Interview with Alan T. Moffet

Sullivan  00:01

Just talking with Al Moffet at Groningen on 11 August, '78. Can you tell me now when you first came in contact with radio astronomy and under what circumstances was that?

Moffet 00:14

I guess my first contact with radio astronomy was while I was still an undergraduate, I went to an IEEE, which was then IRE, meeting in New York City, and heard Ron Bracewell give a talk about his radioheliograph, which he called microsteradian beam. And he impressed me with elegance of that instrument.

Sullivan  00:36

This was--

Moffet 00:37

This while I was an undergraduate at Wesley and majoring in physics and doing some work in nuclear physics.

Sullivan  00:42

And he was talking about the Stanford--

Moffet 00:43

Stanford 10 cm crisscross. I wasn't really quite sure as I became a graduate student whether I wanted to do physics or to do astronomy. I had a strong bent for electronics. And I applied to Stanford and Caltech and perhaps MIT, I don't now recall whether I ever finished that application. And somewhere along about the time it was time to make acceptances, I wrote letters to Caltech and to Stanford saying that while I was accepted in physics there, I wasn't really sure about what I might want to do, radio astronomy, and that I understood that each of them was doing some radio astronomy, was it possible to do that as a physicist? And from Stanford, I got back a rather curt letter from Felix Bloch, who was chairman of the physics department, saying, "Well, radio astronomy here is done in electrical engineering. And if you want to do radio astronomy, you'll have to reapply in electrical engineering."

Sullivan  01:42

Right. It was too late already.

Moffet 01:44

From Bacher, I got back a letter saying, "Sure, come on ahead. We've had people do astronomy who are physicists and astronomers do physics. And so long as it's good science, we don't really care what you call it." So, I made the obvious choice.

Sullivan  02:00

And so, when did you get to Caltech, then?

Moffet 02:02

In the fall of 1957.

Sullivan  02:05

And was there any radio astronomy going on, in fact?

Moffet 02:08

Yes. By that time. Although in my first winter, I was busy taking courses and I really didn't have anything to do with the astronomy department except to go to seminars. But during that time they had a 32 ft antenna in Palomar and were doing 21 cm - a 21 cm survey of the northern Milky Way, which did eventually get published. By the time that spring came around and I was looking for summer employment, I went over to see John Bolton. In fact, I found him pulling out from a mold one section of a fiberglass surface for that 32 ft dish. It was to be resurfaced with a cast fiberglass surface, which was supposed to make it good up to fairly short wavelengths. A lot of effort was spent on that 32 ft dish, moving it from Palomar to the Owens Valley and setting it up. And in the Owens Valley, it's never really done anything at all. It's sitting around--

Sullivan  03:05

It's still [crosstalk].

Moffet 03:06

--reminding us of our past sins. Anyway, Bolton said, yeah, we could take somebody on. And so that summer and subsequent summers, I worked for the radio astronomy group and--

Sullivan  03:17

Did that dish ever do any high-frequency stuff?

Moffet 03:19

No. After it moved to the Owens Valley, it really did nothing at all. By the time it was set up and working, the first 27 m antenna was getting ready to work. And it was obvious it was more important to work with the bigger dishes. So, although several experiments were proposed for it, and indeed, at one later stage, it was pulled down off of its stationary mount and mounted on a carriage with some old surplus bomber wheels to carry it off across the desert as a radio-linked outstation, that project was never carried through. So, it hasn't really done anything since it left Palomar.

Sullivan  03:53

So, what did you do in those first couple of summers?

Moffet 03:56

Well, the first summer, we were putting together the first 27 m antenna there was a 22 MHz wire array, an array of dipoles. Two pieces of it separated by several hundred meters with which they were looking at the [pass?] transit of Cassiopeia and the Crab every day. And they did do a study of the Crab Nebula occultation behind the solar corona within one year which was worthwhile at the time.

Sullivan  04:25

Barry Clark was [telling me?] [crosstalk]--

Moffet 04:27

Barry Clark was supervising that thing at the time. He was then an undergraduate. So, I handled the business end of a theodolite while surveying in baseline and surveying in the pieces of the first 27 m antenna as it was welded together there.

Sullivan  04:45

So, it wasn't a matter of a subcontractor putting the antenna together. It seems like--

Moffet 04:50

Oh, it was but we were responsible for checking to see that they lined things up correctly as they welded them together. In other words, we were the inspectors on the job. The antenna was put together by a steel erection firm. The fabrication was done by a small steel manufacturing house in Phoenix, Arizona, to the Caltech design. And I guess I got involved with wiring together some of the drive circuitry on the antenna and a few other odds and ends during that first summer. By the time the second summer came around up there we were already using it. And I wound up doing some of the routine observing of sources at 960 MHz. Later went into the CTA list that Harris and Roberts-- or Roberts and Harris published.

Sullivan  05:49

Right. This was the summer of '59?

Moffet 05:50

Summer of '59 then. And I have a sort of a proprietary feel for CTA 21 because I found it by virtue of falling asleep one night during that survey and letting the antenna drive too far and a strange bump showed up on the record which later turned out to be a real source. The next night I went back and looked more carefully in that location. And so that was the first source I ever found that somebody else hadn't found already.

Sullivan  06:17

And what was the basic idea of the CTA survey?

Moffet 06:20

Basically, then we were checking the reliability of the 3C survey. The 2C disaster had by then become a fairly well-known fact. And we had a preprint of the 3C list and we were checking all the stronger 3C sources at least to see first of all, whether they really were there and secondly, to measure their spectra over the range of 159 to 960 MHz. And I think we were of considerable assurance to the Cambridge people when we told them that, yes, we were able to find at least most of the strong sources that they said were there.

Sullivan  07:00

Right now. And now in fact this just showed up some interesting sources.

Moffet 07:04

It did. We discovered two or three sources that were not in the Cambridge list although they should have been. CTA 21 and CTA 102 are the most important of those. There was one in a later survey that was called CTD 93 which turns out to be an interesting source also. It's a small diameter component. For some reason, other people have tended to leave those names attached to them even though they have very little connection with other important surveys. But mostly I think the most important part of that was to show the reliability of the 3C and to show that there was a rather narrow spread in spectral index. I think that was perhaps the first demonstration of that for a significant number of sources at least.

Sullivan  07:49

Now, this is all still with a single dish?

Moffet 07:50

All with one dish, yeah.

Sullivan  07:52

I guess the second dish was being built at this time, or.

Moffet 07:54

The second dish was built just a few months later than the first one. So, it was available for some work during that summer. And I think during that summer, Radhakrishnan was there trying to make a 21 cm spectrograph work on that dish. Maybe that he arrived later and started that work. I'd have to check to see just his day of arrival there. Maybe that he did most of that work the second summer. The first summer, yes, I guess that first summer, the other dish was being used daytimes by the solar people from Convair, who had a 500 to 1000 MHz solar dynamic spectrograph, and we more or less let them have the other dish while we did some of that first work.

Sullivan  08:42

In '59 [inaudible]?

Moffet 08:43

Let me see. No. In the summer of '59, Rad had the dish closest to the observatory control building for his 21 cm exercises. And we time shared the other dish, the one farther out from the observatory with the solar people. They had it in the daytime and we did the 960 MHz work after dark. We were afraid of sun in the sidelobes, and indeed the daytime observations were not very good because of Sun in the sidelobes. Dickey switched radiometer against a liquid nitrogen cooled resistor most of that time. And one of the exercises was to find which of the thermos bottles would break the first time you poured liquid nitrogen in them. It turns out that most of them will survive it, but once in a while one breaks the first time you get it cold. But they're a lot cheaper than laboratory doers, which are guaranteed to hold liquid nitrogen. Then by the next summer, we had a primitive interferometer going already, and we're beginning to figure out how to measure source positions with it and surveying all of the sources that we'd found with the single dish--

Sullivan  09:58

[inaudible] more accurate positions.

Moffet 10:00

Well, more accurate positions and to begin to get a notion of diameter, compare their fluxes with the single dish fluxes.

Sullivan  10:08

Were you changing the baseline?

Moffet 10:09

No, well, we played around a little with changing of baseline, but mostly we were testing out methods of getting a coherent interferometer. For instance, for a while there was a time when we ran a single local oscillator in a hut in the center between the two antennas and sent the signal from that rather powerful triode oscillator down an open wire parallel line transmission line to each antenna, and then on up to the focus of the antenna. That didn't work too well, and it involved picking up this little hut which sat in the middle between the two antennas each time we changed spacing so that it was exactly in the center between the two antennas. Actually, we ran that way with that little hut, all the time we were doing 960 MHz interferometry there. It wasn't until quite some while later that we developed methods of phase locking independent local oscillators in each antenna and could do away with the center hut. Although we did away with the open wire transmission line fairly [inaudible].

Sullivan  11:19

I was going to say there was some cable burying that--

Moffet 11:21

Yeah, we had buried cables for the IFs and then later on we buried very big diameter coaxial cable with low loss along which we could send oscillator signal. We still use that buried cable, but it carries a low-powered reference signal instead of the full-powered local oscillator. I would have to look up to see what dates we were able to introduce additional frequencies, but essentially we did interferometry only at 31 cm until I was finished with my graduate studies in 1961. But the interferometry was gotten reasonably civil by that time, although we still couldn't track with it. We had to measure with fixed delays at specific hour angles. Tracking was introduced by Bob Wilson and Dave Morris probably during the 1961, '62 period when I was in Bonn.

Sullivan  12:13

That wasn't part of--

Moffet 12:13

And so, for my thesis at least we observed three separate hour angles by putting in a fixed delay line and then catching the sources as they passed the white--

Sullivan  12:21

I see.

Moffet 12:21

--fringe at that particular value of delay.

Sullivan  12:24

But that wasn't part of the original specs for the antenna that they'd be steerable?

Moffet 12:27

Oh, the antennas were steerable, but we didn't have a tracking delay line.

Sullivan  12:30

Oh, you didn't have any delay lines. I see.

Moffet 12:31

And there were various efforts in the subsequent years of making analog computers to drive the delay line on the appropriate function, well, for a baseline of arbitrary orientation.

Sullivan  12:45

Okay. So, your thesis was concerned with the source position [crosstalk]--

Moffet 12:49

For thesis work, I measured not positions but sizes and brightness distributions of radio sources along an east-west baseline. Maltby came during the time that I was finishing up, and he and I measured north-south visibilities along the north-south baseline. So, by the time I left, we had measurements of north-south strip visibility functions, east-west strip visibility functions, and then some of these off-hour angle ones where we moved a fixed amount of delay off of the east-west line. And I suppose I had 130 sources or something or other of which perhaps half were resolved well enough to say something about their structure. And most of those showed up a characteristic double structure. So that was the--

Sullivan  13:36

That was the main result.

Moffet 13:37

The main result from that work was that most radio sources with angular sizes greater than a minute of arc or so are double.

Sullivan  13:46

Was that somewhat surprising or --

Moffet 13:48

It was a little surprising. It had been foreshadowed a bit by work with the small variable spacing interferometer at Nançay by Lequeux but he'd only been able to see four or five radio galaxies because he had very poor sensitivity.

Sullivan  14:03

Of course, he had Cygnus and Centaurus that had always been there.

Moffet 14:06

Cygnus and Centaurus were known from-- well, Cygnus from Jennison and Das Gupta and Jennison and others at Jodrell Bank a long time ago. But other radio galaxies like Virgo A were known to be more or less simple. So the picture was not clear until a reasonable statistical sample was produced. And we had a sample of 30 or so doubles out of perhaps 50 that were resolved well enough-- no. Less than 50 that were resolved well enough to say something about them. And that was a fairly conclusive result. It was reported by John Bolton at the London URSI General Assembly, which I suppose must have been 18 years ago.

Sullivan  14:46

1960.

Moffet 14:46

1960. And I think was received with some interest at that time. Maltby and I published an ApJ Supplement with a full set of stuff in 1962, I guess it was.

Sullivan  15:02

Well, now it seems like you changed fields rather drastically in order to go to Bonn. Did you just want to try something different or?

Moffet 15:08

Well, German was the only foreign language I knew and I wanted a change of scene more than anything else. And the only place in Germany that had any pretext of doing radio astronomy was at Bonn. So, I got a Fulbright scholarship and went off to Bonn.

Sullivan  15:24

I see. And what was going on there at that time?

Moffet 15:27

Well, they had one 25 m antenna on Stockert which is still there, which has a surface that's usable at wavelengths to 10 cm or a bit shorter. They time shared it with our radar research group.

Sullivan  15:42

I see.

Moffet 15:44

And therein lay a good many problems. Their radar research was classified, so there was one room in the base of the antenna that was locked. There was a dual set of controls for the thing and a great big switch that switched over from the radar people to the radio astronomy people.

Sullivan  15:59

And all the touch [inaudible]--

Moffet 16:00

And all the time I was there that winter, or almost all the time I was there, the antenna was down because it had a cracked bearing. So, I didn't do any observing while I was there. But it gave me an opportunity to boil my thesis down into one and a half publishable papers. And that was quite useful. And I certainly enjoyed getting to know a different culture.

Sullivan  16:22

It seems to me that that Stockert dish never really did anything terribly useful. It just made some hydrogen line maps. And they did a little bit with Sgr A and so forth. Is that your impression also?

Moffet 16:33

Yeah. Certainly, if you compare it with Dwingeloo, which was put into commission at about the same time, it was not anywhere near as productive.

Sullivan  16:44

Any idea as to [inaudible]?

Moffet 16:46

Well, in part, the Dutch knew better what they wanted to do. And they had some pretty vigorous people like Westerhout to get things going. And Oort, of course, was the guiding light behind the whole business.

Sullivan  16:59

And Muller, of course.

Moffet 17:00

And Muller did get some good receivers going. At Bonn, they had some personality problems. I don't know how deeply I should go into those. But Priester, who was nominally in charge of the radio astronomy group, couldn't get along with Peter Metzger very well. Metzger was there as an engineer, but he had ambitions to become something greater and better, which eventually he did. And so, he wasn't content to take just the role of an engineer. And they fought bitterly. That certainly held things back. They didn't really have any very good engineer there, so their equipment was not very reliable. And they had bought it all, or almost all of it. And they didn't really understand it I think. The receiver was bought from Telefunken for an immense amount of money. And in the years following that, they bought paramps from various people in Germany for huge sums, which didn't do anything either. And there was a sort of a lack of coordination there until Hachenberg arrived. He arrived just as I left.

Sullivan  18:03

From East Berlin.

Moffet 18:04

Well, he had a position where he lived in West Berlin, but was in charge of radio astronomy work, mainly solar radio astronomy, at the Heinrich Hertz Institute in East Berlin. Which was all very well until the wall was built in the summer of 1961. And then he was obliged to choose whether he would move to East Berlin and become an East German, or whether he would stay in the West and lose his position at the Heinrich Hertz Institute. And I guess it wasn't an entirely straightforward decision for him. But eventually, he decided to come West. And Bonn was in the market for a professor of radio astronomy, and they hired him. And I think history shows that that was a good idea because he certainly started immediately on plans for a large instrument which was eventually realized in the 100 m telescope at Effelsberg, which is probably the world's finest fully steerable radio telescope.

Sullivan  19:00

I have talked to him about those sort of things. Well, okay. You went back to Caltech in '63, would that be?

Moffet 19:07

I returned in the fall of '62 then as a postdoctoral fellow. And then worked some more on radio brightness distributions and miscellaneous other projects.

Sullivan  19:18

Now, the quasi-stellar sources, a couple of them had been identified by that stage, I think. Were you involved at all in that, or?

Moffet 19:25

Not really very closely. Although I observed it from-- well, actually, I was involved in some of the position measurements. But I didn't do the identification work myself. The first quasi-stellar objects had been identified. That is, the optical counterparts had been found before I left in the spring and summer of 1961. Tom Matthews and others had measured pretty good positions for a number of small-diameter 3C sources. And 3C 48, the position was good enough to pretty unambiguously identify it with an optical counterpart which was more or less starlike, except that there is a little bit of filamentary material that can be seen sticking above and below it on good scale photographs from Palomar. But the optical spectra of 3C 48 were not easily interpreted. And there was some uncertainty as to what that was all about. And then a couple of others were found. 3C 147.

Sullivan  20:30

Well, you don't need--

Moffet 20:31

And so on with that one.

Sullivan  20:32

I talked to people about that--

Moffet 20:33

And they're still not clear. And the matter did not get clarified until 3C 273 was identified as a result of Hazard’s occultation measurement in Australia. And that showed that it was in two pieces. And that's why the centroid of the radio position is measured with lower resolution methods, was not on any obvious optical object. It was between a fuzzy thing, which is the jet, and a starlight thing, which everybody thought was too bright to be any relation to a radio source. When that occultation measurement showed what the situation there was, then the spectrum of 3C 273 immediately showed redshifted Balmer lines. And then with that as the key, it was possible to understand 3C 48. And from that Maarten, principally, went by sort of a bootstrap technique and untangled the spectrum of a lot of other ones.

Sullivan  21:25

I guess another difference when you went back was that John Bolton had resigned by that time.

Moffet 21:29

John Bolton had left before I finished, actually. He left in December of 1960 to assume command at Parkes while they were still building it.

Sullivan  21:38

Did that substantially change--?

Moffet 21:39

Very much so, since John is a very strong leader. And he left things somewhat in disarray then with his abrupt departure.

Sullivan  21:49

Gordon Stanley took over, right, at that stage?

Moffet 21:52

Well, he was acting director for a year and a half or something, and then was permanent director for some years until I took over in 19--

Sullivan  22:04

Three years ago?

Moffet 22:05

1975. Yeah.

Sullivan  22:07

I also asked Al why aperture synthesis had not been done at Hat Creek or Owens Valley back in the early '60s when in principle, it could have. And he had several points to make. First of all. he thought that it probably was not phase stable enough. However, when I pointed out why wasn't there an effort made to make it more phase stable to do it, his answer was, "Well, you also need to have a good mini-computer besides the general purpose computer for offline reduction. Aperture synthesis requires a good online mini-computer and they just were not available around 1960, '61 at a reasonable price. And he also said that there were just several other programs going on such as source positions and hydrogen line interferometry a little bit later, and you can only do so many things. So, it seems like they still sort of missed the boat if they'd made the effort to concentrate on that. But it is true, you can only do so many things at one time. Well, that ends the interview with Alan Moffet on 11 August '78 at Groningen.