfold.

Chapter 10

Quantifying mm-Wave Astronomy In the centimeter-wavelength range (300 MHz to 30 GHz) of traditional radio astronomy, calibration of intensity is more or less straightforward. The transmission of the terrestrial atmosphere plays only a small role in the intensities of incoming cosmic radio waves. Absolute calibration at these wavelengths results from inserting objects of known temperatures in front of the feed of the radio receiver installed on the radio telescope. Typically, these are absorber materials at room temperature (≈ 300K) and at the temperature of liquid nitrogen (≈ 77K). These calibrators then establish the intensity units in terms of the effective temperature of the antenna, called “antenna temperature” or TA and quoted in units of Kelvins. Think of a radio telescope as a large, focused thermometer. Knowing the effective collecting area of the telescope, the astronomer converts these temperature units into the standard physical units of brightness or flux density. The situation is very different at millimeter wavelengths. Unlike centimeter wavelengths, the terrestrial atmosphere is a powerful attenuator of radio waves. The principal absorbers are molecular oxygen and water vapor, giving atmospheric transmission curves like those shown in Figure 1.4, which is why these telescopes are usually located at high, dry sites. The absorption varies with path length through the atmosphere, that is, with the elevation angle of the astronomical target with respect to the telescope. Observing at a lower elevation means a longer path length through the atmosphere, more extinction of the incoming astronomical radiation, and correspondingly a higher noise background because of the radiation emitted from the increased atmosphere along the path. In fact, millimeter-wave telescopes use 107

108

CHAPTER 10. QUANTIFYING MM-WAVE ASTRONOMY

this increase in background noise to estimate and correct for the atmospheric extinction. Furthermore, most large millimeter-wave telescopes weren’t very stable mechanically in the 1970s. As they changed elevation angle, gravity distorted their parabolic surfaces. This distortion changed their efficiencies as radiative collectors. The distortion varied from one telescope to another, making it difficult to compare observations quantitatively. The new observations were now good enough to allow astrophysical calculations regarding stellar evolution and quantitative models of interstellar chemistry. In the early years of millimeter-wave astronomy, accurate intensity calibration was not a problem. Astronomers’ expectations were low. They were mainly interested in whether or not they could detect something. But as the field grew and the sensitivities of the receivers improved, their expectations changed. Measuring intensities accurately of both continuum sources and spectral lines became important. What was needed was a reliable standard of brightness outside the atmosphere—standard “candles” that could be observed from time to time to provide an absolute intensity calibration independent of any particular millimeter-wave radio telescope. Cleverly, Bobby Ulich and Bob Haas solved this important problem [69]. They carefully observed a number of bright planets and spectral line objects with the small but highly stable millimeter-wave telescope of the MWO operated by the University of Texas in the Davis Mountains of Texas. Analyzing the details of how millimeter-wave signals pass through the atmosphere and through the telescope optics and electronics, they produced a system for absolute calibration that allowed observations made by different millimeterwave telescopes to be compared quantitatively in terms of a standardized unit of TA∗ —the asterisk is important here. Using this system, they showed that independent observations made with the 36-ft telescope agreed perfectly with those made by the MWO. Subsequently, Marc Kutner and Bobby Ulich [70] refined this technique by recommending a more useful intensity unit they called TR∗ . Much later, I, Jaap Baars, and John Cocke [71] showed how to convert these now standardized but still peculiar observational units into the ordinary radio astronomical units of TA used at centimeter wavelengths and into the physical units of ergs s−1 cm−2 for any of the principal millimeter-wave radio telescopes. And to gild this astrophysical lily, we also discussed the physical interpretation of internal radial velocities detected for nearby astronomical gas clouds as well as for the extremely distant objects that were subject to the laws of General Relativity.

109 Apart from the paramount astronomical discoveries themselves, I think these calibration papers, as a group, were one of the most important contributions of the NRAO’s Tucson Operations. Millimeter-wave astronomy is now on a sound quantitative footing because of them.

Chapter 11

Scheduling In a national observatory open to everyone on the basis of merit, perhaps the most important obligation is scheduling the telescopes. This procedure is always complicated. Technical equipment often changes unexpectedly from week to week or month to month. The most sensitive equipment such as receivers involves unproven prototypes. In fact, when a receiver becomes reliable, it is usually no longer at the cutting edge of sensitivity and must be replaced—with another unproven prototype. For a telescope sensitive to sunlight like the 36-ft telescope, some observing programs must be done during seasons when the target astronomical objects transit at night.

11.1

The Initial Schedulers

In the early years, beginning in 1967, Dave Heeschen personally scheduled the 36-ft telescope from Charlottesville. An experimental facility, the telescope could be assigned to a small number of astronomers with rather large blocks of time. With the advent of molecular astronomy, scheduling became a much more complex process. Several receivers were available, the number of scientific proposals greatly increased, and the time required to assess and dovetail individual programs became too large for the NRAO director. Soon after the 36-ft telescope became operational, Bill Howard (Assistant to the Director) in Charlottesville began scheduling the 36-ft telescope in addition to his usual scheduling of the telescopes at Green Bank. 111

CHAPTER 11. SCHEDULING

112

11.2

Local Control

By 1974, the scheduling demands for Tucson became even more complicated. The number of receivers had increased further. Since every receiver was a prototype with inherently low reliability, scheduling agility was required to fill in a suitable program whenever a receiver failed. Consequently, I took over the scheduling because, being in Tucson, I could often anticipate the prognosis and status of our one-of-a-kind receivers. The large number of diverse proposals made this a big job, At my request, Mike Hollis wrote FORTRAN programs into which I could enter the proposed programs and their ratings, filed by frequency. When a receiver failed, it was then easy to sift the proposals quickly by frequency to find the most appropriate one suited to whatever receiver was available. In 1990 after I returned from a sabbatical year in Bonn, I wrote a much more sophisticated system (PropMgr) for tracking proposals and referee reports for the 12-m telescope. The language was dBASE 5. Barry Clark had written a similar system for the VLA in the same language. After several years, scheduler Phil Jewell had to abandon PropMgr when our capable but independent-minded secretary, Nancy Clarke, refused to continue entering information into it. She was unhappy that she was unable to make changes in the dBASE code herself. In life, it’s often the clerical staff who determine whether or not procedures are successfully implemented. Ranking the proposals proved difficult. Most proposals experienced a considerable range of referee ratings. They involved different astronomical objects, observing techniques, and scientific objects. Some proposals were written well; some, poorly. Yet, apart from the writing skills, the scientific essence must be identified and evaluated. The choice is often between oranges, bananas, apples, and kiwi fruit. It was easy to identify the best and the worst but most fell in between. Choosing among them unavoidably involved personal judgement. Sometimes, a referee did not understand the value of the proposal. Initially, we had four referees who saw all proposals. Later, we expanded the number to five to minimize the chances of a split verdict. These referees included spectroscopists as well as continuum observers. Expecting a continuum expert to understand the nuances of a spectroscopic proposal was unrealistic—and vice versa. I always thought the scheduling system worked well until Spring, 1978. That year, Lew Snyder (Illinois) wrote to the chairman of the AUI-appointed Visiting Committee, Marshall Cohen (Caltech), arguing that I was scheduling the 36-ft telescope irresponsibly. Lew felt that the most productive use

11.2. LOCAL CONTROL

113

of the telescope was to search for molecular lines and that all other programs (continuum observations of quasars or of planets, exploration of gaseous nebula, mapping the structure of the Milky Way, and the like) should neither be considered nor scheduled. To present our case, Dave Heeschen and I prepared all sorts of histograms to describe the range of referee ratings, the range of research topics, and the result of the scheduling process. Our defense was persuasive. However, just prior to the meeting, Marshall Cohen took me to lunch and told me that this was the first time anyone had questioned the NRAO objectivity in selecting proposals and that I should be ashamed that I had been involved in this scandal. I was devastated—as well as completely innocent of any wrong-doing. The data showed that the NRAO (in this case, I) was doing well with a difficult situation. Figure 11.1 gives the dispersion of the ratings. The highest possible rating was 4. Of 148 proposals, the five referees agreed on only 25. For 50 of them, the dispersion was about 0.4. For some, the referees disagreed considerably. Figure 11.2 shows those selected from this group. Note that this was no rigid threshhold. Because some judgment was required because of the dispersion in the ratings, some proposals with the same rating (2.0) were selected while some were not. Another essential aspect of the scheduling problem was the fraction of time on the 36-ft telescope awarded to the NRAO staff members. Were we (I, actually) awarding too much time to the NRAO staff members? Table 11.1 shows the breakdown of scheduled time for proposals scheduled from March 1977 through February 1978—one year. The data set includes 123 visitors and students from 45 institutions. Most of these students worked with faculty. A few worked with NRAO staff members. Table 11.1: Distribution of Time Scheduled on the 36-ft Telescope Group Visitors Students Research Associates Permanent NRAO Staff

Percentage Usage 73 12 01 14

Nonetheless, the allegation that the NRAO wasn’t scheduling a telescope effectively was a serious one. It had to be soundly refuted by the community

114

CHAPTER 11. SCHEDULING

Figure 11.1: The histogram shows the dispersion of the average rating of referees for proposals submitted to the 36-ft telescope for the first half of 1978. The largest possible rating was 4.

11.2. LOCAL CONTROL

115

Figure 11.2: The dotted region of the histogram indicates the proposals selected for scheduling. Note that there is not a rigid mathematical threshhold for selected proposals.

116

CHAPTER 11. SCHEDULING

as well. To do this, after the Visiting Committee had met, Dave Heeschen mailed [72] this information in a letter to all current users of the 36-ft telescope, asking for their comments and their suggestions. Many replies came back, including one from Charles Townes, a Nobel Prize winner in physics who had changed his research interest to radio astronomy. All the letters stated that they found the competitive situation at the 36-ft daunting but felt the NRAO was doing a good job in a difficult situation. No one agreed with the charge made by Lew Snyder. To ward off future complaints, Heeschen’s solution was to ask Dave Hogg, then associate director of the NRAO for Operations, to review my selected proposals before I scheduled them. This procedure worked well and continued to the day the NRAO closed the telescope (by then, the 12-m telescope).

11.3

Paranoia and the Law of the Jungle

From the day molecular lines from interstellar objects were discovered, the astronomers became extremely competitive—at the 36-ft telescope, at least. While this behavior is not unknown in the research world, it seemed to be the general rule in the millimeter-wave astronomy community. That many of these astronomers used the same instrument exacerbated the problem. For example, Lew Snyder and colleagues regularly used a “burn bag” for every scrap of paper they generated at the telescope. Their fear was that a frequency or chemical formula might fall into the hands of competitors, thereby allowing competitors to search for the same molecules. As far as I know, Lew and all of his students still use the same strategy. Another incident involved “poaching.” Phil Solomon (SUNY at Stony Brook) had been scheduled on the 36-ft to investigate the structure of specific gaseous nebulae with the CO emission line, known as proposal S110, which was co-authored with Nick Scoville and Bob de Zafra. On Friday, February 7, 1975, he telephoned me from the telescope to warn me that he was changing his observing program to search for CO emission from nearby galaxies. When I reminded him that NRAO policy required him to stick to his original program, he said he was going to observe the galaxies anyway. I then told him that he should not do this because a group preceding him and following him had been scheduled explicitly to search for CO emission from nearby galaxies (proposal R65 by Lee Rickard, Pat Palmer, and Mark Morris) and, hence, he would be poaching on a refereed, already-scheduled program. He observed anyway, not only poaching, but also writing a draft paper at the telescope announcing his detection of CO in other galaxies.

11.3. PARANOIA AND THE LAW OF THE JUNGLE

117

Recently reviewing the observing logs, I found that Solomon had taken me in. From January 31, the very first day of his observing run and days before he telephoned me, he had been ignoring his approved program to search for CO in external galaxies. The logs falsely identify these galaxies by cover names like “S2-IR,” “Off1,” “Off2,” and others so that the NRAO telescope operators (Paul Rhodes and Werner Scharlach) would not know he was not following the source list given in his observing proposal. The coordinates listed in the log for these objects clearly identify them as the galaxies observed earlier by the Rickard group. Mark Morris had discovered CO in some of these galaxies two weeks earlier when the approved proposal R65 was first scheduled, from January 15 through 18. The lead proposer, Lee Rickard, was ill and could not travel to the telescope. Returning to his post-doctoral position at Caltech, Mark mentioned these detections to several colleagues, including one of Solomon’s observing partners for S110, Nick Scoville [73]. Evidently this news reached Phil Solomon just before he went to the 36-ft telescope to observe his approved program of Galactic molecular clouds, scheduled from January 31 through February 7. Obviously, he wanted to take the credit for these extragalactic discoveries and, hence, re-observed them under false names. He quickly submitted a paper to the Astrophysical Journal Letters, claiming first detection while knowing the claim was false. This was not only brazenly dishonest but also extremely unkind to the lead author of proposal R65, graduate student Lee Rickard, who wanted to include the detections as part of his doctoral thesis. If Solomon had any defense, it might be that he had been scheduled on the telescope much earlier to search for CO in galaxies but had been unsuccessful [73]. However, the NRAO scheduling system considered past, scheduled proposals as having been completed, a policy that all astronomers understood. Soon, each group had submitted a detection paper to the same journal. Barry Turner [74] recalled a telephone conference between Morris, Palmer, and himself to decide how to proceed. They decided that Pat Palmer would travel to Boston to meet with the editor, Alexander Dalgarno, to explain the situation. Not only had Solomon poached, but his detection paper [75] got to Prof. Dalgarno two days ahead of the “legitimate” one [76]. Dalgarno decided to ask Arno Penzias to adjudicate. By then, Nick Scoville had removed his name from the Solomon paper. Arno recommended that the Ap. J. Letters publish both papers but with the Rickard one being first. In addition, Arno required Solomon and de Zafra to include a specific reference to the Rickard et al. paper, which they did. This solution still left some frustration on both sides but seemed to be the best possible solution to a

118

CHAPTER 11. SCHEDULING

difficult problem for the journal. The incident presented a profound problem for the NRAO as well. Vested by the NSF with the responsibility for operating and scheduling radio telescopes, the NRAO was involved to some extent with how the telescopes were used. The NRAO management split on the topic. Heeschen argued that the NRAO role did not include policing how the telescopes were used, only facilitating access to them. Others argued that the NRAO had an obligation to protect the interests of scheduled observers and to punish transgressions to set an example. In the end, we agreed to ban Phil from all the NRAO telescopes for one year but that decision was not unanimous. In retrospect, I think the punishment worked. I don’t know of another incident like this one. Furthermore, Phil Solomon, a gifted scientist, continued to make substantive contributions to astronomy.

Chapter 12

Improving the Performance of the 36-ft Telescope Astronomers and staff quickly recognized that the surface accuracy of the 36-ft telescope fell significantly short of the design specifications. Being a monolithic mirror machined from welded aluminum plate, the surface presented a challenge to fix, which evidently appealed to our “can do” staff. There were several options. One was to attempt to repair the surface itself.

12.1

“Foiling” the 36-ft Telescope

During the summer of 1976 when the atmospheric moisture vitiated astronomical observing, John Payne, Mike Hollis, and John Findlay [77] measured the telescope surface with a small mechanical cart designed by John Findlay. This cart—essentially, a spherometer like those used by opticians—measured the depth between its two axles as it was hauled from the center of the dish surface along a radius to the outer edge, in effect measuring the differential curvature of the surface (over the distance between the cart axles) as a function of distance along that track. Integrating these data produces the surface figure along the cart track at that azimuth angle. Combining these surface “cuts” led to a map of the antenna surface from which the design surface was subtracted, leaving a contour map of the deviations from the target design. Figure 12.1 shows the cart. The process was messy [78]. First, the NRAO staff removed the white titanium oxide surface paint with gallons of paint remover (Jasco). The experiment needed direct access to the bare aluminum. Removing the paint splattered it all over the normally pristine concrete floor of the dome. 119

120

CHAPTER 12. IMPROVING TELESCOPE PERFORMANCE

Figure 12.1: The spherometer developed by John Findlay for measuring the surface figure. Basically, it was a device to differentiate the curvature at one azimuth angle running from the center to the lip of the reflecting surface, with a linear resolution of the wheel base (dr/d ≡ depth/wheelbase). After the measurements had been completed, the data were processed on the CDC computer at KPNO in Tucson to produce a contour map of the surface errors, which was printed onto a transparency. With the dome door closed to darken the interior, a projector at prime focus projected the contours onto the antenna surface, shown in Figure 12.2. Felt-tipped pens then marked the contours on the aluminum surface. The final stage was to fill in the surface “holes” with appropriate layers of the 0.005 in-thick adhesive aluminum foil and, in this way, to correct the deficiencies of the surface figure. This proved to be an iterative process. After the foil had been applied, radio astronomical observations were made as soon as possible to check the corrections. Invariably, the corrections proved to be less effective than expected. So, the entire process was repeated over a period of weeks—without success. Often, previously laid foil had to be removed to accommodate the latest contour maps. Despite its ingenuity, the technique was evidently spoiled by the instabilities of the telescope surface. No one realized that at the time. First, it took hours to acquire the data with the cart. During this time, the ambient temperature undoubtedly changed. Second, the cart data had to be made with the telescope in the vertical or “stowed” position—not an elevation angle normally used for astronomical observing. As the telescope changed elevation to track an actual source, the changed gravity vector distorted the

12.1. “FOILING” THE 36-FT TELESCOPE

121

Figure 12.2: A nighttime view of the 36-ft telescope showing error contours projected onto its surface. Using Sharpie marking pens, employees traced these contours and, in daylight, filled them with aluminum foil of appropriate thicknesses to correct the errors of the parabolic surface figure.

122

CHAPTER 12. IMPROVING TELESCOPE PERFORMANCE

figure of its surface. So, not only did the surface change during the measurements, but the measurements did not describe the telescope surface as it was actually used. Third, after integration, the error in the surface measurements accumulated as the cart moved from the center toward the lip of the dish. Thus the largest error involved the largest area of the telescope surface—the very opposite of what was needed.

12.2

The Teepee

In the mid-1970s, the telescope was in great demand by astronomers. Unfortunately, it could not be used if sunlight struck the surface, that is, fully used during the day for sources in directions near the Sun. The solar heating would warp the telescope surface, requiring as much as four hours to recover depending upon the amount of exposure. To maximize the astronomical “throughput,” we needed to find a way to keep sunlight off the antenna. The first attempt was to install a removable sunscreen in the slit of the astrodome. This sunscreen consisted of a thin, cord-reinforced, sandwiched plastic fabric known as Griffolyn T-55, sewn to our specifications by the Sullivan Awning Company of San Francisco. This fabric had a white exterior that reflected both sunlight and infra-red, and a black interior to reduce transmission of sunlight. Its radio frequency transmission was extremely high. As I recall, the company’s salesman was Gene Wetmore, who had previously worked at the 36-ft telescope as chief telescope operator. He understood exactly what was needed and delivered what we required. In spite of the excellent transmission characteristics, the sun screen was easily torn by wind gusts. It was just too fragile for regular use. It was time for Plan B. Plan B proved extremely successful. It consisted of installing a conical Griffolyn sunshield extending from the telescope surface to its apex, hence the name “teepee.” Figure 12.3 shows the sunshield in place. To maintain the figure of the reflector and to prevent the telescope surface from heating through a greenhouse effect, cool air was continually pumped into the cone at a low pressure. Zippered panels provided access to the receivers. There were small penalties paid for its use. First, reduced flexibility. It took several hours to install or remove the teepee. Once installed, it was usually left up for days if not weeks. Second, reduced nighttime sensitivity. While the Griffolyn fabric had high transmission at millimeter wavelengths, it was not perfectly transparent. Further, its absorption increased toward higher frequencies. Although the teepee greatly increased the usefulness of

12.2. THE TEEPEE

123

Figure 12.3: Northward-looking photograph of the 36-ft telescope encased in a Griffolyn “teepee” to prevent sunlight from striking the reflector. Also shown are the two cables stretched across the slit which maintained the slit dimensions, and the two lightning rods on top of each arch which reduced (we hoped) lightning strikes. the telescope during the day, there was a small deterioration at night because of this absorbtion, compared to night observations when the teepee was not installed. While also sensitive to the wind, the teepee was protected by being inside the astrodome. On those occasions when the wind blew directly through the open slit onto the teepee, astronomers usually selected another object in their observing list, until the dome azimuth changed so that the earlier target could be safely observed. Although wear and tear inevitably led to a finite lifetime for the teepees, they cost only about $3k each—small compared to the operating cost for the extra observing time they facilitated— and one or two new ones were

124

CHAPTER 12. IMPROVING TELESCOPE PERFORMANCE

always kept on hand.

Chapter 13

The 25-m Telescope Neither the foiling nor the teepee solved the sensitivity of the telescope adequately. It was time to consider replacing the 36-ft telescope.

13.1

What Should We Build?

In early 1974, David Heeschen organized an ad hoc committee to investigate a replacement for the 36-ft telescope. Its first chairman, carrying the title Project Scientist, was Barry Turner. Also on the committee were John Findlay, Sebastian von Hoerner, Campbell Wade, and me. The general plan was to scale the never-funded NRAO design for a 65-m, centimeterwave telescope down to 25 m, enclose it in a astrodome, and locate it on an excellent site. Structural engineering was done by Lee King, Woon-Yin Wong, and Sebastian von Hoerner. Figure 13.1 shows the design. Each of us brought something unique to the table. Barry was a leader in millimeter-wave astronomy, John Findlay had supervised the design of the 65-m telescope and had managed construction of the 36-ft telescope, Campbell Wade had done the site search and evaluation for the VLA which was then under construction, Sebastian von Hoerner was the father of a modern design for radio telescopes, and I was the site manager of the 36-ft telescope operation. From time to time, other NRAO employees contributed to the planning including Dale Webb, the Tucson business manager who knew the operating costs of the 36-ft telescope firsthand. The telescope design consisted of an “equally soft” truss to support a reflecting surface comprised of highly accurate, individual panels. This back structure would deform as it tilted such that the entire surface always maintained a paraboloidal shape that could be accommodated by repositioning 125

126

CHAPTER 13. THE 25-M TELESCOPE

Figure 13.1: The structure of the 25m, millimeter-wave telescope proposed to replaced the 36-ft telescope. A human figure has been placed on the stairway to give scale.

13.2. WHERE SHOULD WE BUILD IT?

127

the focus along the axis of symmetry—a principle known as homologous deformation [79]. A wheel-and-track with a central pintle bearing would support the entire structure. To resist additional deformation induced by wind and by solar heating, the proposed 25-m telescope would be enclosed in an astrodome like that used by the 36-ft telescope. Because the site of the 25-m telescope would probably involve high altitude and higher winds, the new astrodome would be constructed of steel rather than fabric.

13.2

Where Should We Build It?

An important question was where to site it. By reviewing maps of the western United States and making exploratory visits, the committee considered eight sites to be viable candidates on the basis of road access and proximity to cities, as listed in Table 13.1. The column labeled “Wv ” gave an estimate of atmosphere moisture in terms of precipitable water vapor on a good day; that labeled “Clear Days,” the approximate number of cloud-free days per year. Site exploration was fun. I remember well taking a helicopter from Denver to the summit of Mt. Evans, Colorado, in winter, driving to the summit of Pikes Peak after it had been closed for the season, driving up an awful access road to White Mountain near Bishop, California, with Berkeley’s Dave Cudaback, and flying to Hawaii to visit Mauna Kea on the “Big Island” – to mention just a few of these trips. The scenery was just as breath-taking as the high altitude. Each site listed in Table 13.1 had strengths and weaknesses. None was clearly superior to the others, although some were clearly inferior to the others. Mauna Kea was the closest to the equator and would give the largest declination coverage of the sky but would be expensive to build on and to operate from. Mt. Lemmon was particularly frustrating because, in spite of its wonderful attributes for observing and for operations, it was the site of powerful radio transmitters that could interfere with the operation of the radio telescope. “Could” was an important word. Even if the interference proved to be impotent at millimeter wavelengths, co-locating a radio telescope with known radio transmitters would certainly undermine the “no tolerance” position taken by radio astronomers for years in frequency-allocation committees with respect to radio interference. A site of particular interest to our committee was “White Mountain,” in the Inyo Mountain range immediately east of Bishop, California. For many

CHAPTER 13. THE 25-M TELESCOPE

128

Table 13.1: Sites Considered for the 25-m Millimeter-Wave Telescope Site Kitt Peak Mauna Kea Mt. Epaulet Mt. Hopkins Mt. Lemmon Pikes Peak South Baldy White Mt.

Nearest City Tucson, AZ Hilo, HI Denver, CO Tucson, AZ Tucson, AZ Co. Springs, CO Magdalena, NM Bishop, CA

N. Lat. 31 58 19 49 39 34 31 41 32 26 38 50 37 58 37 37

W. Long. 111 36 155 29 105 38 110 53 110 47 105 02 107 11 118 14

Elev.(ft) 6,750 13,630 13,450 8,585 9,190 13,950 10,800 13,000

Wv (mm) 3.3 ± 0.8 2.5 ± 0.2

2.3 ± 0.8

1.5 ± 0.6

Clear Days 260 230 160 240 260 160 260 260

years, the University of California at Berkeley had operated a laboratory (the Nello Pace Laboratory) for high altitude research on a 12,470-ft saddle near Mt. Barcroft, a locale we called White Mt. The research involved a wide range of topics such as the effects of altitude upon pilot judgement, the porosity of the shells of chicken eggs, cosmic-ray physics, cosmic micro-wave background emission, astronomical infra-red emission, and other topics. The site contained a large laboratory and dormitory building as well as a number of smaller storage buildings. Two aspects of any site were paramount to the new radio telescope: weather and access. And here lay the shortcomings of White Mt. Being a northerly site, White Mt. experienced storms from time to time that would impair observing. Unlike Mauna Kea, which experienced clouds at dawn every morning due to the diurnal up-welling of moist air carried by the trade winds, White Mt.’s cloudy weather could persist for days. The Mauna Kea situation appeared more tolerable because, except for a storm, a visiting observer should be able to observe some during every day—or so everyone thought at that time. So, the lower number of clear days of Mauna Kea compared to White Mt. was a misleading statistic with regard to astronomical observing. The second aspect was fast, easy access for astronomers and our support crew. At that time, Barcroft could be reached only by road or helicopter. The road trip took hours. One drove up Westgard Pass and then turned north on a poorly maintained, unpaved road that wended its way upward. The drive could easily take at least 3 hours, on a good day in the summer. It was impassable in storms or in winter. Upgrading this to an all-weather road would have been expensive but also perhaps impossible. In those days, the US Forest Service strongly pursued a policy of protecting the back country by preventing enhanced access. Their Bishop representative assured us that

13.3. PREPARING THE FORMAL PROPOSAL

129

they would oppose an upgrade to the “White Mountain Road” because some of the land was being considered for inclusion in the National Wilderness Preservation System. To supply and staff their Barcroft Laboratory, the University of California had used a helicopter for several years. For a while, they actually owned their own helicopter before deciding it was more economical to lease one when required. Frankly, I thought this access system would be too expensive and probably too risky for an NRAO telescope. On those occasions when I chartered a helicopter to get from Bishop to the summit area, the ex-Viet Nam pilots terrified me by swinging the helicopter from side to side along the canyon walls to use the “ground effect” enroute to the Barcroft Laboratory. This seemed dangerous, despite all the literature that Dave Cudaback sent me attesting to the safety of helicopters and all the telephone calls he arranged from helicopter-safety officials. On one trip, the pilot actually sported a cast on his wrist for an injury obtained during a “crash” the previous week! For me, a more attractive alternative would be an aerial tramway like those used by ski areas and small Swiss villages. Consequently, the von Rols Corporation provided a detailed design of a tramway for one of the canyons (Milner Canyon) leading from Bishop to the Barcroft Laboratory area. They estimated a construction cost of about $8M, as I recall, expensive but much less than the cost of constructing an all-weather highway, which the Forest Service would have opposed. Unfortunately, further investigation revealed that tramways were expensive to maintain and not always trouble-free [80], as the IRAM observatory was to learn tragically on July 1, 1999, when twenty people were killed enroute to their millimeter-wave interferometer on Plateau de Bure, France.

13.3

Preparing the Formal Proposal

Barry Turner drafted what we had learned into a report [81] circulated internally for comment, later to be known as “Volume I” of the proposal for the 25-m telescope. This report also contained a description of the proposed telescope, a description of radomes and astrodomes that would be required, a brief discussion of sites, and preliminary cost estimates. Work continued on this project but with an organizational change. Having dutifully served for a year, Barry Turner asked to return to research. Soon I became project manager and re-organized the committee into a form that was more comfortable for me. I arranged for an internal working com-

130

CHAPTER 13. THE 25-M TELESCOPE

mittee of John Findlay, Sebastian von Hoerner, Hein Hvatum, Buck Perry, Bobby Ulich, Campbell Wade, Dale Webb, and myself, augmented by the structural engineers Lee King and Woon-Yin Wong and an external committee of millimeter-wave astronomers to evaluate our progress from time to time. We spent the next two years refining the concept described in the earlier report. The design of both telescope and astrodome (Figure 13.2) evolved. We examined options for the surface plates. We prepared a PERT criticalpath plan for construction, and detailed estimates for the costs of construction and of operation. To set the surface initially, John Findlay developed a spherometer appropriate to measuring the surface of the 25-m telescope. Opticians use these routinely to measure the curvature of spectacle lenses. Findlay’s was more sophisticated. It determined the curvature of the telescope surface along any given azimuth angle of the surface. This instrument measured the differential along a path over the surface which, when integrated, gave the shape of that path. Figure 12.1 shows the device. Figure 13.3 shows two of the carts on the Green Bank test track. Most importantly, we greatly expanded our search for telescope sites and decided that Mauna Kea would be the best of all those we investigated. Figures 13.5 and 13.6 give maps of the island; Figs. 13.7 and 13.8, the details of the road from the saddle to the summit; and Figure 13.9, the topography and telescope locations as of 1977. Its low latitude would allow access to the largest range of astronomical objects. Our recommendations appeared in July, 1977, in a second report [82] that AUI submitted as a formal proposal to the NSF for funding. Figure 13.10 shows a composite panoramic view of the Mauna Kea summit, looking north. Figure 13.4 shows a drawing of how the telescope would have appeared.

13.4

Negotiating for a Mauna Kea Site

Negotiations with the University of Hawaii were interesting. First, the summit of Mauna Kea lay within a scientific preserve established by the Hawaii legislature to benefit the people of Hawaii. One of my instructors in graduate school, John Jefferies, directed the Institute for Astronomy and represented the interests of the University of Hawaii. He was responsible for the creation of Mauna Kea as an international observatory. The precedent was that every telescope built within the preserve allot 10% of its observing time to the Institute for Astronomy, thereby fulfilling the requirement of benefitting the

13.4. NEGOTIATING FOR A MAUNA KEA SITE

131

Figure 13.2: The metal astrodome ultimately proposed for the 25-m millimeter-wave telescope. Lee King designed it. The figure at the lower right gives the scale.

132

CHAPTER 13. THE 25-M TELESCOPE

Figure 13.3: Two spherometers on the test track in the basement of the Jansky Laboratory in Green Bank. Shown is mechanical engineer John Ralston, who actually built the carts and the track. NRAO photo.

13.4. NEGOTIATING FOR A MAUNA KEA SITE

133

Figure 13.4: An artist’s rendering of the proposed 25-m telescope in place in “Millimeter-Wave Valley” on the summit of Mauna Kea, Hawaii. This site is next to where the James Clerk Maxwell Telescope is now situated. NRAO photo. people of Hawaii. Being a national observatory, the NRAO really could not cede time without peer review. After fruitless discussions, the AUI vice president, Carl Amthor, flew to Hawaii to negotiate a deal personally. It was then decided that the AUI/NRAO contribution would be to pay for a buried power line between the island power grid and the summit of Mauna Kea. The estimate for this construction was $2M. In addition, the NRAO would make annual payments to support the infrastructure of Mauna Kea observatories, carried out by an organization known as the Mauna Kea Support Organization. The operations plan was, essentially, to copy the infra-structure of the 36-ft telescope in Arizona to Hawaii. The low-altitude base would be located in an astronomy center in Waimea, Hawaii, near the operations center of the historic Parker Ranch, which would be adjacent to the support center for the Canada-France-Hawaii optical telescope. A microwave link would connect this base office with the 25-m telescope.

134

CHAPTER 13. THE 25-M TELESCOPE

Figure 13.5: The island of Hawaii, showing the location of its five volcanoes. The “saddle road” runs westward from Hilo, across the volcanic saddle between Mauna Kea and Mauna Loa. A site,“millimeter-wave valley,” a few hundred meters west of the Mauna Kea summit, was selected for the 25-m telescope. Hale Pohaku is the 9,500-ft location of the “mid-level facility” of dormitory rooms used to house astronomers. The village of Waimea, just south of the volcano Kohala, was planned to be the operations base. Since this map was drawn, Kiluea Crater has erupted into a major active volcano. NRAO drawing.

13.4. NEGOTIATING FOR A MAUNA KEA SITE

135

Figure 13.6: A three-dimensional topographic map of Mauna Kea. The town of Hilo is to the far right. The upper left arrow marks the location of Waimea, our proposed base operations; the other arrow, the summit of Mauna Kea. The “saddle road” runs more or less east to west from Hilo. Note the lines marking the extensive drainage stream beds to the northeast, indicating the large amount of precipitation deposited on that quadrant of Mauna Kea by the trade winds. In contrast, the southwest region of Mauna Kea is arid because it lies in the rain shadow. Author photo.

136

CHAPTER 13. THE 25-M TELESCOPE

the slope increases more or less regularly en route to the summit. Hale Pohaku is the location of the housing facility for astronomers. Numbers indicate altitude in feet above sea level. Hein Hvatum and I had planned to ride our bicycles from Hilo to the summit, but the opportunity never appeared. NRAO drawing.

13.4. NEGOTIATING FOR A MAUNA KEA SITE

137

Figure 13.8: An aerial photograph of the access road to the summit of Mauna Kea in 1977, looking northward. The old dormitory complex at Hale Pohaku may be seen at the lower right. Bobby Ulich photo.

Throughout the planning process, additional negotiations were necessary with citizens of the Island of Hawaii. The county mayor, Herbert Matayoshi, strongly opposed construction of more telescopes on Mauna Kea for aesthetic reasons, often referring to them as pimples on the face of a beautiful mountain. At one meeting in his office, he became so angry at the prospect of the NRAO building a telescope that he swivelled his office chair around to face the wall behind him, indicating that the meeting was over. Mayor Matayoshi’s opposition never wavered. The minutes of an April 1980 meeting of the Mauna Kea Users’ Committee in Tucson state, “In anticipation of this [NSF construction money], Dr. Jerry Tape, then president of Associated Universities, Inc., and Dr. Mort Roberts visited Hawaii to

138

CHAPTER 13. THE 25-M TELESCOPE

Figure 13.9: Topographic map of the summit of Mauna Kea in 1977. Marked are the 61-cm Air Force telescope, the 61-cm Planetary Telescope of the University of Hawaii, the 3.6-m Canada-France-Hawaii Telescope (CFHT), the 3-m NASA Infra-Red Telescope (NIRT), the 3.8-m United Kingdom Infra-Red Telescope (UKIRT), and the site for the proposed 25-m millimeterwave telescope of the NRAO in “millimeter-wave valley.” The bold lines indicate roads existing at that time. Elevation contour intervals are 40 feet. Institute for Astronomy drawing.

13.4. NEGOTIATING FOR A MAUNA KEA SITE

139

Figure 13.10: Panoramic view of the summit of Mauna Kea in 1977, looking north. The Toyota “jeep” near the road branch shows the scale. Optical telescopes were located on the large cinder cone at the right center. See Figure 13.9 for the exact locations. The site proposed for the 25-m telescope is at the left base of the cinder cones in the lava flow known as “MillimeterWave Valley.” Composite photo by Bobby Ulich. meet with various individuals in the State and County government to introduce the [25-m telescope] project, where they were greeted graciously and positively by all they met except the Mayor of the Big Island.” [83] The Hawaii Audubon Society opposed increased activity on Mauna Kea because of risk to the Palila bird habitat. This endangered bird was the last remaining finch-billed Hawaiian honeycreeper to exist on the island of Hawaii. It fed only on the m¨ amane tree that grew on Mauna Kea between 6,500 and 10,000 ft altitudes. Unbeknownst to me was that over 50% of bird species endemic to Hawaii had become extinct because of changes to critical habitat by early settlers and competition from foreign bird species introduced to control insects. Mae Mull was the Audubon spokesperson for the Big Island. After several years of meeting with her, I became a convert to her point of view. But it was too late for remedies other than minimizing traffic up and down the mountain. Hawaiian sportsmen (hunters) also opposed construction of telescopes on Mauna Kea. They wanted to hunt feral deer and pigs on the mountain and realized that telescope activities would curtail these activities. They and the Audubon Society joined forces to oppose us—strange bedfellows, I thought. Supporting us was the business community on Hawaii. Particularly in Hilo, they wanted to encourage new business activity. Telescopes would surely bring that to them, especially if the support facilities were to be located in Hilo. I gave a talk to the Hilo Rotary Club, which expressed support for our telescope, and attended at least one island Town Meeting. Through many meetings, we were able to smooth the way for the 25-m

140

CHAPTER 13. THE 25-M TELESCOPE

telescope and looked forward to its construction. Officials from the University of Hawaii were always helpful. President Fujio Matsuda, a civil engineer by training, met with us several times. Vice President Harold S. Masumoto, an attorney, always had his door open to us, as did the representative of the University of Hawaii at Hilo, Mary Matayoshi. John Jefferies and his assistant, Ginger Plasch, did everything possible to help us solidify our plan for construction and operations.

13.5

The Funding Process

The project was ready for funding. Hein Hvatum and Gilbert “Buck” Perry prepared a PERT analysis for construction. AUI submitted the formal proposal to the NSF in 1977. It should have been an excellent time for this proposal. Bill Howard had arrived at the NSF in January 1977 to head the Astronomy Division. He was looking for a large, ready-to-go “Centers” project to push. The 25-m telescope was just what he wanted. But now the road toward funding became rocky. The nearer the project came to funding, the more the opposition hardened. In retrospect, this reaction should have been expected but I was much too naive to have predicted this. First, university support was ambivalent if not actually schizophrenic. When their delegates attended review meetings of the 25-meter project, they always left enthusiastic. Arriving home, they again focused on their own difficulties of getting NSF money—particularly if their proposal competed with one submitted by the NRAO. In the case of Berkeley, Caltech, Columbia, and the University of Massachusetts, this competition was especially direct. Each of these institutions had a millimeter-wave telescope of their own and, understandably, were interested in the continued vitality of their own instruments. The NRAO project competed directly with this goal and with the universities’ share of the limited NSF operating funds. Second, another group existed within the radio astronomy community with their own agenda—one that did not include a new millimeter-wave telescope. US astronomers working with Very Long Baseline Interferometry (VLBI) had been observing with an ad hoc assembly of radio telescopes for many years. This group wanted a dedicated array of telescopes available for their observations on a fulltime basis, the Very Long Baseline Array (VLBA). Furthermore, many of this group did not understand molecular spectroscopy and did not relate to its successes. I remember well overhear-

13.5. THE FUNDING PROCESS

141

ing a conversation in the parking lot of a Socorro motel between Heeschen, Ken Kellermann, a senior VLBI scientist on the NRAO staff, and Marshall Cohen, a Caltech professor who worked exclusively with VLBI. Kellermann was saying something like, “Well, Dave, the 25m project is interesting and all that but we both know the science is inferior to that coming out of VLBI.” Marshall Cohen agreed. Knowing Dave Heeschen’s broad understanding of radio astronomy, I’m not certain that he shared those extremely narrow sentiments and he certainly did not endorse the comment then. This conversation illustrates that some of the VLBI community were lukewarm, at best, toward a new millimeter-wave telescope. They were concerned about the competition with the prospects for funding the VLBA. Meanwhile, we had one indicator that the NSF supported the 25-m telescope. Bill Howard telephoned that he had funds to support half of the project but could not guarantee that the next year’s budget could provide the second half of those funds. Absolutely certain that the NSF would surely fund this important project, we elected to wait until all of the funds could be found. In retrospect, that was a big mistake. Third, in the spring of 1979 a last-minute competitor appeared. Frank Drake, then director of the National Astronomy and Ionosphere Center (NAIC) in Arecibo, Puerto Rico, and, Ed Grayzeck, a faculty member of the University of Nevada at Las Vegas presented a new idea. They proposed that the NSF (with the help of the Fleischman Foundation) support construction of a 35-m spherical millimeter-wave telescope (see Figure 13.11) designed like the Arecibo Telescope. The proposed site was Charleston Peak near Las Vegas, with a ski resort and a paved road to the top. Furthermore, Drake claimed that the spherical design would be as effective as the NRAO design but much cheaper to build and operate. To investigate manufacturing methods, the NAIC actually built a prototype 60-ft dish near Ithaca, New York, by cutting Styrofoam blocks with a cutter at the end of a long boom [6]. The plan included enclosing the smaller millimeter-wave version in an inexpensive radome. This challenge had to be met head on. The NRAO design was expensive, and its technical advantages had to be shown to be worth its extra cost. As we saw the situation, the NSF had to see solid support for one design or the other to place either proposal in a queue for funding. Bill Howard told me later that he did not regard the Drake proposal as a significant threat. At the time, we did not know that, however. [84]. To choose, the NSF organized a subcommittee of its standing Advisory Committee for Astronomical Sciences to hear presentations of both projects. This committee included prominent astronomers: Alan Barrett (Chair), Bob

142

CHAPTER 13. THE 25-M TELESCOPE

Figure 13.11: Artist’s sketch of the fixed spherical, millimeter-wave reflector proposed by Frank Drake and Ed Grayzeck in 1979 for Charleston Peak near Las Vegas, Nevada. The 35-m dish would be enclosed within a radome that is not shown. Like the NAIC telescope in Arecibo, the radio beam would be “pointed” by moving a feed. The figure at the bottom illustrates the scale. Sketch from Frank Drake. Gehrz, Carl Heiles, Dick Huguenin, Don Johnson, Bob Leighton, Al Moffet, Don Osterbrock, John Ruze, Phil Solomon, Lew Snyder, Pat Thaddeus, Charles Townes (Nobel Laureate), Paul Vanden Bout, and Bob Wilson (Nobel Laureate). The charges to the committee were: “(1) From . . . deliberations on possible advantages and disadvantages of both the 25-meter fully steerable telescope and the 35-meter fixed spherical telescope, how should the NSF proceed with its funding plans for millimeter wave astronomy in the near future?” and “(2) In the event one of these two . . . telescopes will soon be funded, at what site should it be located?”

13.5. THE FUNDING PROCESS

143

The committee met in Washington, D.C., on July 16 and 17 in 1979. Also present were Mort Roberts (the NRAO director at the time), Bob Hughes (the AUI president), Sebastian von Hoerner, Ted Riffe (the NRAO business manager), and sundry NSF staff members. A coin was flipped and I was selected to present first. Frank Drake went second. I don’t recall ever being so nervous. Not only was I new to presentations to national committees but Drake is a gifted speaker. Mort had told me earlier that the NRAO proposal was inseparable from the Mauna Kea site. He felt that a national observatory had no business in building less than a cutting edge instrument. The low-latitude Mauna Kea site was essential to ensuring that a wide range of declination would be available. Accordingly, he instructed all of us not to mention alternative sites under any circumstances. However, the Mauna Kea site added a huge cost to the project. In their deliberations, the committee was aware of this and summoned me back to their room to speak about other sites. One reason that the topic arose was that George Rieke, then deputy director of Steward Observatory, insisted on attending this meeting to persuade the committee that the NRAO build the 25-telescope on Mt. Bigelow rather than on Mauna Kea. Despite Mort’s injunction, I really had no choice but, technically, I had let him down. In October, the committee’s report [85] appeared. The committee was “unanimous in recommending without reservation that the NSF fund immediately the 25-meter millimeter wave telescope, as proposed by the NRAO. The NRAO design is an excellent response to the needs of millimeter wavelength astronomy; it has been thoroughly studied, and realistic plans exist for the construction, management, operations, and maintenance of the instrument and facility. The NRAO is the best qualified institution in the country to carry out this program, with a long history of effective operation of comparable national facilities and highly-qualified scientific, engineering, and technical staff.” and “The committee considered at great length the choice of site for the telescope. This choice quickly narrowed to one between Mauna Kea, Hawaii, and a site in the Santa Catalina Mountains near Tucson, Arizona. . . . Mauna Kea appears to be, by a considerable margin, the best site in the United States for the instrument . . . The Committee feels that if the increased costs

144

CHAPTER 13. THE 25-M TELESCOPE . . . on Mauna Kea were to jeopardize the entire project, then the same instrument should be funded by the NSF for . . . a site in the Santa Catalina Mountains of Arizona. Therefore, the Committee strongly recommends the NSF fund immediately the construction and operation of the 25-meter telescope on Mauna Kea.

Another hurdle soon appeared. In 1979, the Decade Review of Astronomy for the 1980s was being prepared [86]. This report would rank and recommend new starts for astronomy for the 1980s. It was important that the 25-m telescope receive the highest rank if the NSF brass were to support it for funding. As was customary then, drafts of the report were exchanged regularly between the review chairman, George Field, and Bill Howard. In one of these drafts—possibly, the 5th or 6th—Bill Howard noted that laudatory adjectives for the 25-m telescope had disappeared. Alarmed, Bill telephoned Mort and asked him to call Field and reaffirm the NRAO’s support for the telescope [84]. Bill suspected this telephone call was never made, because the critical adjectives were never restored. Bill suspected that Mort was now waffling between supporting the 25-m telescope and the newly proposed VLBA telescope. The VLBA community had certainly been pressuring Mort for support. Not wanting to ram the 25-m telescope down the community’s throat, Bill began to wonder if the earlier strong support for the 25-m telescope had now waned. Not all of the millimeter-wave astronomy committee were negative regarding our prospects for funding. I remember a visit of Bob Leighton and Gerry Neugebauer to Tucson in 1978 to meet with Dave Heeschen and me regarding support of their mm-wave telescope planned for Mauna Kea. I don’t remember the date, however. Not wanting Caltech to get into administration of another telescope, they suggested that the NRAO manage their telescope as well, reserving half of the observing time for Caltech staff. Clearly, they expected the NSF to fund the 25-m telescope. In January, 1980, the NSF had requested $1.7M to start construction, and the 1981 budget proposed by President Carter included this request. But by early March, the administration had removed $70M from the NSF request for FY1980, and their Astronomy Division decided to postpone the startup money for the 25-m telescope [83]. By the end of 1981, the NSF not only had approved the project but the Presidential budget request for FY1982 again asked Congress for funding to build the 25-telescope on Mauna Kea. What a victory! Right? But no. It was only one battle in the war;

13.5. THE FUNDING PROCESS

145

it was a false summit on the route to the top. This proved to be the last budget request of Jimmy Carter. Inflation was running high, as were interest rates. The US economy was not particularly sound and Ronald Reagan soundly defeated Carter for the presidency. In November, 1981, the incoming administration tore up the Carter budget request and postponed all new starts. The 25-m telescope was one of the casualties. Had the war been lost? Not quite but losses were mounting. The important Field report actually appeared in 1982. In a presumably well-intentioned move, the decadal review committee decided to consider the 25-m project as a “done deal” and not to rank it in their listing. Perhaps they were concerned about delaying the aging telescope project further because it had been submitted for funding in 1977. The proposal was old. Having the Field Committee consider the project would add another year of delay. For whatever reason, the report omitted the telescope from its formal rankings of new projects although the verbiage in the report repeatedly stated the committee considered the project to be an excellent one and assumed the telescope to be under construction. The result was that the opposition to the telescope stressed that the Field committee had not ranked the project and, hence, it should no longer be considered a viable proposal. The topranked projects were the VLBA and the New Technology Telescope (NTT) of KPNO. Seven years had passed since Volume I of the 25-m Telescope Report had first appeared, and five years since AUI had submitted the proposal to the NSF for funding. Finally, the optical astronomers eventually found their chance to axe the project for good. In a meeting of the Astronomy Advisory Committee (AAC) on April 5 and 6, 1982, the members debated the 25-m telescope project. By this time, I had resigned from the project, and the NRAO was represented by Mort Roberts, Hein Hvatum, Harvey Liszt, and Barry Turner. At that time, the AAC consisted of Jacques Beckers (chair), Eric Becklin, Bernie Burke, Riccardo Giacconi, Fred Gillett, Dave Hogg, Roberta Humphreys, Dick McCray, Don Osterbrock, P. Pesch, and Joe Taylor. After considerable discussion, Osterbrock moved that the AAC vote to support the 25-m telescope project for Mauna Kea. The resolution failed more or less along optical/radio lines, with only Burke, Hogg, and Osterbrock voting for it. Becklin abstained. Subsequently, Beckers asked Giacconi to draft a resolution to support concurrently the VLBA and the NTT, a motion that was passed by acclamation. It’s ironic that, many years later as president of AUI, Riccardo Giacconi found himself in charge of the ultimate replacement for the NRAO 36-ft telescope, the very expensive and politically complicated ALMA array. God

146

CHAPTER 13. THE 25-M TELESCOPE

is just. It was surely punishment for his earlier misjudgment. The 25-m telescope had been killed. The AAC had selected what was to be an unexpectedly expensive telescope (VLBA), with a smaller collecting area compared to the older composite VLBI array, to be the next major initiative for US radio astronomy. For years afterward, I believed it was a bad decision. However, at this writing, the VLBA has made exciting discoveries and undoubtedly will continue to do so. As intended, it is much easier to use than the old composite array despite its lower sensitivity. The question in 1982 for millimeter-wave astronomers was what to do now? After the AAC vote, Bob Wilson invited millimeter-wave astronomers to meet in New Jersey for a post-mortem. NRAO staff members were excluded so the community would be free to criticize the NRAO if it wanted. In early 1983, in an almost parallel effort, the NSF organized a sub-committee to the AAC, with Charlie Lada, Pat Palmer, Lew Snyder, Jack Welch, and chaired by Alan Barrett, to advise them what the NSF should do next. I recommended that the NSF build the 25-m telescope within a cloth astrodome and with pressed ESSCO aluminum panels in the Santa Catalina range near Tucson for about $9M, an option specifically recommended by the earlier Barrett committee in 1979 if there were difficulties funding the 25-m telescope for Mauna Kea—which there certainly had been. The lower altitude site in Tucson not only had a well-developed infra-structure including an existing dormitory but its milder climate would allow less expensive choices for the design of the telescope. Alas, it was not to be. The winds of astronomical interest had changed direction by this time. The committee had no interest in a single-dish telescope for millimeter waves. The official report [87] appeared in early 1983 with a recommendation to design and build “a national large millimeterwave array that would have a resolution of 1 at 115 GHz (λ2.6 mm), a collecting area of 1,000-2,000 m2 , and good imaging at 300 GHz (λ1 mm).” This report stimulated the NRAO to design the Millimeter Wave Array (MMA) that, ultimately, became ALMA, now being constructed at a 16,500-ft site in northern Chile. The delay in funding the 25-m telescope ultimately will produce an extremely powerful millimeter-wave telescope that has everything: collecting area, angular resolution, and a site of incredibly low atmospheric extinction—and even a very large price tag. Meanwhile, the IRAM 30-m millimeter-wave telescope on Pico Veleta, the 10-m Heinrich Hertz Telescope (HHT) on Mt. Graham, the 15-m James Clerk Maxwell Telescope (JCMT) on Mauna Kea, and the Caltech 10.4-m telescope on Mauna Kea have become the leaders in single-dish millimeterwave astronomy, a richly-productive field of research pioneered by the origi-

13.6. A NEW 12-M SURFACE FOR THE 36-FT TELESCOPE

147

nal NRAO 36-ft telescope. Regrettably, the most sensitive of these telescopes have generally not been open to US astronomers except via collaboration. Figure 13.12 compares the collecting area of these principal millimeter-wave telescopes available to astronomers at the end of 2003.

13.6

A New 12-m Surface for the 36-ft Telescope

While the proposal to build the 25-m telescope slowly crept through the NSF bureaucracy, US millimeter-wave astronomers were hurting. The experimental but now outdated 36-ft telescope was the only single-dish type available to them without regard to institutional affiliation or travel expense, and with a subsidy for travel expenses. The IRAM announced that their new 30-m millimeter-wave telescope under construction near Granada, Spain, would be available only to astronomers from IRAM member countries—and only on a percentage relating to their country’s share of the construction and operating costs: Germany, 45%; France, 45%; and Spain, 10%. I felt this was an exceedingly uncharitable policy imposed by their selfish national administrators, considering that European astronomers had been and were continuing to enjoy equal access to all the NRAO radio telescopes at no charge, including the heavily oversubscribed 36-ft telescope and the VLA. Evidently, the Europeans were planning to follow up astronomical discoveries largely made at US observatories while excluding US astronomers not based at European observatories. But that was Europe at that time: politics before science. The Tucson staff indeed had investigated replacement reflectors. Bob Leighton, a Caltech physics professor, had designed and was constructing an inexpensive but highly accurate 34-ft (10.4-m) millimeter-wave telescope for the Caltech observatory in the Owens Valley, California. He was doing this on a “shoe-string” budget, using simple devices like water hoses to transfer levels from one side of the dish to the other and a lawnmower engine to turn a cutter to shape the aluminum honeycomb panels used in the parabolic surface. In 1976, John Payne, Dale Webb, and I visited Bob to determine whether we could build a copy ourselves in Tucson. After the visit, we felt that Tucson Operations simply would not have adequate staff to make one and maintain the 36-ft telescope at the same time. In April, 1980 [88], I wrote Mort Roberts (with a copy to Hein Hvatum) a memorandum describing the problems with the 36-ft reflector, summarizing the intense but unsuccessful efforts of the Tucson staff over the years to improve it by working on the surface and feed designs, and installing a sun

CHAPTER 13. THE 25-M TELESCOPE

148

Single-Dish mm-Wave Telescopes in 2003 450

400 IRAM 30m (70µm)

350

2

Effective Collecting Area (m )

300

250

200

proposed NRAO 25m m (75µm)

150

JCMT 15m (25µm) 100

NRAO 12m (75µm) Caltech 10.4m (15µm) 50 HHT 10m (13µm) NRAO 36ft (140 4 µm) 40 m 0 100

200

300

400

500

600

Frequency (GHz)

Figure 13.12: Collecting area of single-dish millimeter-wave telescopes available in 2003 plotted against frequency. The calculation assumes the Ruze efficiency formula, the RMS surface accuracy given in the identification label, and a zero-frequency aperture efficiency of 60%. It does not include the atmospheric transmission associated with each telescope. The University of Arizona now operates the “NRAO 12m” telescope. Author drawing.

13.6. A NEW 12-M SURFACE FOR THE 36-FT TELESCOPE

149

shield to minimize thermal distortion. I also told him that Bob Leighton had just telephoned me to offer to lend us “Dish No. 0,” as he called his prototype telescope, which would be finished within a month or so. There were two catches. First, the word “loan.” Quite possibly, the loan could be permanent but Bob could not be sure at this time. Second, he mentioned that the loan (or gift) would probably involve some unspecified consideration. Because of the tight schedule required to build the high-accuracy version intended for Mauna Kea, Bob said he needed a yes or no answer within a couple of months (say, June). He was willing to move the dish only once: to Tucson or to the Owens Valley. To supplement my memorandum, I immediately telephoned Hein to tell him about the offer. Evidently, after some months (“fall of 1980” [89]), Hein Hvatum asked Bill Horne to investigate. On January 9, 1981—way beyond the deadline— Bill wrote [89] Hein that the Leighton dish wasn’t the best option for the 36-ft mount and, hence, should not be considered by the NRAO. To reach this conclusion, he and Lee King had met with Bob Leighton to discuss his design and, subsequently, made some calculations. Bill’s memo stated that the NRAO would be better off with an entirely new design. Today, the Leighton dishes still comprise the amazingly productive millimeter-wave interferometer at Caltech’s Owens Valley Radio Observatory (OVRO), and the high-altitude version on Mauna Kea continues to operate very well. But Bill may have been correct that they weren’t suitable for a quick replacement to the flawed 36-ft reflector despite their eventual success as millimeter-wave reflectors. By November of 1980, I was even more concerned that nothing had happened regarding the Leighton offer. The NRAO needed to do something quickly about the 36-ft telescope to maintain research momentum. I wrote a memorandum [90] to the NRAO engineering staff, with a copy to Mort, and organized an ad hoc meeting in Charlottesville for Tuesday, December 9th. The topic was to be whether or not the NRAO could refurbish the 36-ft surface or replace it. We all knew that a new surface for the 36-ft telescope might obviate the need for the 25-m telescope and destroy any possibility for funding. Alternatively, concomitant new discoveries would strengthen the case for the 25-m telescope. It was risky. At the meeting, the consensus was to go for a new surface. I believe that most of us were growing weary of waiting for the funding of a proposal brought to NSF attention in 1975. On behalf of the NRAO, AUI had formally submitted it to the NSF in 1977, to provide a decent telescope for an extraordinarily productive, new research area in astronomy pioneered by US astronomers. In my opinion, the NSF should have funded it without

150

CHAPTER 13. THE 25-M TELESCOPE

hesitation. Then, as in recent years, the NSF funding for highly productive,“cutting edge” astronomy was painfully slow and usually inadequate. It took over ten years for them to decide to fund the successor project to the 25-m telescope, ALMA in Chile, a radio telescope backed by most astronomers and, monetarily, by the Europeans as well. An additional problem—discussed above—with the 25-m telescope was that, in some sense, it competed with activities at US universities, such as the millimeter-wave interferometers of Caltech and of U. C. Berkeley and with the 15-m millimeter-wave telescope of the University of Massachusetts. Somehow, a superbly maintained, large national millimeter-wave telescope—the 25-m telescope—on an excellent site (Mauna Kea) was seen as a scientific threat and a competitor for NSF operating funds rather than as a complement to those university’s research activities. At the Charlottesville meeting, John Findlay suggested that, first, the NRAO should make a strong case for the inadequacy of the 36-ft surface. Specifically, he suggested that the NRAO quantitatively investigate the thermal hysteresis of the telescope by installing a theodolite at the vertex of the dish and measure the position of the surface edge throughout a day. Indeed he did this. The measurements convincingly showed the 36-ft reflector to be so thermally unstable that nothing could fix it. While we had known this for many years, these specific measurements demonstrated the problem in a quantitative, straight-forward way that anyone could understand clearly. Accordingly, the NRAO decided to proceed with replacing the 36-ft surface with a new one while continuing to press the case for a new 25-m telescope on Mauna Kea. The plan was that the new surface should be inexpensive but highly accurate and stable. Lee King, Bill Horne, and John Findlay proposed a rigid space-frame overlaid with the highly accurate but inexpensive, pressed aluminum surface panels made by Environmental Space Systems Corporation (ESSCO) of Concord, Massachusetts. They selected a diameter of 12 m, the largest diameter possible allowed by the width of the slit of the astrodome. The initial setting of the panels was to be made by referencing them to a jig template, which pivoted from point to point around the lip of the new dish. The final setting would be done by holography. A 1984 article [91] in “Sky and Telescope” magazine describes the construction of the new surface. Findlay very much wanted it to be his project. He told Hein Hvatum that neither Hein nor I (the site manager at the time) were to be involved; neither of us was to visit the site until construction had been completed. His reasoning wasn’t based upon animosity but rather,

13.6. A NEW 12-M SURFACE FOR THE 36-FT TELESCOPE

151

perhaps, a desire to take full responsibility (and to receive full credit) for what he knew would be his last NRAO construction project—for him, a replay of the successful 300-ft telescope project he suggested and managed in the early 1960s in Green Bank. There may well have been another reason. Findlay had been project manager for the original 36-ft telescope but had left for Arecibo before the telescope had been completed and while the construction problems had appeared. It’s possible that he saw the resurfacing project to be an opportunity to atone for his earlier failure to follow a millimeter-wave project to completion. We’ll never know. Our Tucson employees did not know Findlay’s conditions and constantly criticized me for not participating more in the actual construction of the new surface. In the end, I did manage to foil Findlay a little by helping attach many of the ESSCO surface panels. But mostly, I kept to my Tucson office, just as Findlay had requested, taking advantage of the quiet to write some overdue research papers. I had also committed to a lecture series in Brazil, which unfortunately kept me from participating in the check-out of the 12-m telescope, a procedure delayed by months when the construction schedule slipped. The resurfacing went very well. Figure 13.13 shows the removal of the old, welded-aluminum, 36-ft reflector. Figure 13.14 shows the installation of the new back structure to the tower. Figure 13.15 shows the initial attachment procedure for the ESSCO surface panels. Figure 13.16 shows the mechanical measuring jig mounted on the new surface. Figure 13.17 shows a side view of the resurfaced telescope, and Figure 13.18 shows the details of the back structure and receiver support. The initial setting of the surface was good enough to focus astronomical radio waves, a necessity for proceeding with the holographic setting of the panels. This holographic technique compares the direct radio waves from a distant transmitter to those reflected from the telescope surface. In this case, the transmitter was a 38-MHz beacon of a satellite (LES-8) operated by MIT Lincoln Laboratory as part of a system to detect nuclear detonations. Located very far from the telescope, the direct satellite signal was a perfect plane-wave front. If the reflector surface was perfectly parabolic, the signal reflected from the surface would be a perfect spherical wave front, mathematically transformed to a plane-wave front. The direct and (transformed) reflected wave fronts were then compared. Any differences indicated imperfections in the telescope surface, which could then be plotted as a contour map. Having determined the surface errors from the holography, the staff ad-

152

CHAPTER 13. THE 25-M TELESCOPE

Figure 13.13: A crane lowers the old 36-ft reflector to the ground. In the background, the new 12-m back structure awaits lifting into the dome. Paul Rhodes photo.

justed the one or more of the four double-pitched bolts to move the panel edges closer to the design figure. The sophisticated computer program not only produced a contour map of the surface errors but it also listed the exact number of turns each bolt was to be moved. Furthermore, while the actual observations were made over a range of elevations, the program corrected these into what was needed for an “ideal” elevation for Galactic millimeter observations. The mathematical (NASTRAN) model of the back structure supporting the panels predicted exactly how the surface should deform with elevation angle. This level of adjustment did not result from a one-time process. First, the satellite moved in a vertically oriented “figure 8” orbit, changing its elevation by about 30◦ . Even though the mathematical modeling of the surface flexure as a function of elevation was very good, it wasn’t perfect. Hence, correction of the surface measurements to a single reference elevation contained some errors. Second, the panels were not completely stiff. The support bolts at the four corners over-constrained the panel. It was easy to

13.6. A NEW 12-M SURFACE FOR THE 36-FT TELESCOPE

153

Figure 13.14: The new space frame or “back structure” of the 12-m reflector is lowered toward the support tower. Note the stiff octagonal section at the center. Sections of the fabric-covered astrodome and roll-up slit are seen in the background. On the dome wall (left) is the galvanized pipe supporting the infra-red detectors that sense the telescope motion with respect to the astrodome. Paul Rhodes photo. unintentionally distort the panel surface while adjusting. For this reason, the holographic adjustment extended over several years and several sessions. As I recall, the final RMS error of the surface achieved by this process reached about 90 µm and was stable—much, much better than the old 36-ft surface. Figure 13.19 shows the performance as of February 22, 1983. After a few years, the improvements went even further. With the help of John Davis and Charlie Mayer from the University of Texas, the NRAO designed and machined a subreflector with a negative image of the surface errors. This gave a huge reduction to the effective surface errors. Although that process did not correct the surface errors perfectly, it did lead to an effective RMS error perhaps as low as 55 µm. If true, this was a remarkable achievement.With this refinement, the NRAO millimeter telescope, although with a new reflector, had achieved its original design goal, set in the 1964

154

CHAPTER 13. THE 25-M TELESCOPE

Figure 13.15: John Ralston, at left, using a theodolite to guide Paul Rhodes (hard hat) and Bill Horne in the initial adjustment of a newly attached ESSCO surface panel. Author photo.

Figure 13.16: The mechanical measurement jig used for the initial setting of the 12-m surface panels. The central end (left) lies upon a pivot; the outboard end (right), upon one of the fiducial pins set into the rim of the back structure. The vertical sensors seen on the truss transmit position information to computer. Author photo.

13.6. A NEW 12-M SURFACE FOR THE 36-FT TELESCOPE

155

Figure 13.17: Edge-on view of the 12-m telescope. Unlike the 36-ft telescope, the 12-m reflector surface consists of individual panels. The cooled receiver now lies behind the surface. The pier (tower) is the same as for the 36-ft telescope. Author photo.

156

CHAPTER 13. THE 25-M TELESCOPE

Figure 13.18: The back of the NRAO 12-m surface. The tubular truss supporting the panels cantilevers out from a stiff polygon structure near the center. The lettering indicates the new position of the cooled receivers behind the surface, allowing the mounting of up to four remotely selectable receivers and greatly facilitating access for service. The two engineers give scale. Author photo.

13.6. A NEW 12-M SURFACE FOR THE 36-FT TELESCOPE

157

Figure 13.19: The effective area of the new 12-m surface plotted against wavelength and frequency as measured on February 22, 1983 before the correcting subreflector. For comparison, the figure also shows the performance of the old 36-ft surface as well as the goal for the new surface. Author drawing

158

CHAPTER 13. THE 25-M TELESCOPE

request for proposals. The process had taken about 35 years. Making the holographic maps was productive in itself. According to Darrel Emerson, the early maps did not have enough resolution. They were only 65 by 65 points and would take several sessions to achieve. This gave Darrel an opportunity to implement “on-the-fly” mapping, a technique of recording data while the telescope was actually moving—made possible by the new, faster computer system. The new technique produced 129-by-129 point maps or, even, 257-by-257 point maps quite quickly. Making a 65-by65 map could take as little time as 20 minutes. These almost “snap-shot” maps showed that the surface figure changed at dawn, when sunlight began to heat the reflector surface and probably because the atmospheric turbulence affected the wavefront transmitted from LES-8. Comparing these maps with independent measurements of aperture efficiency showed good agreement, which hadn’t been the case for the early maps that had taken several sessions to make. But sometimes two steps forward are followed by a step backward. One day in 1989, two key struts in the elevation support to the back structure were found to have loosened [92]. In addition to noting a gradual deterioration in the pointing characteristics, Phil Jewell had discovered the exact cause while routinely making pointing corrections, this time with an optical telescope and charge-coupled device (CCD) camera attached to the telescope structure. There was a hysteresis in the elevation error, which suggested something loose along that axis. To make repairs, the telescope manager asked our telescope mechanic, Stan Sullivan, to replace the loose bolts with welds in the summer of 1989. In the following April, new holographic maps showed that this re-attachment had warped the figure of the dish, increasing the RMS error to about 125 µm! Figure 13.20 shows the distortion. This discovery led to another round of holographic measurements and construction of a second shaped subreflector. For some unknown reason, the new subreflector only corrected RMS to about 70 µm—short of the low value achieved earlier. The surface was still excellent but not as good as it once had been. Nonetheless, the new 12-m reflector was far better than the original 36-ft surface.

13.6. A NEW 12-M SURFACE FOR THE 36-FT TELESCOPE

159

Figure 13.20: This “holographic” image shows the distribution of surface errors across the surface of the 12-m reflector. The cross marks the quadrupod that supports the subreflector at the apex. It is rotated slightly with respect to the top and bottom to avoid holes in the aluminum panels. The dark areas at the top and bottom indicate the “clam-shell” distortion to the surface introduced by welding the loosened strut. Darrel Emerson photo.

Chapter 14

Odds and Ends Looking back, I recall a number of interesting periods that held an interest all their own.

14.1

NSF Reorganization

One event sticks in my mind but without a date. Around 1974, a senior NSF official visited our Forbes offices on a familiarization trip. I remember neither his name nor his position with certainty, and maybe that’s just as well. I suspect this was Thomas Owen, a retired Rear Admiral and the assistant director for Astronomy, Aeronomy, Earth, and Ocean Sciences (AAEO), accompanied by Dan Hunt who had just moved from the Head of National Centers to head the NSF Astronomy Division. But I do remember his message: the NSF had reorganized along research disciplines. Now, all astronomy, including national centers and university grants, were to be in a single astronomy division. Prior to this visit, all of the national observatories funded by the NSF were grouped in a separate division, a “national centers” division. The rationale was that, since their funding served all astronomers, their funding should not compete head-to-head with grants to individual astronomers or to smaller, university observatories. Partly because of this, university astronomers generally saw the national observatories as facilitating rather than competing with their research. It was a big, happy family. As one might guess, this reorganization of the NSF changed all of that. Because there was never enough NSF money to go around, the new organization meant that individual astronomers began to see the national observatories as competitors. Furthermore, this relationship was somewhat 161

162

CHAPTER 14. ODDS AND ENDS

ambivalent. While using national facilities, the astronomers appreciated how hard the staff worked to ensure their success. But returning to their home institutions, they were again faced with the realities of inadequate money. It was easy for the university astronomers to resent the comparatively financial comfort of the national observatories whose telescopes they had just used. The one big happy family was falling apart. In subsequent years, the NSF astronomy division heads were aware of this problem. Bill Howard was the first director of the Astronomy Division to have both the Grants and Centers Programs under him. At that time, the annual funding ratio of the national centers to university grants for astronomy was two to one. For consistency, Bill carried a pocket Hewlett Packard calculator programmed to maintain this ratio. To maintain it, Bill occasionally borrowed from the allocation to one center to fund something of another center, thereby keeping the ratio at 2.0. I don’t know whether this ratio persisted in the funding decisions of his successors. The pressures to lower it must have been enormous. In time, this ratio did decrease. Granting $700k for a construction “kit” for the 14-m millimeter-wave telescope at the Quabbin Reservoir proposed by the University of Massachusetts together with Amherst College, Hampshire College, Mount Holyoke College, and Smith College, raised the Grants side of the budget and lowered Bill’s target ratio below 2:1 [84]. In subsequent years, a number of significant additional grants was required to finish construction of this telescope. In any case, I believe this ill-advised organizational change within the NSF has heavily penalized the US astronomical community. The new funding climate divided astronomers rather than united them, significantly lessening community support for new national telescopes and consequently delaying if not killing their chances for funding. In fact, one director of a major university observatory always refers to Kitt Peak National Observatory as “the KPNO National Bank”—an undeserved slur on a national observatory largely responsible for the tremendous growth in US optical astronomy. As will be seen below, this NSF-induced ambivalence eventually killed the NRAO proposal to build a large, state-of-the-art millimeter telescope on Mauna Kea to replace the over-subscribed but flawed 36-ft telescope. This killing was tragic, considering that the older telescope had opened up a new, hugely successful venue for astronomical research. Other nations did not employ this irrational system of funding. They quickly built millimeter-wave telescopes to explore this newly opened area of research. Unfortunately, some of these were not available to US-based astronomers.

14.2. BARRY GOLDWATER’S VISIT

14.2

163

Barry Goldwater’s Visit

For his election day, November 4, 1974, KPNO had invited Senator Barry Goldwater to visit Kitt Peak to tour the facility. Around noon, Barry helicoptered in. On hand were Leo Goldberg, the KPNO director, Dick Doane, the KPNO mountain superintendent, a number of other representatives from KPNO, and some reporters. After lunch, the first stop was the observation platform of the 4-m optical telescope. Looking out over the summit area, Goldwater asked about the large, cloth-covered dome on the southwest ridge. When told it was a radio telescope, Goldwater asked if he could visit it. The KPNO hosts duly chauffeured him to our 36-ft telescope. At the telescope were graduate students Tom Bania and Jay Lockman, finishing up an observing session assigned to the two of them, Butler Burton, and me to map the large-scale distribution of CO emission in the Galaxy. Butler and I had left the observing run early. When Goldwater arrived, he was fascinated. A long-time radio amateur, he was especially interested in the technology. Furthermore, the relaxed, unplanned visit provided a great opportunity for the Arizona Senator to learn about what was happening in the world of radio astronomy. Jay Lockman [93] described the situation, “Needless to say, Tom and I looked the part of graduate students in 1974. I had a ponytail down to my waist and Tom had a handlebar mustache wider than his head. That look was not common on the mountain (or in Goldwater’s entourage) so we felt a bit conspicuous. If Goldwater noticed he made no sign though.” At his request, Tom and Jay showed him all the electronics they were using and, finally, displayed one of the CO spectral lines on the screen. When Goldwater asked what frequency the line was, they told him 115 GHz and the United States Senator and former Republican candidate for President exclaimed, “Holy shit!” [94]. He stayed for a long while, while the KPNO officials paced outside and looked at their watches, and reporters took notes. Later, Dick Doane told me that this would be the last impromptu visit of a government official to Kitt Peak. From then on, stops would be scheduled. Lockman and Bania’s candor, and the millimeter-wave telescope’s technical appeal, had unintentionally derailed KPNO’s plan. KPNO officials should have foreseen the situation. Unlike radio telescopes, optical telescopes are dormant during the day and probably would have bored Goldwater. Goldwater visited Kitt Peak again in the winter of 1991-92 after he had retired from the Senate. This time, he specifically came to tour the 12-m telescope with Darrel Emerson. While attending an amateur radio conference in Phoenix, Darrel met Goldwater, remarked that he knew that

CHAPTER 14. ODDS AND ENDS

164

the Senator had visited the 36-ft telescope in the past, and invited him to visit again verbally and, later, followed this up by a letter [62]. Goldwater was delighted to accept, again helicoptered to Kitt Peak, and spent the day learning about the refurbished telescope from the NRAO staff. To make the most of an opportunity, Dale Webb had rummaged through his library, had recovered his copy of Goldwater’s 1960 book “The Conscience of a Conservative,” and had Goldwater autograph it. It must have been a great day for Barry Goldwater as well!

14.3

The Chinese Visit

Soon after my wife and I arrived in Tucson, Bart Bok invited us to dinner. Professor Bok was an early patron of radio astronomy. While he was a member of the Harvard astronomy faculty in the 1950s, he had served as mentor for many graduate students entering the then new field of radio astronomy. Over his career, 55 graduate students had received their doctorates with Bok as their advisor. After Harvard, Bok and his astronomer wife, Priscilla, had gone to the Australian National University to energize their astronomy program, then, to the University of Arizona to build their department. Over the next few years, we kept in touch with each other. Bok’s interest in me was more than just courtesy. Passionately interested in the Milky Way, he was well aware of our research program of using carbon monoxide emission to map the cold molecular gas in the Galaxy. Ultimately, this contact resulted in an invitation to write an article on Galactic CO for Scientific American [48]. At a meeting of the American Astronomical Society, Bok introduced me to C. C. Lin. An expert in galactic kinematics, Professor C. C. Lin of MIT had been a close friend of Bok for many years. Lin had started spending summers in his native China, continuing his research and advising Chinese astronomers about new opportunities for observing. Based partly on his conversations with Bok and with me about the work being carried out at the 36-ft telescope, Lin recommended that the Chinese build a radio telescope for millimeter wavelengths, which they could use to search for molecules and to explore the interstellar medium of the Milky Way. In November 1976, to investigate the recommendation further, the Chinese sent a delegation of five astronomers and engineers to Tucson, to spend a week talking with the NRAO Tucson staff. Their leader was an older astronomer, Shou-Guan Wang, from Beijing Observatory, who had received his Ph.D. many years earlier in England.

14.3. THE CHINESE VISIT

165

Accompanying them was a woman from the Chinese consulate in Washington and two well-dressed but well-armed guards from the US Department of State. At that time, the US did not have full diplomatic relations with China, and only a consular office existed in Washington. The group also included a Chinese interpreter. The daily routine was always the same. Around 8AM, the mini-van with Chinese and guards left the Plaza International Hotel in Tucson and drove to the NRAO offices on Forbes Boulevard. They entered our conference room, took seats, produced small notebooks, and listened to a presentation from one or more of our staff. They took copious notes, occasionally asking a question through their interpreter. On one day we took them to Kitt Peak to show them the details of the 36-ft telescope and its equipment. Except for the leader, the consular official, and the interpreter, we did not know the particular field of expertise of any of the others. At the end of the day, they returned to the Plaza where—according to the guards—they discussed what they had heard and seen in front of the political attache. I thought it would be nice to invite them to my house for a dinner and asked the State Department guards if this might be possible. “Yes,” they said, and we selected an evening. Very soon, the CIA sent me an informative FAX regarding suitable food, advising me to avoid dairy products like milk and cheese because adult Chinese could not digest them. Accordingly, spouses of our staff members prepared appropriate dishes. But the most startling thing happened the day before the dinner. One of the guards visited all my neighbors to ensure that my dinner would not produce any kind of international incident, i.e, it would not be a “set-up” for a political gaffe. At the dinner itself, one guard remained outside, walking around my house, fully armed, to make certain no incident would occur. The other guard, wearing a suit and undoubtedly also well-armed, joined our dinner party and mingled with the guests. Later, when I asked why all that protection was necessary, the guards—one had been a lawyer— explained that they were simply duplicating the treatment given US visitors to China at that time. The visit evidently went very well. On the evening before their last day, the Chinese organized an excellent banquet for us, including Leo Goldberg and John Schaefer (the University of Arizona president at the time) who had recently returned from an official scientific visit to China (the NRAO director, Dave Heeschen, had been a member of that delegation), and other local officials at the Old Peking Restaurant on Speedway Avenue. On the next afternoon when we finally parted from the last office visit, they presented many of us with small gifts: silk scarves, small Chinese

166

CHAPTER 14. ODDS AND ENDS

paintings, etc. I hadn’t been prepared for the exchanging of gifts, so I gave each of them a US Government ball-point pen that the NRAO bought by the carton. Immediately after they left, Neil Albaugh came to me and asked if I’d ever taken one of these pens apart. I said, “No,” and disassembled the one he handed me. On the brass ink tube was carefully lettered the words “Made in Taiwan.” Good Lord! I had just committed a huge gaffe. After a few minutes of enjoying my discomfort, Neil grinned and told me not to worry. He had personally lettered that one just to stir me up! The outcome of that visit was the later construction of a millimeterwave telescope in western China. The ESSCO Corporation in Concord, Massachusetts, made the telescope parts in the form of a “kit” and MilliTech, Inc. (an offshoot of University of Massachusetts astronomy department) sold them the receiver and filter-banks. The NRAO gave them the FORTH datareduction software that we used at Kitt Peak—which Betty Stobie went to China to install. Eventually, the Chinese sent several astronomers to work at the NRAO to learn about modern radio astronomy.

14.4

AUI Board Meetings in Tucson

One of the most enjoyable activities of Tucson Operations was hosting the periodic “spouses” meeting of the AUI Board. Because the AUI Board members served for no fee, AUI had a policy of inviting spouses to one of their annual meetings. AUI was willing to spend whatever it took to make these visits enjoyable. The funds came from the fees paid to AUI to manage Brookhaven National Laboratory and the NRAO and were not subject to government procurement regulations. The first visit to Tucson during my time came in March, 1976. I was concerned that the trustees and spouses would be bored on the bus ride from the Arizona Inn to the 36-ft telescope on Kitt Peak, and back. Consequently, I spent two weeks reading everything I could on the history of southern Arizona and learning about the local botany from the staff of the ArizonaSonora Desert Museum. Dale and I even visited a cotton farm (Buckalew Farm) on the road to Kitt Peak so we would be able to explain to our bus riders the crop (long-stable “pima” cotton), the laser-leveled fields, and the role of adding sulfuric acid to the irrigation water. The day before the trip, I prepared boxes of cuttings that I could pass out as the bus moved through different botanical zones. Although the AUI board enjoyed the trip, I was the major beneficiary of the preparation by getting to know much more regarding the Tucson environment to which I’d moved only one and a half

14.4. AUI BOARD MEETINGS IN TUCSON

167

years earlier. The posh Arizona Inn—an Arizona legend in its own right—hosted the AUI visitors. Two notable events occurred. First, a disgruntled, recently fired, ex-employee of the Inn drove down Elm Street one evening, emptying a revolver into the Inn’s heavily paneled front door. That certainly created an aura of the Old West—or of the urban gangs that were still to come. Second, thieves broke into the room of the Board president and stole his wife’s new mink coat, an event that was much less admired by the Board members. Eventually, the Inn’s insurance paid for the coat, but only after AUI had engaged a prestigious Tucson law firm, who persuaded the Arizona Inn that they were indeed responsible for the loss despite claims to the contrary. Dinner was held in The Old Pueblo Club, a private Tucson club on the top floor of Tucson’s highest building, with magnificent views of Tucson’s night skyline. As was AUI’s custom, they invited representatives from KPNO and Steward Observatory to acknowledge the help they were providing the NRAO millimeter-wave operation in Tucson. Another visit in October, 1980, involved the same bus trip but a more ambitious dinner under the stars in the desert near Sabino Canyon. We used a US Forest Service group picnic area, surrounded it with hundreds of luminaria (candles in sand-weighted paper bags), and placed tens of yellowglobed Coleman white-gas lanterns behind the giant Sahuaro cacti and flowering ocatillo plants to make a theatrical rendering of a beautiful Sonoran Desert setting. On the afternoon of the party, the NRAO employees discovered the picnic area to be infested with red ants. Fortunately, liberal does of chlorodane—now banned but then legal—eliminated them by party time. A Tucson private club catered with cold vichyssoise, assorted vegetables, fresh French bread, beef filet mignon (cooked at the site) with a Bearnaise sauce, after-dinner salad, with baked Alaska flamb´e for dessert. Tuxedoed waiters and gowned waitresses served us. We had a full bar in this remote place, with red and white wines with dinner, and cordials afterward. It was a success as well as a lot of fun. This party was the first one for the newly appointed AUI president Bob Hughes. I remember well because, during the evening, Bob asked me to order an expensive, new toolbox for our mountain mechanic, Stan Sullivan. Evidently, Stan had approached Bob Hughes, explained his problem, and found a way to get around my earlier, “Let me think about it.” Naturally, I ordered the toolbox, which turned out to be an excellent investment. It soon became a game to see if we could outdo the previous dinner. Over the years we tried a number of variations for the spouses’ dinner. Perhaps the most successful was a “wash party” in October, 1990—a

168

CHAPTER 14. ODDS AND ENDS

Figure 14.1: An army ambulance—acquired through GSA excess property to be used as a truck at the NRAO’s VLA telescope—carries party-goers from a double-decker bus parked on Cloud Road, south across the Tanque Verde Wash, to AUI’s Wash Party on the south (left) bank of the wash on the grounds of Tucson Country Club. Note the hay bales for benches and the fire in the foreground. The Santa Catalina mountains lie in the background. NRAO photo. Tucson term for a party held in any of Tucson’s ephemeral, sandy-bottomed, dry water courses. The AUI president, Bob Hughes, advised us that financial times were tight and we should try to hide the financial opulence of AUI from our Tucson guests—particularly from those associated with KPNO. Fancy restaurants were ruled out. Our solution was to locate a suitable wash on the grounds of Tucson Country Club, where we could organize excellent food and drink in a cowboylike environment. Guests arrived from the north, were ferried across the Tanque Verde Wash in WW II four-wheel-drive ambulances (one appears in Figure 14.1), unaware they were only several hundred yards from the hidden TCC clubhouse and kitchens. TCC staff ferried the food from the kitchens using golf carts. We even rented portable toilets so guests had no reason to approach the plush but unseen clubhouse. Dale Webb built a dance floor.

14.5. CHANGING NRAO DIRECTORS

169

We hired a cowboy singer to sing “traditional” ballads during dinner and a small western band to accompany dancing. There was a bonfire, of course. The food was wonderful, thanks to enthusiastic support from Fred Carter, the TCC catering manager at the time. Furthermore, the guests included the optical astronomer Helmut Abt—long-time editor of The Astrophysical Journal —who kindly pointed out the constellations and cosmic objects to non-astronomer guests and optically ignorant NRAO radio astronomers like me.

14.5

Changing NRAO Directors

Changing NRAO directors had an influence on Tucson Operations. Since the inception of the 36-ft telescope, the NRAO director had been Dave Heeschen. He was an outstanding observatory director who delegated authority, encouraged imagination and resourcefulness, and rewarded outstanding performance. Wisely, he also demanded accountability; he expected his managers to perform well. All of us in Tucson loved working for him. After sixteen years as the NRAO director, Dave decided it was time to do something else. AUI formed a search committee. Finding a successor proved difficult. They met for about a year. The highest-ranked candidates did not want to leave their present positions. On my own, I visited Pat Thaddeus in New York to persuade him to reconsider, because I wanted someone empathetic to millimeter-wave astronomy. He told me he was very happy where he was and, especially, wanted a position where he could focus on research, which the NRAO directorship could never be. Finally, Morton Roberts—a member of the search committee and a senior NRAO astronomer—reluctantly and unselfishly agreed to take the NRAO directorship. Unfortunately for him, he was following an exceptionally gifted and widely admired director who would be a “hard act” to follow. Mort worked diligently at his job as director but, in my opinion, was not a natural manager of people. I found he tended to avoid and, occasionally, even isolate people who disagreed with him. On the other hand, he was highly knowledgeable about what was happening in astronomical research. Where I had an excellent rapport with Dave Heeschen, the chemistry between Mort and me wasn’t as good. I had been assistant director for eleven years and had successfully built the Tucson Operations into a highly reliable and efficient facility, just as Dave Heeschen originally had asked me to do. I had also managed the 25-m telescope project to the point where President Carter had twice submitted it to the Congress for funding.

170

CHAPTER 14. ODDS AND ENDS

However, in doing this work, I had also been highly activist and, in that process, had surely irritated some of the people I worked with. While I was on a sabbatical semester in Berkeley, Mort Roberts telephoned to tell me he was replacing me with Bob Brown. In retrospect, Mort probably did the right thing. Good personal chemistry is an element essential to effective management regardless of who’s “right” and who’s “wrong.” And I was somewhat burnt out. It was time for me to go. This was a strange period in the NRAO history. The newly appointed NSF director, Erich Bloch, strongly favored the funding of industry-related research rather than basic research. A former IBM engineer and vicepresident appointed by Reagan, Bloch served as director of the NSF from September 1984 through August 1990. Mathematics, chemistry, physics, and astronomy were out of favor in his vision of the NSF. Engineering sciences that produced results immediately available to US industry were in. Consequently, the NRAO’s funding was unusually tight, and AUI’s proposal for the highly needed 25-m telescope for Mauna Kea was going nowhere. Universities were complaining that too much NSF money was going to the national centers, and vice versa. Radio astronomers seemed to be polarized either toward the 25-m telescope or a stand-alone VLB array. Optical astronomers felt radio astronomers in general were receiving too much money, especially since the VLA had just been completed. The NRAO morale was generally low, observatory wide. Employees at one site were pointing accusing fingers at those at another, often passing along rumors that were patently false. I thought Mort did a number of good things for the observatory. He was an enthusiastic, accomplished astronomer who understood that the principal “mission” of the NRAO was to support astronomers, especially visiting astronomers. He wisely kept most of Heeschen’s policies and management structure in place; they had proven to be effective. His appointments of Martha Haynes, Ron Ekers, and Bob Brown as site managers for Green Bank, the VLA, and Tucson, respectively, were superb choices. Six years later, Mort returned to research and AUI was again looking for a new director for the NRAO. The new search committee ran into the same problems as its predecessor. As earlier, the highest-ranked candidates turned the NRAO down; they were happy in their present positions. Finally, Paul Vanden Bout, an astronomy professor at the University of Texas and a member of the search committee, agreed to accept the appointment. Paul was personable, politically skilled, and an experienced manager. Unlike Mort, Paul was not primarily known for his astronomical research. Rather, he knew well how observatories operated and was familiar with the technical

14.5. CHANGING NRAO DIRECTORS

171

aspects of millimeter-wave astronomy. He had directed the Millimeter-Wave Observatory (MWO) at the University of Texas, after organizing it as a cooperative venture with other organizations. And he would be a new member of the NRAO family, bringing a new vision of how we should best operate. This time, there were many changes to the observatory, many of which affected Tucson Operations. Paul hired a new business manager, Jim Desmond, to replace the fiscally conservative but extraordinarily competent Ted Riffe. Jim came to the NRAO from Brookhaven National Laboratory, which was a much larger organization (at one time, 5,000 employees) and had a different operating culture than the comparatively small, 400-employee NRAO. Operating funds became compartmented. No longer did each assistant director have commitment authority for every NRAO account. No longer did every assistant director and site business manager have access to all observatory expense records. No longer did the NRAO senior business manager in Charlottesville personally negotiate the five-year NSF contract; these negotiations were delegated to an impersonal New York City law office that may not have understood the historical reasons behind some of the special clauses. No longer were the site managers encouraged to comment on financial or administrative procedures before they were instituted (they were able to comment on astronomy-related procedures). Consequently, the NRAO acquired a different ambience. While there may have been good administrative reasons for these changes, they tended to isolate each observatory site from the others rather than emphasizing a “several sites but one-observatory” concept. Although, Paul Vanden Bout’s gregarious personality lowered these inter-site fences somewhat, it did not remove them. Furthermore, the new organization tended to emphasize the importance of a “head office” in Charlottesville, rather than the importance of making as many operational decisions at the sites as possible, where the NRAO most often interacted with the visiting astronomers for whom the NRAO was created. But, as things turned out, the NRAO was on the eve of its largest expansion, and Paul was the perfect person to interact with the NSF and the AUI board regarding this matter. After sixteen or so years, Paul Vanden Bout stepped down and, in September, 2002, AUI appointed K.-Y. (Fred) Lo as the NRAO director. Fred believed that the NRAO was not as organized as it should be, appointed several new oversight committees, and focused upon the daunting task of building the newly funded and costly ALMA telescope. He closed Tucson Operations because he believed it was no longer needed (see §16.3). Over the years, the NRAO had slowly progressed from an informal but effective organization under Dave Heeschen to a highly structured one under

172

CHAPTER 14. ODDS AND ENDS

Fred Lo. In fairness, the huge cost of the ALMA telescope as well as its international funding may have precluded the original, small, cosy NRAO. Time will tell which system works best.

Chapter 15

The MMA and ALMA The NRAO was now poised to enter the most ambitious part of its history.

15.1

The Millimeter-Wave Array

Much of what we had learned from the 36-ft telescope and from preparing the proposal for the 25-m telescope would prove useful again. In early 1993, Bob Brown and an NRAO astronomer associated with the VLA, Frazer Owen, were leading a design group for what was the MMA. It was an exciting idea, and I asked Bob Brown if I could help. Three years earlier in response to the recommendation [87] of the second Barrett Panel to build a national millimeter-wavelength synthesis array, AUI had submitted a proposal [95] to the NSF to construct an array of 40 8-m antennas, comprising a total collecting area of 2,010 m2 . These antennas could be configured from a compact array of 70 m in diameter to a widely spaced array of 3 km. The array could cover frequencies from 68 (λ4.4 mm) through 366 GHz (λ0.82 mm). In my view, the proposal had three major problems. First, the cost estimate of $120M was certainly too low, no matter which site was selected. Second, while no specific site was recommended, those listed all lay in the northern hemisphere. As we had painfully learned from the 25-m telescope project, northern hemisphere sites would generate only ambivalent support from universities like Caltech and UC Berkeley who were operating smaller millimeter-wave arrays. Ultimately, this would kill the proposal. Third, it wasn’t big enough to be operated by a US national observatory although the proposal followed the specific recommendation of the Barrett Panel. Any motivated university, like Caltech or UC Berkeley, could have found public 173

174

CHAPTER 15. THE MMA AND ALMA

or private funds to build and operate an array of this size themselves. Earlier, in October 1984, I had written a memorandum [96] that raised the eyebrows of the MMA planners. To maximize the usefulness of the MMA to US astronomers and to minimize direct competition with existing and planned arrays of US universities, I suggested that the NRAO build the largest possible millimeter-wave array in Chile. At the time, my preference was Cerro Morado, a plateau near Cerro Tololo Interamerican Observatory (CTIO) operated by AURA. Frazer Owen especially believed my suggestion to be “off the wall.” He preferred that the MMA be constructed on South Baldy, a mountain very near the VLA site in New Mexico where he worked. A Chilean site would work well. In 1983, I had unsuccessfully tried to persuade the NRAO director at that time, Mort Roberts, to build a clone of the 36-ft telescope on Cerro Morado. My plan was to operate the high-frequency receivers with this new telescope during the moist Arizona summers, when millimeter-wave observing was impossible, and vice versa. This would not only provide excellent facilities to astronomers year round but would give access to Southern hemisphere sites. Pat Thaddeus of the NASA’s Institute for Planetary Studies was already observing near there with his 1.2-m millimeter-wave telescope. In 1987, in a collaboration with European Southern Observatory (ESO), Chalmers University’s millimeterwave observatory in Onsala, Sweden, did exactly what I had earlier proposed to Mort. They erected a 15-m submillimeter-wave telescope (Swedish-ESO Submillimeter Telescope or SEST) on ESO’s La Silla site, not very far from Cerro Morado. Bob Brown, now associate director of the NRAO, welcomed me to the MMA project. I was certain we could work well together. We once shared an office in Charlottesville. And when Bob replaced me in 1984 for a year as assistant director for Tucson Operations, we got along well. I asked Bob if I could be spared serving on a committee. I dislike them intensely, believing them to move toward conservative, mediocre decisions and to waste committee members’ time by ponderous deliberations. We agreed that I could function on my own, with responsibility for developing more accurate cost estimates and for generating an operations plan for the proposed sites. As usual in the early stages of a project, the cost estimates for MMA components carried large uncertainties. These uncertainties were often asymmetrical, that is, an estimate might have uncertainties of +30% and -10%. In addition, the estimates for individual components could be bimodal, that is, either $40k or $20k, depending upon which design was selected. I wondered how to deal with these situations. By luck, Campbell Wade, a retired NRAO astronomer associated with

15.1. THE MILLIMETER-WAVE ARRAY

175

the VLA, showed me a spreadsheet program using “Fuzzy Logic.” Called FuziCalc, it was perfect for handling the asymmetrical estimates for the MMA. It always amazes me how the proper tool can sometimes slide into your life. As I recall, Bob Hjellming had introduced Cam to the program, which Cam was playfully using to estimate the time required to drive from Socorro to visit his son at the State University of New Mexico at Las Cruces. But what a tool! A Knoxville, Tennessee, company was marketing a spreadsheet program that used Fuzzy Logic for arithmetic operations involving the spreadsheet cells. Fuzzy Logic is a form of set theory developed in the mid-1960s principally by Lofti A. Zadeh of the Computer Science Department of UC Berkeley. A typical application is controlling elevators in high-rise buildings, where the load in each elevator car can vary from empty to ten or so people. To minimize transit time without discomfort to the passengers, the appropriate power delivered by the electric motors should vary depending upon the weight of the elevator car. By incrementing the electric power and sensing the corresponding change in acceleration for each increment, Fuzzy Logic controllers empirically determine the most appropriate acceleration and de-acceleration curves for each trip. A fixed mathematical algorithm could not possibly work in this situation, which requires the “trial and error” approach intrinsic to Fuzzy Logic calculations. While I was estimating the cost of the three North American sites, the NRAO decided to investigate a Chilean site. In early 1994, the NRAO had learned that the Japanese were considering building a large millimeter-wave array on Mauna Kea, where the National Optical Observatory of Japan (NAOJ) was building a large 8.2-m optical telescope called Subaru. Not only would two large millimeter-wave arrays in the northern hemisphere not make sense, the NSF would be unlikely to fund the US millimeter-wave array under these circumstances. Another site had to be considered. Rejected in 1984, Chile now seemed to a possible alternative. Fortunately, there was a lot of information regarding possible astronomical sites in Chile. ESO had surveyed the dry Atacama Desert of northern Chile in the early 1980s, ultimately identifying the Cerro Paranal site for their Very Large optical Telescope array (VLT). The Japanese had extensively explored the same regions to look for sites suitable for their proposed submillimeterwave array, beginning in 1992. In 1992, Phillipe Raffin and Alan Kusunoki of the Smithsonian Astrophysical Observatory (SAO) had searched the Atacama Desert, looking for a suitable site for the submillimeter-wave array (SMA) proposed by the SAO. Although the SAO eventually selected the summit region of Mauna Kea for this array, their informal report [97] on Chilean sites was a useful travel

176

CHAPTER 15. THE MMA AND ALMA

guide. Where exactly should one look first? The Atacama Desert is huge. The Japanese group had excluded all sites above 4,000 m. Fortunately, Angel Otarola, ´ as a student of geographical engineering in Santiago, had worked from 1984 to 1990 with a German university group investigating the crustal dynamics of Region II from the Pacific coast near Antofagasta all the way east to the Argentine town of Salta. Unlike the study by the Japanese astronomers, these studies included many regions above 4,000 m. Angel knew many of them. Furthermore, scientists from ESO and the Onsala Space Observatory had also explored [98] the northern part of Chile in 1991 to find a site suitable for a European millimeter-wave interferometer, later called the Large Southern Array (LSA). This group included Lars Baath and Angel Ot´ a´rola. With information from ESO, the Japanese and the German studies, and with the SAO report in hand, in May 1994, Paul Vanden Bout (the NRAO director at the time), Bob Brown, Hern´ an Quintana from Universidad Catolica, ´ Riccardo Giovanelli from Cornell, and Angel Ot´ arola, ´ then with ESO, explored high-altitude sites in northern Chile [99]. Bob Brown remembers the trip this way [100]. There were two rented four-wheel drive vehicles. First, the group visited two sites near Ollaque listed in the SAO report but the access roads to these were poor. At one point, Bob’s vehicle hit a bump causing the well-traveled infra-red hygrometer (see Figure 15.1) to hit the roof so hard that it never worked again. Exhausted, the group returned to their Calama hotel, arriving around midnight. Bob wanted a high-altitude site with easy access. The cost of building a paved all-weather road would be too great for the MMA project. Looking over his maps, Bob selected the road from San Pedro de Atacama to the Jama pass as the best way to reach the high altitude terrain. Furthermore, Angel was certain that Chile would pave this road. The next morning, the group set out up the Jama road. The condition of the road was terrible, with deep accumulations of fine sand waiting to stop the upward momentum of the vehicles, and with several sets of parallel tracks which other vehicles had made to avoid the sandy ruts. Around 4,000 m (13,100 ft), the road dust clogged the air filter of Bob’s Isuzu, and its upward progress slowed to a crawl. They stopped, and some of the group stayed there, taking photographs of the wonderful scenery. Meanwhile, Bob and Hern´ an Quintana switched to the other vehicle and continued upward to the Paso de Jama itself, confident that somewhere in the surrounding terrain would be the perfect site for the MMA. But they had run out of time, and the detailed exploration of that area had to await another visit to

15.1. THE MILLIMETER-WAVE ARRAY

177

Figure 15.1: The infra-red hygrometer designed by Frank Low in Green Bank around 1963, based upon a design developed at the US Weather Bureau [101]. It compared the absorption of solar radiation through the atmosphere at 9,359 ˚ A to the unabsorbed radiation seen at 8,800 ˚ A. Portable and reliable, this one was particularly well-traveled, having been used for the site surveys for the VLA, the 25-m telescope, the MMA, and Universidad Nacional Autonoma de M´´exico before meeting its end in Chile in May, 1994. The famous pilot, Francis Gary Powers, provided the original “above the atmosphere” reference calibration for the design from his U2 aircraft. I believe Dave Hogg assembled this one. My fountain pen illustrates the size. Author photo.

178

CHAPTER 15. THE MMA AND ALMA

Figure 15.2: A composite photograph of the ALMA site at the Llano de Chajnantor taken in 1994, looking west from an overlook at about 18,000 ft. Photo by Simon Radford. Chile. A follow-up trip in October, 1994, included Hugh Van Horn and Bill Harris from the NSF, Paul Vanden Bout, Bob Brown, and Geraldo Villadares. Ot´arola was not present on this trip. This time, Bob and Geraldo managed to reach what is now known as the ALMA site. Vanden Bout did not reach it because Harris became uncomfortable with the altitude, and Paul elected to drive him back to a lower altitude. The site lay at 16,500 ft, east of the village of San Pedro de Atacama. I later named this site Llano de Chajnantor. As soon as Bob saw it, he knew it would be perfect for the MMA, which later became ALMA. When they returned to the United States, Bob Brown telephoned me to ask me to include a Chilean site in my costing. Figure 15.3 shows the cost estimates I produced in March 1995 for all four sites considered viable for the MMA. These calculations were a direct result of understanding what’s required to operate a visitor-oriented millimeter-wave telescope and of procedures we had developed for the 25-m proposal. But in this case, I had used FuziCalc. Soon after, in November, 1994, an NRAO delegation of technically oriented people, including Peter Napier, Frazer Owen, Simon Radford, and Juan Uson, together with Angel Ot´ a´rola from ESO, visited the Llano de Chajnantor site specifically. Peter and Simon climbed to the summit of Cerro Chajnantor to take the photograph in Figure 15.2—a tough climb even from the 16,500-ft Llano; the summit is about 18,500 ft above mean sea level.

15.1. THE MILLIMETER-WAVE ARRAY

179

Figure 15.3: The summary page for preliminary Fuzzy Logic cost estimates for construction of the MMA on South Baldy in New Mexico’s Magdalena Mountains, Springerville, Arizona, Mauna Kea, Hawaii, and Llano de Chanantor in Chile in thousands of 1994 dollars. The slightly asymmetrical “belief graphs” show the cost uncertainties for each site. The possibility that the cryogenic receiver cooling could have either a single (less expensive) or double dewar (more expensive) gives bimodal peaks.

CHAPTER 15. THE MMA AND ALMA

180

It was now time to develop a plan for operating the MMA in Chile as well. I went to Chile in April 1995 and visited all of the principal optical observatories, spoke with everyone I met about how best to operate in Chile, and devised a detailed plan [102] for operating the MMA in Chile. I estimated the annual operating cost to be $7.5M in 1994 dollars.

15.2

ALMA

By 1998, ESO and the NRAO merged their projects into a much larger array to be known by 1999 as the Atacama Large Millimeter-wave Array or ALMA. To make my earlier report relevant to ALMA, I needed only to modify it for the larger array planned for the same site and, of course, to refine what I had learned earlier by making additional trips to Chile, including visits to new copper mines being developed at sites above 12,000 ft. Because of the superb Atacama site, the ALMA plan had even more demanding objectives than the MMA. Well-suited to the thin atmosphere of high elevations, the ALMA antennas will operate to almost 950 GHz (λ350 µm). The array itself will be much larger than the originally planned MMA, having sixty-four 12-m antennas and, at its highest frequency, an angular resolution of 10 milliarcseconds—ten times better than the VLA or the Hubble Space Telescope. This angle is the diameter of a US dime seen from approximately 2,300 miles. Surviving these additional trips to Chile was a learning experience in itself. The most direct route from Tucson was an overnight American Airlines flight from Dallas to Santiago. The NRAO allowed only coach travel. I quickly learned to fly in mid-week and to reserve the center seat of a row of three towards the rear of the Boeing 767 aircraft. With my “Gold Card,” American would protect a seat on one side of me. With the lower mid-week load factor, the chances were excellent that no one wanted the other seat, and I could stretch out for a cramped but welcome sleep. Chile also brought some unexpected experiences. One Saturday morning in Santiago, my wife, Julie, and I were walking from our hotel to an open market located in a poorer section of the city. Suddenly, I felt someone put his hand in the left pocket of my trousers. Instinctively, I put his head in a headlock and tried to wrestle him to the sidewalk. But he had an accomplice who moved in and ripped off one of the legs of my trousers, grabbed my wallet, and ran off. Running after him with trousers in tatters and underwear showing, and shouting “ladrones, ladrones” (thieves, thieves), I actually gained on him enough that he tossed out my wallet to divert my

15.2. ALMA

181

pursuit—after having taken my money. Picking up the wallet and seeing that my credit cards were all there, I embarrassedly hurried into a nearby department store and bought a new pair of trousers. The next morning, my leg muscles were so stiff I could hardly walk. There’s a big difference between the physical condition of a 60-year old office worker and a 20-year old thief! There were enjoyable times as well. Tired of spending weekends alone in Santiago, I remembered that a high-school classmate, Dave Gould (Phillips Academy, Andover, class of 1955), lived in Santiago. I telephoned him. He remembered me well and invited me to dinner. We immediately reconnected even though decades had passed. A Stanford-trained economist raised in Granby, Colorado, he had spent most of his professional career in South America, retiring at age 60 from the UN Economic Commission for Latin America and the Caribbean (known as CEPAL, in Spanish). He was married to a Peruvian, Nora, whom he had met while serving in the Peace Corps in Peru in the 1960s. Visiting them soon became a pleasant weekend habit when I was in Santiago. On one visit, I asked him why he owned a new diesel Land Rover when he lived in the middle of a large, sophisticated city. He told me that his dream had been to make a car-camping trip to the southern tip of South America but Nora refused to go with him. “I’ll go with you,” I said. Returning to Tucson, I described the possibility to my geologist wife, Julie. She immediately said, “I want to go, too.” When Nora heard that Julie was also going, she relented and decided to come along as well. The trip was on. What a terrific trip! We took a month to drive from Santiago to Ushuaia and back, taking a 3-day ferry south from Puerto Montt through the Chilean archipelago to Puerto Natales and, later, crossing the Magellan Strait from Punta Arenas to Isla del Tierra Fuego to reach the Beagle Channel. The return trip went through Argentine Patagonia, through the always windy steppes and across the southern Andes to Cochrane, Chile. We took the winding Carretera Austral back to Santiago. Often the roads were terrible but the Land Rover performed flawlessly. At times, we needed our GPS to know where we were. Lodging was a combination of camping and guest houses. But Nora hated the camping, much preferring the comfort of the guest houses. On the return trip, thieves stole their tent from the car’s luggage rack while we stayed in a guesthouse in the Argentine town of San Carlos de Bariloche. When I broke this bad news to Nora, she cheered, saying “God is looking out for me!” Despite the time I spent in Chile, my Spanish never became fluent. To acquire what little skill I had, I had spent three weeks at an intensive lan-

182

CHAPTER 15. THE MMA AND ALMA

guage school in Antigua, Guatemala, in 1994. Immediately following this experience, my Spanish-speaking wife told me I had come a long way and that my accent wasn’t too bad. Alas, in a short time, my newly acquired skills began to ebb. Although I continued to read well and speak somewhat, I had difficulties understanding. In part, this was because I had blown away much of my hearing as an 11-year old child, when I played with explosives. But it could also be the way my brain was wired. On one visit to Santiago, I remember sharing a cab with Bob Brown who could understand rather well but could not speak because he did not know the grammar. We experienced the absurd situation where the driver would ask questions, Bob would translate into English for me, and I would answer the driver in Spanish. ALMA-related trips could also be fun and, sometimes, a little challenging. On one occasion, Eduardo Hardy and I flew to La Paz, Bolivia, to investigate how well oxygen enrichment devices worked. I was planning to recommend their installation in the support buildings at the 16,500-ft ALMA site. Born in Argentina, raised in Santiago, and a former astronomy professor at Quebec’s Laval University, Edy had joined the ALMA project in late 1997 as the AUI representative in Santiago. He dealt with the politics between AUI, ESO, and the Republic of Chile, I dealt with planning site construction and operations. We were a good team. The La Paz airport at El Alto on the altiplano lies at 13,300 ft. The city of La Paz itself is built in a nearby ravine, with the best hotels situated near the bottom to maximize oxygen for their guests. For the same reason, their bars and dining rooms were situated on the lowest floors, often in basements. How a visitor could imbibe at that altitude, I’ll never know. After walking around La Paz for a few hours, we more or less crawled back to our hotel because of the altitude. The next day, with a company representative, we visited several medical clinics and one glass-blowing factory. These organizations used US-made enrichment devices to raise the partial pressure of oxygen in their patient’s rooms, in the case of the clinics, and, for the glass factory, the melting furnace. Evidently, the devices were effective, completely reliable, and easy to service—good news for my site planning. A couple of years later, Caltech field-tested these enrichment devices in the small laboratory they built at the ALMA site for a cosmic background experiment. They also found them to work well. Returning to Chile, the aircraft stopped for two hours at Chile’s tiny, northernmost town of Arica on its route to Calama. During the stop-over, Edy and I took a cab to explore the town thoroughly, joking that we could get rich by writing a best-selling guidebook for visitors. The town really was in-

15.2. ALMA

183

teresting, with a church designed by the famous Gustav Eiffel of Eiffel-tower fame, a good archaeological museum, and a military museum documenting the “War of the Pacific” of 1879-83 in which Chile substantially expanded its northern territory. Modifications to the earlier reports resulted in “ALMA in Chile, A Plan for Operations and Site Construction” that summarized my four years of investigation [103], appearing in May, 2000. This was my “Swan Song” with regard to the ALMA project. I had completed the task given to me and returned to the NRAO research staff where I (and a Russian colleague) finished an overdue astrophysics book [104] about radio recombination spectra, which I had agreed to write in 1993. Enriched by what I had learned traveling through Chile since the first report, this plan was much more detailed. It estimated the cost of construction of support facilities to be $90M in 2000 dollars, excluding construction of the ALMA instrument itself. Annual operating costs would be about $33M, based upon the exchange rate between the peso and the US dollars in 2000. Bob Brown combined these estimates with the costs of the ALMA array itself, delaying some of the equipping cost from construction budget to the operations budget, thereby obtaining a total construction cost of about $550M in 2000 dollars. The NSF, the National Research Council of Canada (NRC), ESO, and Spain divided this cost so that it split almost equally between North America and Europe, with a cooperative involvement of the Republic of Chile. ALMA is a direct result of the NRAO’s years of experience with the 36-ft telescope and the 12-m telescope, and with the proposal for the unfunded 25-m telescope. We certainly did not do it alone. The operational details and array design also benefitted greatly from the NRAO’s experience with the VLA in New Mexico—especially from the work of Mark Holdaway, Peter Napier, Frazer Owen, and Dick Sramek. Augmenting all of this is the experience of the ALMA partners operating their own telescopes in Chile, in Europe, and in Hawaii. The experience of a sister NSF-funded observatory, CTIO, in Chile was especially helpful. The NSF funded the Design and Development phase of ALMA in FY1998 and provided the first construction money in FY2000. ALMA officially broke ground for construction in Chile on November 6, 2003. The NRAO’s successor to the 36-ft was finally under way, even though thirty years had passed from the first request to breaking ground!

184

CHAPTER 15. THE MMA AND ALMA

Figure 15.4: An artistic sketch of ALMA’s 64 antennas on the 5,000 m (16,500 ft) Llano de Chajnantor site, viewed toward the northeast. This configuration is the “close-packed” array. The diameter of each antenna is 25 m. In the background are the service facilities. Note the tractor at the lower right, moving an antenna to a new position. European Southern Observatory drawing.

Chapter 16

The Twilight Years Closing the millimeter-wave observing facilities on Kitt Peak and, later, the NRAO offices in Tucson was a painful process for everyone. It occurred in stages.

16.1

Closing the 12-m Telescope—Part 1

The first thing I remember regarding closing the 12-m telescope was an event that occurred the morning after I returned from a sabbatical year in Bonn. On October 10, 1989, I arrived in our Tucson offices and learned that the NRAO director, Paul Vanden Bout, had unexpectedly flown to Tucson to inform the Tucson staff, at an early morning meeting, that the NRAO had decided to close the 12-m telescope. What had occurred was that the NSF proposed allocation to the NRAO would have been too small to continue operations as usual. Erich Bloch, the current NSF director (also see the discussion in §14.5), seemed to follow the philosophy of the Reagan administration regarding basic research. This administration felt that basic research should be funded mainly within corporations and not by the taxpayer. Consequently, Bloch reduced or froze funding for basic sciences but allocated new NSF money to applied research that might benefit US industry. Although there is certainly an overlap, science is not always engineering. Usually, “science” involves understanding whereas “engineering” involves a product. Given the specific wording of the 1950 Act establishing the NSF “to promote the progress of science,” it appeared that the Reagan people had hijacked the agency. Astronomy—and, hence, the NRAO—was in trouble. The annual NSF allocation to the NRAO seldom involved a dialogue. 185

186

CHAPTER 16. THE TWILIGHT YEARS

With arcane logic, senior NSF officials allotted money to the NRAO annually with little input from the NRAO. During my term as an NRAO assistant director, I saw the allocation as a process in which the annual grant was based upon the previous year’s allocation and a guess as to how much money would be granted to the NSF Astronomy Division for the next year. Excellent performance or need evidently weren’t definitive factors. The NRAO and its sister national observatories always seemed to be at the bottom of an arbitrary process within the NSF. For example, the previous Astronomy Division head, Bill Howard, tried to follow a “2:1 rule” no matter how much money his section received, that is, twice as much money for the national astronomy centers that all astronomers used than the total for individual grants. It was a nerve-wracking, frustrating situation to be in. Also to his credit, Bill instituted and maintained a ranked list of new initiatives for astronomy. Because allocations to astronomy were always too small and unpredictable in amount, this kind of planning allowed the Astronomy Division to adapt to whatever bones they were thrown. In a breakfast meeting with Darrel Emerson and Dale Webb, Paul had decided to gamble with the 12-m telescope to persuade the NSF to increase their allocation. According to Dale, Paul said, “I can only play this game once. If the NSF does not grant additional money ($2M) for operations, I am prepared to go through with my threat to close the 12-m telescope.” Later that morning, he shared the same logic with the entire staff of Tucson Operations at a meeting I missed. Later that day, when I discussed it with him, Paul told me that if I had any influence on getting additional money out of the NSF, I should get to that as soon as possible. Accordingly, I contacted Senator Dennis DeConcini and met with Congressman Jim Kolbe to see if they could persuade the NSF to increase its allocation to the NRAO by $2M. I was unable to meet faceto-face with Senator DeConcini (whom I knew socially) but was in almost daily contact with his staff. I telephoned their offices to assess progress and, in turn, their staff called me to report on developments. Meanwhile, millimeter-wave astronomers wrote letters and made telephone calls to Paul Vanden Bout, to NSF officials, and to the congressional representatives. Arizona State University professor Lucy Ziurys, in particular, organized much of this campaign. During this crisis, an acerbic but hilarious, serialized saga began to appear. Titled “The Rise and Fall of the Twelfth Reich,” the saga consisted of anonymously authored, fictional episodes of the attack of Charlottesville management on the staff of the 12-m telescope. Lucy M. Ziurys [105] told me, “ ‘The Rise and Fall’ was written over a period of about a year. The

16.1. CLOSING THE 12-M TELESCOPE—PART 1

187

first three chapters appeared after PVB [Paul Vanden Bout] announced his intention to close the 12 m. Chapters were added as time went on. L. J. Sage contributed scenes 6 and 7, with slight modifications. I wrote the rest, with help from the 12 m operators.” My contact with Senator DeConcini paid off—surely assisted by the many letters written by concerned astronomers. The Congressional Record [106] reported this dialogue between Senator Barbara Mikulski, a ranking member of the Senate Appropriations Subcommittee for Veterans Affairs, Housing, and Urban Development and Independent Agencies that includes the NSF, and US Senator Dennis DeConcini, a ranking member of the Senate Appropriations Subcommittee on Treasury, Postal Service, and General Government. “ASTRONOMY OBSERVATORIES IN ARIZONA Mr. DeCONCINI. Mr.President. I commend Senator Mikulski for her efforts in placing proper priority and emphasis upon radio and optical astronomy. As the Senate committee report noted: Astronomy research has taken a back seat to other priorities of the National Science Foundation. Astronomy research must receive adequate resources to insure the viability of the national astronomy centers. I agree with the committee’s statement. Most university astronomers depend upon the national observatories for their research. The National Radio Astronomy Observatory and the National Optical Astronomy Observatory operate in Arizona. These facilities operate world-class telescopes for visiting astronomers. Last year alone the Tucson radio telescope housed 127 astronomers from 58 institutions, including representatives from 10 foreign countries. Recently, several Arizona newspapers have carried articles that express grave concern over the financial viability of the Kitt Peak facility during fiscal year 1990. Therefore, I would like to ask the Senator from Maryland if the bill now before the Senate provides adequate funding to the National Science Foundation to insure the continued operation of these observatories in fiscal year 1990? Ms. MIKULSKI. The fiscal year 1990 appropriations for the National Science Foundation’s “research and related activities” adequately funds astronomy research endeavors. The committee has expressed its concern to the National Science Foundation regarding the priority of astronomy research. Accordingly, we are

CHAPTER 16. THE TWILIGHT YEARS

188

confident that the National Science Foundation will distribute the research funds equitably to insure the continued success of US astronomy research in addition to the continued operation of the National Radio Astronomy Observatory’s millimeter-wave research facility [the emphasis is mine] and the National Optical Astronomy Observatory in Arizona.” Responding to this pressure, the NSF gave $2M more to the NRAO for FY90 to continue operations of the 12-m telescope. This was democracy in action. We had dodged the bullet by bringing our problems to the attention of our elected representatives. Our victory came at a price. The episode had substantially damaged our attitude toward the NRAO management. Prior to this episode, Tucson employees always believed the Charlottesville management looked out for them. Consequently, they had a “can-do” attitude toward their work and were happy to work extra hours to assist visiting astronomers. After this “bet the observatory” incident, the Tucson staff felt unappreciated. They no longer trusted Charlottesville management. In particular, Vanden Bout was no longer considered to be a friend of Tucson Operations, as shown by the bitterness of “The Rise and Fall” saga. And, I suspect, the relationship was also damaged from Charlottesville’s viewpoint. For perhaps the first time, Vanden Bout must have viewed us as capable of causing trouble for the NRAO as a whole.

16.2

Closing the 12-m Telescope—Part 2

The “closing” chapter was not finished. Eleven years later, on February 22, 2000, the NRAO issued a press release announcing that the 12-m millimeterwave telescope would close on July 1, 2000. “ ‘The action was made necessary by the current and anticipated budget for the observatory,’ Vanden Bout said...‘We understood that ALMA would eventually replace the 12 Meter Telescope, but we had hoped to continue operating the 12 Meter until ALMA began interim operations, probably some time in 2005. That is not possible, and we are forced to close the 12 Meter this year.’ ” Such was the fate of one of the most productive telescopes ever constructed. Eventually, the NSF and the NRAO transferred the telescope and all of its auxiliary buildings on Kitt Peak to the University of Arizona, whose astronomy department is operating it as I write. Were there unstated politics behind the closing? I don’t know. But I suspect that the NRAO and its legal entity, AUI, wanted to start the

16.3. CLOSING OF THE NRAO’S “TUCSON OPERATIONS”

189

momentum of transferring activity from a small telescope that now competed head-to-head with university facilities to the construction of ALMA. Unlike the 12-m telescope, ALMA would be a huge, unique facility that would augment university research in the United States rather than compete with it. I see support of huge, rather than small, facilities as the appropriate long-term role for a US national observatory like the NRAO. Looking back, astronomers should recognize that the 36-ft and its 12-m successor filled exactly that same role years earlier. Indeed, those very ventures led to the designing and funding of ALMA.

16.3

Closing of the NRAO’s “Tucson Operations”

After the closing of the 12-m telescope, the remaining Tucson staff spent their time applying their expertise with millimeter-wave hardware to designing equipment for ALMA. I believe that I alone remained as a member of the NRAO basic research staff pursuing astronomical research in the Tucson offices. AUI announced on November 9, 2001, that Paul Vanden Bout would step down as director of the NRAO on June 1, 2002. On September 1, 2002, Kwok-Yung (Fred) Lo replaced him. After spending about six months visiting all the NRAO sites, Fred felt it was in the best interests of the ALMA project to close Tucson Operations and move their employees to the NRAO headquarters in Charlottesville. There, a larger concentration of ALMA employees would promote better communications and less travel. Furthermore, there was only one (± one, depending upon the season) time zone between Charlottesville and Chile, compared to the three or four between Tucson and Chile. By June 2006 (or before), the NRAO’s Tucson Operations were to be closed. Remaining would be the two technicians supporting the Kitt Peak element of the VLBA, which was managed through the NRAO offices in Socorro, New Mexico. While the end of the Tucson era had finally arrived, the ALMA telescope would carry on the millimeter-wave activity into which the NRAO had entered in 1960. ALMA was a direct result of AUI’s original support of a 36-ft millimeter-wave telescope for the new National Radio Astronomy Observatory. May ALMA be as productive as its NRAO predecessors!

Appendix A

Time Line 1956, November 17, contract signed between AUI and the NSF, creating the National Radio Astronomy Observatory (NRAO) 1957, October 17, dedication of the NRAO. Lloyd V. Berkner is the first observatory director 1959, July, Otto Struve becomes the second NRAO director 1960, Frank D. Drake flies to Dallas to interview Frank J. Low, then working for Texas Instruments 1961, Frank J. Low arrives in Green Bank 1962, feasibility contract for a millimeter-wave telescope awarded to Rohr Corporation 1962, October 1, David S. Heeschen becomes the third NRAO director 1964, AUI awards a construction contract for the 36-ft telescope to Rohr Corporation 1964, Frank Low moves to Tucson, becoming a research professor of astronomy at the University of Arizona 1967, first scheduled operation of the 36-ft telescope 1969, E. E. (Ned) Conklin arrives in Tucson as the first astronomer/manager 1970, Robert Wilson, Keith Jefferts, and Arno Penzias [35] detect CO emission with the 36-ft telescope 1971, Charles Moore installs a FORTH computer system at the 36-ft telescope 1973, October 1, Mark A. Gordon becomes the first NRAO assistant director for the newly created “Tucson Operations” 191

192

APPENDIX A. TIME LINE

1978, Morton S. Roberts becomes the fourth NRAO director 1982, work begins on replacing the 36-ft surface with a 12-m one 1984, June 1, Robert L. Brown becomes the second (temporary) assistant director for Tucson Operations 1985, Paul A. Vanden Bout becomes the fifth NRAO director 1985, July 5, David E. Hogg becomes the third (temporary) assistant director for Tucson Operations 1986, November 1, Darrel T. Emerson becomes the fourth assistant director for Tucson Operations 2000, February 22, Paul A. Vanden Bout announces the closing of the 12-m telescope 2002, September 1, K.-Y. (Fred) Lo becomes the sixth NRAO director of the NRAO 2006, June 1, the date when Tucson Operations is scheduled to close

Appendix B

List of Tucson Employees Biemesderfer, Christopher D. Ade, Peter A. Albanna, Sarmad H. Albaugh, Neil P. Anderson, James Andre, Philippe Jacques-Antoin

Biller, Beth Bishop, John C. Bloomingdale, Richard W. Bowles, Larissa M. Branch, Clair

Andrews, Sean M.

Brasso, Thomas H.

Armstrong, Justin

Brod, Landford G.

Arora, Radhe Shyam

Brooks, Michael J.

Atencio, Nelson (VLBA only)

Brown, Margie M.

Baker, Walter Gregory

Bundy, Tad C.

Ballou, Gary D.

Burhans, Ralph W.

Balonek, Thomas J.

Burnell, James D.

Banda, Juana I

Cardarella, Donald J.

Bass, Daniel L.

Carrad, Graeme

Bates, Ronald W. (VLBA only)

Chang, Jack Jui Lin

Becker, Ralph

Chase, Dennis A.

Behrens, George Jr.

Cheng, Jingquan

Bessett, Rodney

Clark, Anthony

Bezkocka, Peter P.

Clark, Cedric Duane

Bielas, Michael S.

Clarke, Jeffrey S. 193

194

APPENDIX B. LIST OF TUCSON EMPLOYEES

Clarke, Nancy

Freund, Robert W.

Cochran, Jackie

Fuller, Gary

Collenberg, George M.

Gacon, Frank Stanley

Conklin, Edward K.

Galhouse, Stephen L.

Conner, Eugene Paul

Garagnon, Bruno

Cuadra, Rodrigo Andres Brito

Gasho, Victor L.

Cull, Selby C.

Gay, Pamela Lynn

D’Addario, Larry

Gensheimer, Paul David

Daniel, Billy J.

Giddings, Dale R.

Davis, Evan Rodier

Gobin, Maria C.

Davis, Gerald L. Jr.

Gordon, Mark A.

Davis, Jesse E.

Grammer, Wes L.

Dionne, Ronald Alan

Greve, Paul

Douglas, Rosalie G.

Groppi, Christopher

Dowd, Andrew

Grove, George

Dressel, Linda L.

Gust, William

Edwards, Suzan

Hagar, Lee P.

Emerson, Darrel Trevor

Hagen, Jeffrey

Emerson, Nicholas

Hagstrom, Magne Billy

Engel, Lisa

Hale, William Robert

Escalante, Fred Jr.

Halliday, Margaret A.

Essenburg, Alvin E.

Hamed, Julian M. “Tony”

Fagg, Henry Alen

Hamilton, Robert A.

Figueroa, Delia M.

Harmless, Kent M.

Fischer, Lynn S.

Harsha, Nancy L Gunn

Fitzner, John Tracy

Hart, Paul O.

Flynn, Sarah Kathryn

Hay, Cyrus C.

Folkers, Thomas Wesley

Heckler, Christopher Warren

Forster, Vincent

Helfer, Tamara

Foster, Scott

Hersman, Michael S.

195 Highberger, Jaime L.

Lim, Fai (Janne) Jen

Hill, Reuben

Linnaus, Fredrick

Hill, Timothy

Lipscomb, Charles E.

Hogarth, Robert E. Jr.

Long, Kevin

Hogg, David

Lugten, John

Holdaway, Mark

Lynn, James M.

Hollis, Jan M.

Mangum, Jeffrey Gary

Holmstedt, Christian

Marks, Janice

Howard, Richard J.

Marquez, Ivan L.

Jablonka, Paul H.

Martin, Hubert M. III

Jacques, Christophe

Martin, Joan

Jesch, Nicholas (VLBA only)

Martinez, Ricardo

Jewell, Philip Ramer

McBrian, John E. Jr.

Johnson, Keith

McFarlin, Rayford (VLBA only)

Jones, James B.

Mead, Kathryn Nadia

Kemp, Ernest D. Jr.

Meadows, Hollister

Kingsley, Jeffrey

Metcalfe, Mark

Kingsley, Kimberly T.

Middleton, Gerald E.

Kingsley, Robert K.

Miller, Diane Elizabeth

Kogan, Leonid

Miller, John E.

Kolor, Francis P. O’Mahony

Miller, Luther

Krauska, Alexander S.

Montierth, Jeanette Marcroft

Lamb, James W.

Morin, Thomas

Lapedes, Alan S.

Morreale, Jay Philip

Lasater, Martin L.

Morrey, Graham G.

Lasendby, Anthony N.

Mortenson, Gustave E.

Latter, William Bruce

Murphy, Patrick Paul

Lewis, Faye M.

Myers, David

Lewis, Karen M.

Nam, Chong Woo

Lichtenhan, Raymond F.

Neighbours, Jennifer C.

196

APPENDIX B. LIST OF TUCSON EMPLOYEES

O’Conner, Lynda M.

Scharlach, Werner W. G.

Oliver, Stacy

Schartman, Ethan D.

Olson, David B.

Schoknecht, William E.

Patt, Ferdinand

Schraml, Johann

Pauley, Robert G.

Schroeder, James C.

Payne, John

Schroeter, Darrell Frank

Perfetto, Antonio

Schuetz, Ernest Jr.

Peterson, John R.

Schwortz, Andria C.

Pickard, Matthew

Shaklan, Stuart

Pokorny, Martin

Shillue, William

Prestage, Richard

Shopbell, Patrick L.

Radford, Simon

Silver, Ronald

Rather, Elizabeth

Smith, John D.

Rather, John D.

Smith, Luke L.

Raymondson, Daisy

Snyder, Laura

Rector, Travis A.

Sparks, Calvin

Reiland, George Paul

Sperduti, Armand C.

Reimnitz, Jess Michael

Spuhler, Philipp

Reynolds, Donald

Stahl, Harry D.

Rhodes, Paul

Stevens, Robert L.

Rizzo, Joseph W.

Stobie, Elizabeth B.

Rosengard, Rebecca

Street, Oleta R.

Ross, Dewey E.

Stuart, Andrew

Roth, Ryan Matthews

Sullivan, Mark

Routt, Michael

Sullivan, Stanley

Ruiz, Timoteo

Tarr, Norman

Rupp, Phillip

Terrell, William D.

Saffle, James R.

Tester, Martin L.

Salter, Christopher J.

Teyssier, Edward M.

Scarl, James D.

Thomas, T. Maxine

197 Tietz, George A. Treiber, Edward F. Ulich, Bobby L. Urbain, Dennis Vaccari, Andrea Valente, Martin Valentine, Virginia Valladares, Geraldo Waddel, Matt Walker, Christopher K. Walker, Michael Weaver, John Webb, Dale A. Weller, Walter Wells, David Lee Wetmore, Eugene E. White, Terry White, Thomas Williams, Jonathan P. Wilson, Stewart K. Wolf, Grace Wolff, Scott E. Womeldorff, Ronald D. Wordeman, Matthew R. Wren, James A. Wright, Ronald Zhang, Oizhou

Appendix C

Glossary ALMA Atacama Large Millimeter[-wave] Array, a millimeter-wave telescope under construction at a 16,500-ft site in northern Chile, consisting of 64 12-m parabolic antennas. It will be operated by ESO and AUI, with participation by Chilean astronomers. AUI Associated Universities, Inc., a not-for-profit consortium of nine US universities that operates the NRAO for the NSF. AURA Associated Universities for Research in Astronomy, a large not-forprofit consortium that operates KPNO, CTIO, and the Space Telescope Science Institute. autocorrelator A digital device in which a stream of digital numbers (often, 1’s and 0’s) is correlated with itself. Used to make a spectral analysis of radio signals. back end A commonly used term for the electronic components used to detect and to analyze the output of a radio astronomy receiver. bolometer A device to measure the intensity of incident electromagnetic radiation over a defined bandwidth; in effect, a thermometer. Caltech California Institute of Technology located in Pasadena, California. coaxial cable An electrical cable consisting of a central wire surrounding by a dielectric insulator and wrapped by a wire mesh. Used to transmit low-frequency radio waves like those used for television and FM. CTIO Cerro Tololo Interamerican Observatory, an optical observatory operated by AURA and located near La Serena, Chile. 199

200

APPENDIX C. GLOSSARY

Dewar Essentially, a large thermos bottle used to hold cold liquified gases like nitrogen and oxygen. Named after its inventor. dish A radio antenna made with a parabolic surface. ESO European Southern Observatory, an optical observatory operated by a number of European countries, headquartered in Garching, Germany, with observing facilities in La Silla and Cerro Paranal, Chile. feed A device at the focus of a radio antenna that transfers the radio waves into the amplifier. filter bank An array of separately tuned, resonant-frequency filters arranged like the tines of a fork. Used to analyze the spectrum of an incoming radio signal. front end A commonly used term for the high frequency part of a radio astronomy receiver. Haystack Observatory The principal radio and radar astronomy observatory of MIT Lincoln Laboratory. It has a 120-ft radio telescope enclosed in a radome and is located near Westford, Massachusetts. IF The intermediate frequency part of a radio receiver. A mixer converts the frequency incoming radio waves to a lower, intermediate frequency that can be more easily amplified and manipulated. IRAM Institut de Radio Astronomie Millim``etrique, a joint millimeter-wave observatory of France, Germany, and Spain with main offices in Grenoble, France. It operates a 30-m millimeter-wave radio telescope near Granada, Spain, and a millimeter-wave interferometric telescope in the French alps. klystron An oscillator that produces a narrow-band, high-power radio signal and that can be tuned both mechanically and electronically. KPNO Kitt Peak National Observatory, the US national center for optical and infra-red astronomy, based in Tucson, Arizona. local oscillator The artificially generated radio signal injected into a radio receiver to convert the incoming radiation to a lower frequency. MIT Lincoln Laboratory A research laboratory of the Massachusetts Institute of Technology specializing in electronics and computers and principally funded by the US Air Force.

201 MPIfR Max-Planck-Insitut f¨ ffur Radioastronomie, the German institute for radio astronomy based in Bonn, Germany. mixer A non-linear electronic device in which two radio signals are “mixed” together. The outputs are two radio signals with frequencies corresponding to the sum and the difference of the those injected. In radio astronomy, the input frequencies are often the incoming radio signal and an artificial signal from a local oscillator. Often used as a radio receiver at extremely high frequencies. MWO Millimeter-Wave Observatory of the University of Texas, located in Fort Davis, Texas. NAIC National Astronomy and Ionospheric Center. Located near Arecibo, Puerto Rico and operated for the NSF by Cornell University, it operates a 1,000-ft diameter spherical telescope used for radio astronomy, radar astronomy, and ionospheric studies. NASA National Aeronautics and Space Administration, a federal agency with a mission to fund and promote exploration of “space.” This vague term denotes the realm outside of the Earth’s atmosphere, from the top of our atmosphere to the most remote objects in our universe. NRAO National Radio Astronomy Observatory, the US center for radio astronomy, based in Charlottesville, Virginia. NSF National Science Foundation, an agency of the US chartered to fund basic research and promote science education. oscillator In this book an electronic device to generate a radio wave of a specific frequency or wavelength. UKIRT United Kingdom Infrared Telescope, a telescope operated by Great Britain and located on the summit of Mauna Kea in Hawaii. UNIX A widely used computer language invented for large main frame computers but now available for personal computers under the name LINUX. VLA Very Large Array, a radio telescope located near Magdalena, New Mexico, operated by the NRAO. It consists of 27 25-m parabolic antennas mounted on railroad tracks.

202

APPENDIX C. GLOSSARY

VLBA Very Long Baseline Array operated by the NRAO, a continent-sized array of twenty-seven 25-m parabolic antennas mounted on railroad tracks and operated as a single radio telescope. See VLBI. VLBI Very Long Baseline Interferometry. A technique by which widely separated radio telescope are linked to provided extremely high angular resolution. Equipped with a super-accurate clock, each telescope would point at the same object and record the data and time marks on magnetic tape. Later, correlation of these tapes produces a crude image of the radio source with an angular resolution of micro arc-seconds. waveguide A metal pipe with a circular or, more usually, rectangular crosssection used to duct radio waves from one device to another. It is better suited for high-frequency waves than a coaxial cable or a wire because of much lower absorption loss.

Index Cardarella, D. J., 19, 27, 45 Carlstrom, J. E., 94 Carter, J., 36 Cheung, A. C., 30 Clark, B. G., 24, 38, 112 Clark, C. D., 71 Clarke, N., 112 Cochran, J., 55, 56 Cocke, W. J., 108 Cohen, M. H., 112, 113, 141 Conklin, E. E., 15, 19, 24, 30, 31, 42, 43, 45, 46, 54, 55, 61, 69, 82, 191 Cudaback, D. D., 127, 129 Cunningham, G., 66, 67

Abt, H., 169 Ade, P. A., 102 Albaugh, N. P., 19, 26, 27, 45, 59, 166 Albers, H., 51, 53, 54, 59, 61 Amthor, C., 133 Ashton, N. L., 5 Asrael, A., 58 Baars, J. W. M., 108 Baath, L., 176 Balister, M., 36, 40, 41 Bania, T. M., 40, 163 Barrett, A. H., 41, 141, 146, 173 Beckers, J. M., 145 Becklin, E. E., 145 Berkner, L. V., 2–5, 191 Bignell, R. C., 24 Bloch, E., 170, 185 Blum, E., 97 Bok, B. J., 2, 164 Bok, P. F., 164 Brown, H., 36, 37 Brown, R. L., 39, 68, 170, 173, 174, 176, 178, 182, 183, 192 Buhl, D., 26, 32 Burbidge, G., 64 Burke, B. F., 145 Burns, W. R., 84 Burton, W. B., 40, 41, 163

Dalgarno, A., 117 Davis, J. E., 55, 56, 97, 103, 153 DeConcini, D., 186, 187 Desmond, J. L., 171 de Zafra, R., 116 Doane, R., 52, 62, 163 Drake, F. D., 5–7, 9, 141–143, 191 Edrich, J., 99, 100 Ehnebuske, D. L., 24 Ekers, R., 170 Emberson, R. M., 5 Embry, L. A., 4 Emerson, D. T., 69, 72, 85–88, 103, 158, 163, 186, 192 England, J. M., 2

Callendar, F. J., 4 203

INDEX

204 Farris, A., 36 Field, G. B., 38, 144, 145 Findlay, J. W., 5, 13, 16, 17, 119, 120, 125, 130, 150, 151 Fitzner, J. T., 65 Freund, R. F., 24, 55, 56, 65, 78 Galey, M., 52 Gehrz, R. D., 142 Giacconi, R., 145 Gillett, F., 145 Giovanelli, R., 176 Goldberg, L., 51, 61, 64, 163, 165 Goldstein, R. M., 104 Goldwater, B. M., 163, 164 Gordon, J. B., 180–182 Gordon, M. A., 27, 29, 38, 41, 56, 62, 63, 65, 66, 69, 125, 166, 191 Gottesman, S. T., 38 Gould, N., 181 Gould, R. D., 181 Grayzeck, E. J., 141, 142 Greenhalgh, J. P., 36 Greenstein, J. L., 2 Grove, G., 19, 27, 32 Gust, W., 45, 54, 57 Hoglund, ¨ B., 39 Haas, R. W., 108 Hagen, J. P., 2 Hardy, E., 182 Harwood, W. F., 4 Haslam, C. G. T., 87–89 Haynes, M. P., 170 Heeschen, D. S., 2, 5, 12, 13, 27, 35, 39, 41, 42, 51–54, 79, 111, 113, 116, 118, 125, 141, 144, 165, 169, 170, 191

Heiles, C., 142 Hill, R., 56 Hjellming, R. M., 21, 24, 35, 175 Hockenberry, H., 5 Hoff, W. J., 4 Hogg, D. E., 59, 116, 145, 177, 192 Holdaway, M., 183 Hollis, J. M., 32, 45, 81–83, 112, 119 Horne, W. G., 16, 149, 150, 154 Howard, W. E. III, 40, 111, 140, 141, 144, 162, 186 Hudson, J., 24, 83 Hughes, R. E., 70, 143, 167, 168 Huguenin, G. R., 142 Humphreys, R. M., 145 Hungerbuhler, H., 15, 17 Hunt, D., 161 Hvatum, H., 18, 19, 21, 37, 39, 41, 42, 46, 65, 84, 130, 136, 140, 145, 147, 149, 150 Jefferies, J. T., 130, 140 Jefferts, K. B., 30, 32 Jewell, P. R., 24, 69, 88, 112, 158 Johnson, D., 142 Kassim, M., 88 Kellermann, K. I., 141 King, L. J., 125, 130, 131, 149, 150 Kingsley, J. S., 24 Kitchen, E., 19, 21, 24, 79 Klein, U., 87 Kolbe, J., 186 Kraus, J. D., 2 Kreysa, E., 103 Kusunoki, A., 175 Kutner, M. L., 40, 108 Lada, C. J., 146 Lamb, J. W., 99

INDEX Leighton, R. B., 142, 144, 147, 149 Lin, C. C., 164 Lipscomb, C. E., 45 Liszt, H. S., 86, 145 Lo, K.-Y., 171, 189 Lo, K.-Y. (Fred), 192 Lockman, F. J., 39, 40, 163 Logan, H. D., 19, 27 Low, F. J., 5, 7–10, 13, 18, 19, 101, 177, 191 Lynds, B. T., 52 Mangum, J. G., 65 Marymor, J., 52, 63, 67 Masumoto, H. S., 140 Matayoshi, H., 137 Matayoshi, M., 140 Matsuda, F., 140 Mayall, N. U., 10, 12, 27 Mayer, C. E., 153 McBrian, J. E., Jr, 45 McCray, R., 145 Menon, R., 10 Menzel, D. H., 2 Mezger, P. G., 10, 35 Mikulski, B., 187 Minkowski, R., 2 Mitchell, J. M., 4 Moffet, A. T., 142 Moore, C. E., 191 Moore, C. H., 55, 80, 82 Morris, M., 116, 117 Mull, M., 139 Myers, D., 45, 61 Napier, P. J., 178, 183 Nay, G. A., 5 Neugebauer, G., 144 Nolt, I.G., 103 Osterbrock, D. E., 142, 145

205 Otarola ´ M., A., 176, 178 Owen, F. N., 173, 174, 178, 183 Owen, T., 161 Palmer, P., 116, 117, 146 Pauliny-Toth, I. I. K., 35 Payne, J. M., 26, 38, 55, 56, 65, 94, 119, 147 Payne-Geposhkin, C., 2 Penzias, A. A., 30, 32, 41, 117 Perry, G. M., 76, 130, 140 Pesch, P., 145 Plasch, G., 140 Purcell, E. M., 2 Quintana, H., 176 Rabb, J., 52 Radford, S. J. E., 178 Radostitz, J., 103 Raffin, P., 175 Rather, E. D., 31, 45, 55, 81, 82 Rather, J. D., 32, 45, 69 Rhodes, P. J., 45, 50, 56, 57, 117, 154 Rickard, L. J., 116, 117 Rieke, G. H., 143 Riffe, T. R., 51, 52, 67, 143, 171 Roberts, M. S., 39, 53, 83, 137, 143–145, 147, 149, 169, 170, 174, 192 Ross, D. E., 7, 8, 32, 45, 56, 59 Routt, M., 56 Ruttenbery, C. B., 4 Ruze, J., 142 Rydbeck, O., 39 Salter, C. J., 88 Schaefer, J. P., 165 Scharlach, W. W. G., 45, 117 Schraml, J. B., 19, 79

INDEX

206 Schwartz, P. R., 32 Scoville, N. Z., 116, 117 Silver, R., 56 Smith, S. C., 16 Snyder, L. E., 26, 27, 30, 32, 112, 116, 142, 146 Solomon, P., 116–118, 142 Sorochenko, R. L., 29 Sparks, C., 45 Sperduti, A. C., 45, 49 Sramek, R. A., 183 Stevens, R. L., 76 Stobie, E., 56, 83, 84, 166 Stratton, J., 2 Strittmatter, P. A., 65, 67, 68 Struve, O., 191 Sullivan, S., 56, 158, 167 Sunderlin, C. E., 4

von Hoerner, S., 38, 125, 130, 143

Taylor, J. H., Jr., 145 Terrell, W. D., 19, 27 Tester, M. L., 45 Thaddeus, P., 142, 169, 174 Thomas, T. M., 45, 54, 56, 82 Thompson, R. I., 65 Townes, C. H., 30, 31, 116, 142 Turner, B. E., 117, 125, 129, 145 Tuve, M. A., 2, 3

Ziurys, L. M., 186

Ulich, B. L., 32, 45, 53, 73, 108, 130 Uson, J., 178 van de Hulst, H. C., 29 Vanden Bout, P. A., 70, 142, 170, 171, 176, 178, 185, 186, 188, 189, 192 Vandenberg, N. R., 24 Varian, R., 94 Varian, S., 94 Villadares, G., 178

Wade, C. M., 125, 130, 174 Wang, W.-S., 164 Waterman, A. T., 3, 4 Webb, D. A., 54–56, 62, 63, 65–67, 69–71, 73, 125, 130, 147, 164, 166, 168, 186 Weinreb, S., 30, 41, 55, 80, 91, 97, 103, 104 Weller, W., 19, 79 Wetmore, E., 122 Whipple, F. L., 2 Wiesner, J. B., 2 Wilson, R. W., 30, 32, 42, 80, 142, 146 Wilson, W. J., 32 Wong, W.-Y., 125, 130

Bibliography [1] Planning document for the establishment and operation of a radio astronomy observatory. Report to the National Science Foundation, Associated Universities, Inc., 1956. [2] J. M. England. A Patron for Pure Science, The National Science Foundation’s Formative Years, 1945–57. 7 National Science Foundation, Washington, D. C., 1982. [3] D. S. Heeschen. Draft of a Talk on the History of the NRAO, 6 May 1996. Personal Communication to M. A. Gordon. [4] B. K. Malphrus. The History of Radio Astronomy and the National Radio Astronomy Observatory: Evolution Toward Big Science. Krieger Publishing Co., Melborne, Florida, 1996. [This book contains errors.]. [5] F. D. Drake. Telephone conversation with M. A. Gordon, July 1987. [6] F. D. Drake. Email to M. A. Gordon, 25 April 2004. [7] F. J. Low. Conversation with M. A. Gordon, July 1987. [8] G. Grove. 36-[f]t [t]elescope. The [NRAO] Observer, 9(2):22, November 1969. [9] K. Ross. Telephone conversation with M. A. Gordon, 18 May 2004. [10] R. Jastrow and M. H. Thompson. Astronomy: Fundamentals and Frontiers. John Wiley and Sons, second edition, 1974. See pages 52ff. [11] P. G. Mezger. Principal Considerations of Radio Astronomy Observations at Very High Frequencies. Unnumbered NRAO technical report, National Radio Astronomy Observatory, March 1964. 207

208

BIBLIOGRAPHY

[12] R. Menon, N. Albaugh, and D. Logan. Performance of Narrowband 9mm Radiometer. Electronics Division Internal Report 60, National Radio Astronomy Observatory, 1967. [13] J. E. McDonald. Cloudiness over the Southwestern United States and Its Relation to Astronomical Observing. Technical Report 7, Institute of Atmospheric Physics, University of Arizona, 1958. [14] J. W. Findlay. Telephone conversation with M. A. Gordon, July 1987. [15] D. S. Heeschen. Email to M. A. Gordon, June 2003. [16] M. A. Gordon and G. S. Musser. Cold Heart of the Cosmos. Mercury, 26(1):14, 1997. [17] E. K. Conklin. Email to M. A. Gordon, March 2003. [18] S. C. Smith. Email to M. A. Gordon, 13 May 2004. [19] J. W. M. Baars. Conversation with M. A. Gordon, March 2003. [20] George Behrens. Telephone conversation with M. A. Gordon, March 2003. [21] D. E. Ross. The History of the 36-foot. The [NRAO] Observer, 20(4):3, December 1979. [The article actually says 1968, but I think this is a typographical error.]. [22] J. Schraml. Email to M. A. Gordon, March 2003. [23] Barry Clark. Email to M. A. Gordon, 19 June 2003. [24] L. E. Snyder and D. Buhl. Detection of Possible Maser Emission Near 3.48 Millimeters From an Unidentified Molecular Species in Orion. Ap. J., 189:L31, 1974. [25] D. Hartmann and W. B.Burton. Cambridge, 1997. Minutes of the 1944 meeting and their English translation. [26] H. C. van de Hulst. Radio Waves from Space. Naturkunde, 11:230, 1945.

Ned. Tidjschr.

[27] H. J. Ewen and E. M. Purcell. Radiation from Galactic Hdrogen at 1,420 Mc/s. Nature, 168:356, 1951.

BIBLIOGRAPHY

209

[28] C. A. Muller and J. H. Oort. The Interstellar Hydrogen Line at 1,420 Mc/s and an Estimate of Galactic Rotation. Nature, 168:357, 1951. [29] M. A. Gordon and R. L. Sorochenko. Radio Recombination Lines, Their Physics and Astronomical Applications. Kluwer Academic Publishers, Dordrecht, 2002. [30] C. H. Townes. In H. C. van de Hulst, editor, Radio Astronomy, page 92, Cambridge, 1964. I. A. U. Symp. No. 4, Cambridge Univ. Press. [31] S. Weinreb, A. H. Barrett, M. L. Meeks, and J. C. Henry. Nature, 200:829, 1963. [32] A. C. Cheung, D. Rank, C. H. Townes, , D. D. Thornton, and W. J. Welch. Detection of NH3 Molecules in the Interstellar Medium by their Microwave Emission. Phys. Rev. Lett, 21:1701, 1968. [33] A. C. Cheung, D. Rank, C. H. Townes, D. D. Thornton, and W. J. Welch. Detection of Water in Interstellar Regions by its Microwave Radiation. Nature, 221:626, 1969. [34] L. E. Snyder, D. Buhl, B. Zuckerman, and P. Palmer. Microwave Detection of Interstellar Formaldehyde. Phys. Rev. Lett., 22:679, 1969. [35] R. W. Wilson, K. B. Jefferts, and A. A. Penzias. Carbon Monoxide in the Orion Nebula. Ap. J., 161:L43, 1970. [36] M. A. Gordon and M. L. Meeks. Observation of the 94α Hydrogen and 94α Helium Emission Lines in the Orion Nebula. Ap. J., 149:L21, 1967. [37] J. A. Ball. Measurements with Radio-Frequency Spectrometers. In M. L. Meeks, editor, Methods of Experimental Physics, volume 12C, page 46. Academic Press, 1976. See ‘Acknowledgements’ on p 57. [38] S. T. Gottesman and M. A. Gordon. Radio Recombination Lines from Diffuse Interstellar Gas. Ap. J., 162:L93, 1970. [39] M. A. Gordon and S. T. Gottesman. Electron Density and Temperature in the Diffuse Interstellar Medium Determined from Radio Recombination Lines. Ap. J., 168:361, 1971. [40] G. B. Field, D. W. Goldsmith, and H. J. Habing. Cosmic-Ray Heating of the Interstellar Gas. Ap. J., 155:L149, 1969.

210

BIBLIOGRAPHY

[41] M. A. Gordon and M. S. Roberts. The Absense of Formaldehyde Radiation toward Cold Regions of the Galactic Plane. Ap. J., 170:277, 1971. [42] M. A. Gordon and T. Cato. A Longitude Survey of Radio Recombination Lines from the Diffuse Interstellar Medium. Ap. J., 176:587, 1972. [43] T. Bania, W. B. Burton, M. A. Gordon, and F. J. Lockman. LargeScale Distribution of Carbon Monoxide in the Galaxy. Bull. Amer. Astron. Soc., 7:266, 1975. [44] W. B.Burton, M. A. Gordon, T. M. Bania, and F. J. Lockman. The Overall Distribution of Carbon Monoxide in the Plane of the Galaxy. Ap. J., 202:30, 1975. [45] M. A. Gordon and W. B. Burton. Carbon Monoxide in the Galaxy. i - The radial distribution of CO, H2 , and nucleons. Ap. J., 208:346, 1976. [46] W. B. Burton and M. A. Gordon. Carbon Monoxide in the Galaxy. II. The Thickness of the Galactic CO Layer. Ap. J., 207:L189, 1976. [47] W. B. Burton and M. A. Gordon. Carbon Monoxide in the Galaxy. III. The Overall Nature of its Distribution in the Equatorial Plane. Ast. Ap., 63:7, 1978. [48] M. A. Gordon and W. B. Burton. Carbon Monoxide in the Galaxy. Sci. Amer., 240:54, 1979. [49] D. S. Heeschen. Letter to Arno Penzias, 18 March 1969. [50] D. E. Hogg. Letter to Leo Goldberg, July 1975. [51] H. R. Albers. Letter to Daniel Hunt of the NSF, 30 September 1975. [52] M. A. Gordon. Letter to Leo Goldberg, April 1976. [53] M. A. Gordon. Letter to Leo Goldberg, October 1976. [54] D. A. Webb. NRAO Tucson purchase order T07771, August 1991. [55] D. A. Webb. Conversation with M. A. Gordon, March 2003. [56] Contract between AURA and TRICO, Amendment 1, Paragraph 7, 1 March 1966.

BIBLIOGRAPHY

211

[57] D. A. Webb. Conversation with M. A. Gordon, March 2004. [58] C. H. Moore and G. C. Leach. A Language for Interactive Computing. Internal publication, Mohasco Industries, 1970. [59] C. H.Moore. The Evolution of FORTH, an Uusual Language. Byte, August:76, 1980. [60] E. D. Rather, C. H. Moore, and J. M. Hollis. Basic Principles of FORTH Language as Applied to a PDP-11 Computer. Computer Division Internal Report 17, National Radio Astronomy Observatory, 1974. [61] E. Stobie. Conversation with M. A. Gordon, March 2003. [62] D. T. Emerson. Conversation with M. A. Gordon, March 2004. [63] T. Forveille, S. Guilloteau, and R. Lucas. Class: Continuum and Line Analysis Single-dish Software. IRAM Internal Report, 1990. [64] M. A. Gordon. Dust Emission as a Quantitative Probe of Star-Forming Regions. Ast. Ap., 301:853, 1995. [65] C. G. T. Haslam. NOD2 A General System of Analysis for Radioastronomy. Ast. Ap., 15, 1974. [66] D. T. Emerson, U. Klein, and C. G. T. Haslam. A Multiple Beam Technique for Overcoming Atmospheric Limitations to Single-Dish Observations of Extended Radio Sources. Ast. Ap., 76, 1992. [67] J. M. Payne. The Tucson Electronics Group. The [NRAO] Observer, 20(4):25–31, December 1979. [68] S. Weinreb. A Digital Spectral Analysis Technique and its Application to Radio Astronomy. PhD thesis, Massachusetts Institute of Technology, 1963. Also available as Technical Report 412 of MIT’s Research Laboratory of Electronics. [69] B. L. Ulich and R. W. Haas. Absolute Calibration of MillimeterWavelength Spectral Lines. Ap. J. Suppl., 30:247, 1976. [70] M. L. Kutner and B. L. Ulich. Recommendations for Calibration of Millimeter-Wavelength Spectral Line Data. Ap. J., 250:341, 1981.

212

BIBLIOGRAPHY

[71] M. A. Gordon, J. W. M. Baars, and W. J. Cocke. Observations of radio lines from unresolved sources: Telescope coupling, doppler effects, and cosmological corrections. Ast. Ap., 264:337, 1992. [72] D. S. Heeschen. Letter to all Users of the 36-ft Telescope, 1 May 1978. [73] M. Morris. Email to M. A. Gordon, 22 September 2004. [74] B. E. Turner. Email to M. A. Gordon, September 2004. [75] P.M. Solomon and R. de Zafra. Carbon Monoxide in External Galaxies. Ap. J., 199:L79, 1975. [76] L. J. Rickard, P. Palmer, M. Morris, B. Zuckerman, and B. E. Turner. Detection of Extragalactic Carbon Monoxide at Millimeter Wavelengths. Ap. J., 199:L75, 1975. See the acknowledgement on page L77. [77] J. M. Payne, J. M. Hollis, and J. W. Findlay. New Method of Measuring the Shape of Precise Antenna Reflectors. Rev. Sci. Instr., 47:50, 1976. [78] J. M. Hollis. Email to Mark Gordon, July 2003. [79] S. von Hoerner. Design of large steerable antennas. Astron. J., 72:35, 1967. [80] G. Perry. Cable car installations. Technical Report 25-m Telescope Memorandum 23, National Radio Astronomy Observatory, 4 June 1975. [81] A 25-meter telescope for millimeter wavelengths. Internal Report of the National Radio Astronomy Observatory, September 1975. [82] A 25-meter telescope for millimeter wavelengths. Internal Report of the National Radio Astronomy Observatory, Volume II, July 1977. [83] J. T. Jefferies. Minutes of the Mauna Kea Users’ Committee, 22 April 1980, July 1980. [84] W. E. Howard. Telephone conversation with M. A. Gordon, 12 April 2004. [85] Report of the Subcommittee on Millimeter-Wave Facilities of the Advisory Committee for Astronomical Sciences, A. H. Barrett, Chairman. National Science Foundation, 1979.

BIBLIOGRAPHY

213

[86] Astronomy and astrophysics for the 1980s/ Astronomy Survey Committee, Assembly of Mathematical and Physical Sciences, National Research Council. National Academy Press, 1982. [Known as the “Field Report”]. [87] Report of the Subcommittee on Millimeter-and SubmillimeterWavelength Astronomy, A. H. Barrett, Chairman. National Science Foundation, 1983. [88] M. A. Gordon. A new surface for the 36-ft telescope? Memorandum to M. S. Roberts, 17 April 1980. [89] W. G. Horne. Upgrading the 36’ reflector. Memorandum to Hein Hvatum, 9 January 1981. [90] M. A. Gordon. Resurface the 36-ft telescope? Memorandum to M. Balister, J. Findlay, T. Hamed, B. Horne, R. Howard, H. Hvatum, L. King, J. Payne, G. Perry, P. Rhodes, D. Webb, and S. Weinreb, 26 November 1980. [91] M. A. Gordon. A New Surface for an Old Scope. Sky and Telescope, 64(4):326, 1984. [92] D. T. Emerson and P. R. Jewell. Contribution to the NRAO Newsletter, June 1990. [93] F. J. Lockman. Email to M. A. Gordon, March 2004. [94] T. Bania. Conversation with, and Email to, M. A. Gordon, 19 March 2004. [95] Associated Universities, Inc. The Millimeter Array, July 1990. [96] M. A. Gordon. Are we thinking boldly enough? Millimeter-Wave Array Memo Series 25, National Radio Astronomy Observatory, 1 October 1984. [97] P. Raffin and A. Kusunoki. Searching for Submm Sites in Chile. Technical Report SMA Technical Memorandum 59, Smithsonian Astrophysical Observatory, 14 May 1992. [98] A. Otar´ o´la M. Site Survey at Chajnantor for the ALMA Project. Unpublished paper, November 2000. IAU Technical Workshop entitled Astronomical Site Evaluation in the Visible and Radio Range, held in Marrakech, Morocco.

214

BIBLIOGRAPHY

[99] A. Otar´ ola M. Conversation with M. A. Gordon, 23 April 2004. [100] R. L. Brown. Email to M. A. Gordon, 24 April 2004. [101] N. B. Foster, D. T. Volz, and L. W. Foskett. A Spectral Hygrometer for Measuring Total Preciptable Water. Unpublished manuscript of the US Weather Bureau, June 1963. [102] M. A. Gordon. Operation of the MMA in Chile. Technical report, National Radio Astronomy Observatory, 1 June 1995. [103] M. A. Gordon. ALMA in Chile, A Plan for Operations and Site construction. Technical report, National Radio Astronomy Observatory, 30 May 2000. Modified slightly 3 August 2000. [104] M. A. Gordon and R. L. Sorochenko. Radio Recombination Lines, Their Physics and Astronomical Applications. Kluwer Academic Publishers, 2002. ISBN 1-4020-1016-8. [105] L. M. Ziurys. Email to M. A. Gordon, March 2004. [106] Congressional Record—Senate, Page S 14303, 27 October 1989.

Astrophysics and Space Science Library Volume 316: Civic Astronomy - Albany's Dudley Observatory, 1852-2002, by G. Wise Hardbound ISBN 1-4020-2677-3, October 2004 Volume 315: How does the Galaxy Work - A Galactic Tertulia with Don Cox and Ron Reynolds, edited by E. J. Alfaro, E. Pérez, J. Franco Hardbound ISBN 1-4020-2619-6, September 2004 Volume 314: Solar and Space Weather Radiophysics - Current Status and Future Developments, edited by D.E. Gary and C.U. Keller Hardbound ISBN 1-4020-2813-X, August 2004 Volume 313: Adventures in Order and Chaos – A Scientific Autobiography, by G. Contopoulos Hardbound ISBN 1-4020-3039-8, December 2004 Volume 312: High-Velocity Clouds, edited by H. van Woerden, U. Schwarz, B. Wakker Hardbound ISBN 1-4020-2813-X, September 2004 Volume 311: The New ROSETTA Targets- Observations, Simulations and Instrument Performances, edited by L. Colangeli, E. Mazzotta Epifani, P. Palumbo Hardbound ISBN 1-4020-2572-6, September 2004 Volume 310: Organizations and Strategies in Astronomy 5, edited by A. Heck Hardbound ISBN 1-4020-2570-X, September 2004 Volume 309: Soft X-ray Emission from Clusters of Galaxies and Related Phenomena, edited by R. Lieu and J. Mittaz Hardbound ISBN 1-4020-2563-7, September 2004 Volume 308: Supermassive Black Holes in the Distant Universe, edited by A.J. Barger Hardbound ISBN 1-4020-2470-3, August 2004 Volume 307: Polarization in Spectral Lines, by E. Landi Degl’Innocenti and M. Landolfi Hardbound ISBN 1-4020-2414-2, August 2004 Volume 306: Polytropes – Applications in Astrophysics and Related Fields, by G.P. Horedt Hardbound ISBN 1-4020-2350-2, September 2004 Volume 305: Astrobiology: Future Perspectives, edited by P. Ehrenfreund, W.M. Irvine, T. Owen, L. Becker, J. Blank, J.R. Brucato, L. Colangeli, S. Derenne, A. Dutrey, D.

Despois, A. Lazcano, F. Robert Hardbound ISBN 1-4020-2304-9, July 2004 Paperback ISBN 1-4020-2587-4, July 2004 Volume 304: Cosmic Gammy-ray Sources, edited by K.S. Cheng and G.E. Romero Hardbound ISBN 1-4020-2255-7, September 2004 Volume 303: Cosmic rays in the Earth’s Atmosphere and Underground, by L.I, Dorman Hardbound ISBN 1-4020-2071-6, August 2004 Volume 302:Stellar Collapse, edited by Chris L. Fryer Hardbound, ISBN 1-4020-1992-0, April 2004 Volume 301: Multiwavelength Cosmology, edited by Manolis Plionis Hardbound, ISBN 1-4020-1971-8, March 2004 Volume 300:Scientific Detectors for Astronomy, y edited by Paola Amico, James W. Beletic, Jenna E. Beletic Hardbound, ISBN 1-4020-1788-X, February 2004 Volume 299: Open Issues in Local Star Fomation, edited by Jacques Lépine, Jane Gregorio-Hetem Hardbound, ISBN 1-4020-1755-3, December 2003 Volume 298: Stellar Astrophysics - A Tribute to Helmut A. Abt, t edited by K.S. Cheng, Kam Ching Leung, T.P. Li Hardbound, ISBN 1-4020-1683-2, November 2003 e by Leonty I. Miroshnichenko Volume 297: Radiation Hazard in Space, Hardbound, ISBN 1-4020-1538-0, September 2003 Volume 296: Organizations and Strategies in Astronomy, volume 44, edited by André Heck Hardbound, ISBN 1-4020-1526-7, October 2003 Volume 295: Integrable Problems of Celestial Mechanics in Spaces of Constant Curvature, by T.G. Vozmischeva Hardbound, ISBN 1-4020-1521-6, October 2003 Volume 294: An Introduction to Plasma Astrophysics and Magnetohydrodynamics, by Marcel Goossens Hardbound, ISBN 1-4020-1429-5, August 2003 Paperback, ISBN 1-4020-1433-3, August 2003

Volume 293: Physics of the Solar System, by Bruno Bertotti, Paolo Farinella, David Vokrouhlický Hardbound, ISBN 1-4020-1428-7, August 2003 Paperback, ISBN 1-4020-1509-7, August 2003 Volume 292: Whatever Shines Should Be Observed, d by Susan M.P. McKenna-Lawlor Hardbound, ISBN 1-4020-1424-4, September 2003 Volume 291: Dynamical Systems and Cosmology, by Alan Coley Hardbound, ISBN 1-4020-1403-1, November 2003 Volume 290: Astronomy Communication, edited by André Heck, Claus Madsen Hardbound, ISBN 1-4020-1345-0, July 2003 Volume 287/8/9: The Future of Small Telescopes in the New Millennium, edited by Terry D. Oswalt Hardbound Set only of 3 volumes, ISBN 1-4020-0951-8, July 2003 Volume 286: Searching the Heavens and the Earth: The History of Jesuit Observatories, by Agustín Udías Hardbound, ISBN 1-4020-1189-X, October 2003 Volume 285: Information Handling in Astronomy - Historical Vistas, edited by André Heck Hardbound, ISBN 1-4020-1178-4, March 2003 Volume 284: Light Pollution: The Global View, w edited by Hugo E. Schwarz Hardbound, ISBN 1-4020-1174-1, April 2003 Volume 283: Mass-Losing Pulsating Stars and Their Circumstellar Matter, r edited by Y. Nakada, M. Honma, M. Seki Hardbound, ISBN 1-4020-1162-8, March 2003 Volume 282: Radio Recombination Lines, by M.A. Gordon, R.L. Sorochenko Hardbound, ISBN 1-4020-1016-8, November 2002 Volume 281: The IGM/Galaxy Connection, edited by Jessica L. Rosenberg, Mary E. Putman Hardbound, ISBN 1-4020-1289-6, April 2003 Volume 280: Organizations and Strategies in Astronomy III, I edited by André Heck Hardbound, ISBN 1-4020-0812-0, September 2002

Volume 279: Plasma Astrophysics , Second Edition, by Arnold O. Benz Hardbound, ISBN 1-4020-0695-0, July 2002 Volume 278: Exploring the Secrets of the Aurora, by Syun-Ichi Akasofu Hardbound, ISBN 1-4020-0685-3, August 2002 r by Arnold Hanslmeier Volume 277: The Sun and Space Weather, Hardbound, ISBN 1-4020-0684-5, July 2002 Volume 276: Modern Theoretical and Observational Cosmology, y edited by Manolis Plionis, Spiros Cotsakis Hardbound, ISBN 1-4020-0808-2, September 2002 Volume 275: History of Oriental Astronomy, y edited by S.M. Razaullah Ansari Hardbound, ISBN 1-4020-0657-8, December 2002 Volume 274: New Quests in Stellar Astrophysics: The Link Between Stars and Cosmology, y edited by Miguel Chávez, Alessandro Bressan, Alberto Buzzoni,Divakara Mayya Hardbound, ISBN 1-4020-0644-6, June 2002 Volume 273: Lunar Gravimetry, y by Rune Floberghagen Hardbound, ISBN 1-4020-0544-X, May 2002 Volume 272:Merging Processes in Galaxy Clusters, edited by L. Feretti, I.M. Gioia, G. Giovannini Hardbound, ISBN 1-4020-0531-8, May 2002 Volume 271: Astronomy-inspired Atomic and Molecular Physics, by A.R.P. Rau Hardbound, ISBN 1-4020-0467-2, March 2002 Volume 270: Dayside and Polar Cap Aurora, by Per Even Sandholt, Herbert C. Carlson, Alv Egeland Hardbound, ISBN 1-4020-0447-8, July 2002 Volume 269: Mechanics of Turbulence of Multicomponent Gases, by Mikhail Ya. Marov, Aleksander V. Kolesnichenko Hardbound, ISBN 1-4020-0103-7, December 2001 Volume 268: Multielement System Design in Astronomy and Radio Science, e by Lazarus E. Kopilovich, Leonid G. Sodin Hardbound, ISBN 1-4020-0069-3, November 2001

Volume 267: The Nature of Unidentified Galactic High-Energy Gamma-Ray Sources, edited by Alberto Carramiñana, Olaf Reimer, David J. Thompson Hardbound, ISBN 1-4020-0010-3, October 2001 Volume 266: Organizations and Strategies in Astronomy II, I edited by André Heck Hardbound, ISBN 0-7923-7172-0, October 2001 Volume 265: Post-AGB Objects as a Phase of Stellar Evolution, edited by R. Szczerba, S.K. Górny Hardbound, ISBN 0-7923-7145-3, July 2001 Volume 264: The Influence of Binaries on Stellar Population Studies, edited by Dany Vanbeveren Hardbound, ISBN 0-7923-7104-6, July 2001 Volume 262: Whistler Phenomena - Short Impulse Propagation, by Csaba Ferencz, Orsolya E. Ferencz, Dániel Hamar, János Lichtenberger Hardbound, ISBN 0-7923-6995-5, June 2001 Volume 261: Collisional Processes in the Solar System, edited by Mikhail Ya. Marov, Hans Rickman Hardbound, ISBN 0-7923-6946-7, May 2001 Volume 260: Solar Cosmic Rays, by Leonty I. Miroshnichenko Hardbound, ISBN 0-7923-6928-9, May 2001 Volume 259: The Dynamic Sun, edited by Arnold Hanslmeier, Mauro Messerotti, Astrid Veronig Hardbound, ISBN 0-7923-6915-7, May 2001 Volume 258: Electrohydrodynamics in Dusty and Dirty Plasmas- GravitoElectrodynamics and EHD, by Hiroshi Kikuchi Hardbound, ISBN 0-7923-6822-3, June 2001 Volume 257: Stellar Pulsation - Nonlinear Studies, edited by Mine Takeuti, Dimitar D. Sasselov Hardbound, ISBN 0-7923-6818-5, March 2001 Volume 256: Organizations and Strategies in Astronomy, y edited by André Heck Hardbound, ISBN 0-7923-6671-9, November 2000 Volume 255: The Evolution of the Milky Way- Stars versus Clusters, edited by Francesca Matteucci, Franco Giovannelli Hardbound, ISBN 0-7923-6679-4, January 2001

Volume 254: Stellar Astrophysics, edited by K.S. Cheng, Hoi Fung Chau, Kwing Lam Chan, Kam Ching Leung Hardbound, ISBN 0-7923-6659-X, November 2000 Volume 253: The Chemical Evolution of the Galaxy, by Francesca Matteucci Paperback, ISBN 1-4020-1652-2, October 2003 Hardbound, ISBN 0-7923-6552-6, June 2001 Volume 252: Optical Detectors for Astronomy II, I edited by Paola Amico, James W. Beletic Hardbound, ISBN 0-7923-6536-4, December 2000 Volume 251: Cosmic Plasma Physics, by Boris V. Somov Hardbound, ISBN 0-7923-6512-7, September 2000 Volume 250: Information Handling in Astronomy, edited by André Heck Hardbound, ISBN 0-7923-6494-5, October 2000 Volume 249: The Neutral Upper Atmosphere, by S.N. Ghosh Hardbound, ISBN 0-7923-6434-1, July 2002 Volume 247: Large Scale Structure Formation, edited by Reza Mansouri, Robert Brandenberger Hardbound, ISBN 0-7923-6411-2, August 2000 Volume 246: The Legacy of J.C. Kapteyn, edited by Piet C. van der Kruit, Klaas van Berkel Paperback, ISBN 1-4020-0374-9, November 2001 Hardbound, ISBN 0-7923-6393-0, August 2000 Volume 245: Waves in Dusty Space Plasmas, by Frank Verheest Paperback, ISBN 1-4020-0373-0, November 2001 Hardbound, ISBN 0-7923-6232-2, April 2000 Volume 244: The Universe, edited by Naresh Dadhich, Ajit Kembhavi Hardbound, ISBN 0-7923-6210-1, August 2000 Volume 243: Solar Polarization, edited by K.N. Nagendra, Jan Olof Stenflo Hardbound, ISBN 0-7923-5814-7, July 1999 Volume 242: Cosmic Perspectives in Space Physics, by Sukumar Biswas Hardbound, ISBN 0-7923-5813-9, June 2000

Volume 241: Millimeter-Wave Astronomy: Molecular Chemistry & Physics in Space, edited by W.F. Wall, Alberto Carramiñana, Luis Carrasco, P.F. Goldsmith Hardbound, ISBN 0-7923-5581-4, May 1999 Volume 240: Numerical Astrophysics, edited by Shoken M. Miyama, Kohji Tomisaka,Tomoyuki Hanawa Hardbound, ISBN 0-7923-5566-0, March 1999 Volume 239: Motions in the Solar Atmosphere, edited by Arnold Hanslmeier, Mauro Messerotti Hardbound, ISBN 0-7923-5507-5, February 1999 Volume 238: Substorms-4, edited by S. Kokubun, Y. Kamide Hardbound, ISBN 0-7923-5465-6, March 1999

For further information about this book series we refer you to the following web site: www.springeronline.com To contact the Publishing Editor for new book proposals: Dr. Harry (J.J.) Blom: [email protected]