Oral History Transcript — Dr. Carl Romney
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
See the catalog record for this interview and search for other interviews in our collection
Interview with Dr. Carl Romney
Carl Romney; January 28, 1998
ABSTRACT: his interview focuses on Carl Romney’s involvement in the technical and political debates related to seismic detection of underground nuclear explosions during the 1950s and 1960s. Trained in seismology at Berkeley (Thesis Advisor: Perry Byerly). Romney began in 1949 to work for Beers and Heroy, an Air Force contractor, which developed methods for the detection of Soviet nuclear explosions. There he became involved in the Department of Defense’s early efforts for the seismic detection of underground explosions. As the DOD’s leading seismic detection specialist Romney played a central role in a number of technical conferences involving U.S. and Soviet scientists. These conferences, which aimed at defined the state of the arts in seismic detection techniques, had direct consequences for the diplomatic negotiations for a nuclear test ban. Romney describes in detail the technical and political controversies of the Geneva Conference of Experts (1958), the Technical Working Group 2 (1959), and the Seismic Research Program Advisory Group (1960). During the 1960s Romney became a key figure in the development of new seismic detection systems for the DOD’s Project Vela Uniform, which aimed at the improvement of U.S. seismic detection.
TranscriptSession I | Session II
Barth:Okay. This is Kai-Henrik Barth speaking. Today is Wednesday, January 28th, 1998. It's about 10:30 a.m. I'm sitting in the office of Dr. Carl Romney in Arlington, D.C. Dr. Romney, this is Part 2 of our interview which we have begun last week on January 20th, and I would like to start out today with the conference, the second Technical Working Group, Technical Working Group number 2 which met in November and December, 1959. Could you tell us a little bit about the background, why did this working group meet and what were the specific issues which were debated?
Romney:I have forgotten whether we talked about the new seismic data tabled in January of '59.
Barth:Maybe we can reintroduce this shortly.
Romney:Well, following HARDTACK II we analyzed the data and I think we discussed some of that, the process of analyzing the data and reviewing it. It was then tabled on January 5th, 1959, in Geneva, and the Soviets promptly blasted it or rejected it out of hand. I don't remember precisely. At any rate, they wanted nothing to do with it, and we proposed, either on tabling it or shortly thereafter, that the data should be reviewed by a working group of the conference and its implications should be cranked into the deliberations, the negotiations for the treaty. And the Russians refused to do that. They did publish, there was an article published by Riznichenko and some other colleagues, at least one other colleague whose names I have forgotten, toward the end of January attacking the paper and particularly attacking the instrumentation used. That was a major point. The Berkner Panel was formed at the end of December, and the Berkner Panel continued meeting during the spring and early summer, and their report was tabled in the conference in Geneva along about late May or early June. I have notes. Should I look at some notes on this?
Romney:Oh, I can't find it. Oh, it was tabled on the 12th of June. And it was a bit upbeat of course that the system, the Geneva system, could be improved, and that with vigorous R&D we expected even more improvements. Shortly thereafter the Technical Working Group 1 convened to consider monitoring measures for high altitude tests, and I believe that was relatively successful. There was a more or less agreed report, but it set a sort of a precedent for having working groups associate with the negotiations and ultimately the Russians agreed to convene a working group, Technical Working Group 2, to look at the new seismic data, among other things. I believe they introduced the idea of that the working group should also consider criteria for on-site inspection, trying to reach agreement on the conditions under which on-site inspection would be called for, and they insisted that the new data were not really needed, that the Experts report was a sufficient basis for continuing with the treaty negotiations, but they would listen, that was the long and short of it. And so we did, we met on the 25th of November and fairly quickly agreed on an agenda which covered both the points that we would seek agreed criteria for initiating on-site inspection and we would consider all relevant data. On the new seismic data front, I presented the data pretty much as covered in the paper of the 5th of January and as subsequently published in the Journal of Geophysical Research, and the key points were that the amplitude appeared to increase as the first power of the yield rather than as the two-thirds power, which we had used during the Experts time. And the implications of that was that for events smaller than around 25 kilotons or thereabouts the magnitudes could be expected to be lower than had previously been predicted, and therefore there would be more earthquakes equivalent to low yield events. There would be fewer equivalent to high yield events of course, but since the small events are much more numerous, it meant the system would have to cope with more earthquakes equivalent [in magnitude] to explosions. That was our conclusion, one conclusion, and the other was that the first motion, which was the principal criterion during the Experts Conference, the first motion was less detectable than we had estimated at the time we were estimating the capability of the Geneva system, and therefore there would be more unidentified events and more need for on-site inspection. This was not news that the Russians accepted easily, and they responded with considerable sarcasm at times, and it was a very contentious atmosphere that developed as we presented this data. And they attacked it on the grounds that the instruments were the wrong instruments, not the instruments agreed to by the Experts, and that we didn't have arrays and therefore that the information on the first motion was irrelevant; that had we used the Experts' true instruments that we would have had better results.
Barth:Can you maybe go into a little bit more detail at this specific point? Because I think this was one of the major contentious issues during the Technical Working Group 2 meetings.
Barth:So to what extent was the Benioff [seismograph], which was used by the American seismologists and the new seismic data measurements different from the instruments, suggested by the Conference of Experts?
Romney:During the Experts Conference, the United States presented their view of the kind of short period instrument, which is what we are talking about for first motion, the kind of short period instrument that should be used, and I believe it was Bob Bacher who actually presented the discussion on this. And I don't recall the details, except that he was clearly talking, in my recollection at least, about the Benioff, which has a peak in the magnification curve at around 3 Hertz and is about 3 to 6 dB down at both 1 Hertz and let's say 6 to 10 hertz, somewhere in that range. And another, well, that's what we thought was being accepted at the time in the Experts Conference. When the written description was agreed to, the statement was that it would have, the instrument would have a peak, no, the instrument would have a maximum magnification of 10 to the 6th at 1 hertz and sufficient pass-band for good signal reception. I'm not, that's not quite accurate, but that's the essence of it. At the time I objected to that description because I thought it was, could be misinterpreted that we were saying the peak in the magnification curve should be centered at 1 hertz. I was voted down by others in our group working out the details of the written agreement through the translators. I was voted down on the grounds that the phrase about sufficient band-width for good signal reception covered the point that the Benioff could fit this description. And it was the Benioff that we had used during HARDTACK II. Now apparently the Russian translation of this didn't contain the ambiguity that exists in the English. We meant that at 1 hertz the maximum gain you needed was 10 to the 6th, and it didn't say that you couldn't have more gain at still other higher frequencies. But apparently the Russian translation did imply that the maximum, the peak of the magnification curve should be at 1 hertz. Now that was consistent with the SVKM instrument that the Soviets used, and we didn't think much of that instrument. I mean, we had seen, we had actually had experiments previously evaluating the SVKM as compared to the Benioff, and so we had some ideas, we had a pretty good idea, that the Benioff would in fact be better at detecting first motion than the SVKM. At any rate, we were, I was and we were, the whole group was attacked at — you know, had we used some other instrument, and I believe Pasechnik at one stage claimed that the SVKM fit the bill. Subsequently, as I think you reminded me, Federov said that no, we are not saying that the SVKM is the instrument that the Experts use, but perhaps it is closer to what the Experts proposed than the Benioff. But, so that was one part of this argument against our data. The other part of course was that we did not deploy arrays, and we tried to argue that one as scientifically as we could that all arrays do is improve the signal-to-noise ratio, and we had adequate signal-to-noise ratio with single vertical Benioffs to see the first motion at many places so that our claim that it was smaller than we had previously thought was substantiated even without the arrays. The arrays would just help confirm it even more strongly. But the argument that the Russians made was this semantic, really semantic argument that our data was disqualified on that account.
Barth:I think there was a third argument made concerning the calibration of the instruments, which doesn't have any effect on the first motion detection but in the Russian view produced some doubt about how carefully the measurements were done.
Romney:Yeah. The instruments were well calibrated for the time certainly, but I think the main argument they made was that there were many other instruments in the United States that had recorded these signals and that we didn't use the data from them, we should have used the data from all of these stations. Well, the reason we didn't is that they were university type stations for the most part and they were not calibrated, and certainly not calibrated within the precision that the instruments that we had installed, especially for the experiment, were calibrated. And they expressed outrage that — I think that's literally the word that was used at one stage — that we had not gone out and calibrated these other instruments. Well of course we had all we could do to organize the experiment that we did organize, number one, and number two, I don't think there's a university professor in the country that wants somebody from the Air Force to come in and calibrate his instruments. It was kind of an absurd criticism, but they did in fact suggest that.
Barth:I found a document, I think Wolfgang Panofsky referred to this once, that the American delegation was in a relatively weak position in this argument because they couldn't disclose the location of the AFTAC stations at this point, or they couldn't show the seismograms produced by certain AFTAC stations at this point. Is this?
Romney:That's partly correct. There were additional data from several AFTAC stations that we had where we had good measurements of magnitude. We actually gave them, the Soviets; we actually gave them some of that data, thinly disguised. I think we, one of the stations was in Alaska, not far from College, Alaska, which was not literally a part of our experiment, but we identified the data as coming from College when in fact it came from an AFTAC station nearby. I know we did at least that, and I think we possibly gave them some other data from Wyoming, and I don't remember that, where there were some AFTAC stations at the time. But, so he has a point, but I think it's not a very important point, because we had sufficient data from the stations that we could disclose to substantiate what we were saying about the magnitudes, I think, and about the first motion, first motion particularly.
Barth:It seems though that the Soviet experts did not challenge the data per se in the sense of, well, “we can mistrust for example your values of the kilotons of the explosions”. They could have started to challenge this as well, and they accepted this basically face value, said okay, BLANCA is 19 kilotons, LOGAN is about 5, and RAINIER is 1.7. Of course, these values, there is no way to check them independently. You have to rely on the values coming from the Atomic Energy Commission. So they basically accepted this data, and took the data from the January 5 paper and your presentation in the beginning of the Technical Working Group and then let loose a different seismological and statistical analysis on these data.
Romney:Yes. No, they never, as far as I know, they never questioned our yields. It was solely the data that we had presented to them a good many months before. They had had ample time to analyze it, but apparently they had not really done anything serious with that data at TWG 2, and I think it was during TWG 2 that they began in sort of real time looking seriously at the data and trying to find grounds for criticism or reinterpretation or both. But it sat in Moscow, must have sat in Moscow for a good many months with no one looking at it.
Barth:Although Riznichenko wrote a paper on this?
Romney:He wrote a quick paper. It was evident that he had seen that stuff, but from the essence of the paper he had not done any serious reanalysis. He didn't present alternative explanations at all. He just criticized it on the grounds of the frequency response characteristics.
Barth:Did you see at this time, I think this article was published in Izvestia even, I think a political…
Romney:Yes. Or Pravda.
Barth:Yeah, political newspaper, not in a scientific journal.
Romney:Yes, that's right.
Barth:Did you see a translation of this at this time?
Romney:Yes, yes. Yeah, I saw that and possibly it was given to the delegation in Geneva or drawn to their attention in Geneva, because I believe I received an English copy of it, from I think I received it from our delegation. I believe I did. I can't guarantee that, but I did see an English translation of it, and I did write to Doyle Northrup who was in Geneva as a member of the team, I wrote him a counter, a minor rebuttal, a letter, maybe two or three pages, rebutting that article. I do remember doing that. I probably still have that [letter]. I think I'm almost sure I have that.
Barth:So there were disagreements between the Soviet and the American experts on the instruments used, on how to handle the data, how to use the magnitude scale. Maybe we should talk about this for a while, because this is a very important point.
Romney:Ultimately that was, that came about. The presentation that Riznichenko made, I've forgotten the date but…
Barth:I think it was in the third meeting.
Barth:It must be the 28th or 27th. I'm not quite sure.
Romney:I think it was later than that. But at any rate, the criticism that Riznichenko made was based on two things. He accepted we measured the magnitudes at small distances, which are distances less than about a thousand kilometers in the local magnitude scale, the M sub L scale, the original Richter magnitude scale, and that was based on the recordings of data at university stations. We knew that the Wood-Anderson torsion seismometers, which are the basis for the ML [M sub L] scale were calibrated in fact. Most of the stations were Caltech or Berkeley stations, and we had great confidence that there was good data.
Barth:Calibration always meaning that a certain earth displacement translates into a certain voltage.
Romney:A certain motion on the photographic recording, yes, yes. And it's very hard for Wood-Andersons not to be calibrated, I must say. It's intrinsically built into the way they are constructed. They are just, not much can go wrong. At distances beyond about 16 degrees, which was the distance where Gutenberg said you can measure the m sub b, the body weight magnitude based on the P-waves, we used our portable stations that had been put out for the purpose and as we previously discussed we also had a bit of AEDS data at that time. The magnitudes that we used or that we deduced using both scales were presented to the Russians. They never challenged those magnitude estimates or re-measured the seismograms that we gave them. They used what we presented, but they took a paper that I believe I had given to Riznichenko, written by Beno Gutenberg in 1956, published by Beno Gutenberg, where he had proposed a relationship between the M sub L and the m sub B scale [mB]. It was based on data from very large magnitude events, magnitude 6 to 7 events, and extrapolated down to low magnitudes based on very little data, if any, down around magnitude 4, which was a relevant magnitude range for these events that we're discussing. Riznichenko used the formula that Gutenberg had presented in that paper, cautiously presented, I must say, and with caveats that this is very tentative. Anyway, Riznichenko used that formula and converted the local magnitudes, the Richter magnitudes to mB by the formula. That increased the magnitudes so that he converted the ML s into an equivalent mB, and those were larger values than what we had measured. He also unveiled the fact that these explosions had been recorded at three Russian stations, and he gave us the magnitudes that had been measured at the three Russian stations. So he introduced that data. Then he looked at the local, the converted local magnitude data, and discovered or pointed out the fact that two of the stations had very low values compared with the others. And so he proposed, he, well, he not proposed, he did eliminate them from his magnitude on the grounds that he had a statistical argument that said those readings at two stations were statistically different from the recordings at the stations giving larger magnitudes.
Barth:This was Woody and Barrett stations.
Romney:This was, yes, Woody and Barrett were the two stations, and Frank Press mentioned either in the meetings or in private meetings that those were two of the best stations he had, and he thought that it was absurd to leave them out of the analysis. And John Tukey gave some lectures on proper statistical tests for rejecting data, and Hans Bethe did the same. At any rate, we never got Riznichenko to change his interpretation. So he eliminated the low values and then he pointed out a trend in the magnitudes, in the distance range from about 16 degrees, which was where we first made our measurements around 16 degrees to around, well, 2500 kilometers. So from about a thousand kilometers to about 2500 kilometers he said there's a trend in these data and so he defined a-new-to-seismology second zone, which extended from around a thousand kilometers to 2500 [kilometers]
Romney:...hundred, and what he called a third zone was from 2,500 [km] on to great distances, including the stations in the Soviet Union. The trend, he drew a line through these points that seemed to increase with distance, although there was some scatter, drew a line through there and said these should be rejected and the only magnitudes that should be used were those at greater distances. And on this basis, he said that the magnitudes were several tenths of a magnitude higher than we had said. In the case of RAINIER it was almost seven-tenths of a magnitude unit greater. In the case of…
Romney:LOGAN, it was a few tenths, and in the case of BLANCA it was, well, a few tenths in BLANCA. And so he came out with new magnitude values where we had proposed for the three events in order of magnitude we had said we thought the magnitude [of RAINIER] was 4.07 ? 0.4, and I'm referring to my notes now, so I know these numbers are accurate. LOGAN we had estimated at 4.4 ? 0.4, and BLANCA 4.8 ? 0.4. For the same three events what Riznichenko called corrected magnitude were 4.7 ? 0.1, 4.95 ? 0.1, and 5.2 ? 0.1. Now the [smaller] uncertainties reflect the fact of course that he had rejected all of the data that were, he had rejected all the data that exhibited scatter from the values he got at great distances. And I will add editorially that the, from all the experience that we subsequently had measuring large explosions, we always measure standard deviations in the .3 to .4 range. There's no reason to expect accuracies of ?1/10th in the magnitude scale, however. Richter would roll over in his grave if he saw an interpretation like that. So that the, my conclusion at the time and still is that it was a very spurious kind of interpretation and that the data that didn't agree with his results are valid, were valid, and that our interpretation while not perfect since we didn't really know the relationship between the ML and mB curve. Nevertheless, I think our interpretation was closer to the truth than his and subsequently I think that's become fairly well established that that is the case. We simply assumed that ML and mB agreed rather than disagreed, as Gutenberg's formula predicted. And after all, they [Gutenberg and Richter] had intended that they would agree, that when the mB scale was created it was intended that it give exactly the same result as the ML. And it was only later that Gutenberg thought, well, maybe we didn't quite do it right. Well, our interpretation was they did it better than they expected. And so we combined all of the data and came up with the values that I just mentioned.
Barth:What were your impressions and the impression of the other American experts after Riznichenko gave his paper, his rebuttal of the American interpretation, was there an impression that he massaged the data too much, that he went overboard, that he went too far in trying to make the data fit a certain political assumption? How would you describe your first impression after listening to his reinterpretation?
Romney:Well, my impression at the time was that he was deliberately distorting the data to make the magnitudes as great as he could. Now after all, and I pointed this out at the time, one could have taken the Gutenberg formula and instead of increasing the ML values they could have decreased the mB values. So there's an essential ambiguity in this kind of analysis that Riznichenko was indulging in. And that would have ended up with net reductions in the magnitudes even below the values that the U.S. had proposed, and so this, well, I won't say this interpretation was silly but I'll just say that there is an essential ambiguity in it and one shouldn't use it. If one doesn't like combining the data, to me that would be a valid objection. You could publish two values, an ML value and an mB value. The ML values you could relate again to numbers of earthquakes worldwide as well as you could the mB values. And when you did that, then the numbers of earthquakes equivalent to these events were very much larger than what the Americans had presented. So I thought that it was not good science at the time, and that he must be politically motivated. Now, as to what I think today: I don't I think — I know Riznichenko a little better from his writings and subsequent personal meetings with him, and I wouldn't accuse him of deliberate distortions but perhaps of a kind of tinted eyeglasses when he looked at the data, tinted by his, you know, the needs of, or the perceived needs of, the Soviet government to minimize the intrusiveness of the control system that was being discussed in Geneva. And so to make the detectability of the, making the detectability of the explosions seem greater, that reduced the pressure to put seismic stations on the territory of the Soviet Union and reduced the need for an agreement on a large number of on-site inspections. So I guess that's as far as I can go in saying how I interpret it now. But…
Barth:The interesting question which is probably very, very hard to answer is to what extent the Soviet seismic experts were actually directed to push the interpretation as far as possible or to what extent their own view of the American conservative evaluation just paralleled the official Soviet view. So it's a very gray zone where direction starts and guidance starts and where your own personal opinion, your autonomous personal opinion comes into play in evaluating these data in a certain format.
Romney:I think they were under very strong guidance to keep the Experts report as the basis for the negotiations. And so I think they were not to accept, or at least argue as hard as they could not to accept, any evidence that seemed to go against the earlier conclusions. I believe that's true and I believe that that's reflected in the words that were said in the negotiations leading up to the Experts Conference, and I believe it's reflected in Federov's words that “we see no need to,” and I think this is an exact quote "we see no reason to change the estimates of the Experts." Very close to his exact words.
Barth:Let's look now at the American side. I think it's very difficult to argue that the Soviet scientists were politically guided and the American scientists were free of political pressure. I think there was also quite some political pressure on the American scientists. It was well known that the Atomic Energy Commission and the Department of Defense were not in favor of a badly monitored Test Ban Treaty, so there was definitely the pressure on the delegation to evaluate the data in a very conservative way in order not to make the same, I wouldn't say mistake, but to come to the same kind of agreement like the Conference of Experts report which, after publishing the report then showed some loopholes and some problems which had made ramifications for national security. Could you say something about the political pressures on the American experts?
Romney:Well, I think there was no net political pressure of any sort that related to how one looks at science. There certainly were major interagency differences within the U.S. government as to the advisability of the test ban itself, and no doubt some of the members of the elegation were much more inclined to emphasize problems than others and to take little discrepancies as more serious than others, but I don't think there was any net guidance. I think we wanted an agreement ultimately with the Soviet Union, but we wanted it on the basis of the best possible interpretation of the data we had. Now some of us probably, I at least was, a bit, was motivated to be a bit conservative in the interpretation because after all we felt we'd been burned by being too optimistic at the time of the Experts and that the subsequent data showed that that optimism was not well justified. So I think some people would have been conservative, but I don't think there was any, I don't believe there was ever any guidance to me from the DOD for instance to take that kind of approach. I think it's something in my own little soul if I did take that attitude.
Barth:What do you remember in terms of feedback from Washington during the meetings? Did Fisk report to Wadsworth or did he report to the Committee of Principals at the White House level? Was there any exchange going on between the Department of State and the Technical Working Groups and the technical experts, the U.S. experts during this conference?
Romney:I'm sure they were kept fully informed. I'm sure both Wadsworth and the back-stopping committee, or whatever we had at the time back in the United States, were kept fully informed. I don't recall any guidance coming in the other direction, though. I don't recall any guidance. I don't believe there was any shift. Usually when we got beyond the guidance that we had there was a pause in the negotiations or the deliberations while you were waiting for the Great White Father back in Washington to make up his mind, or the Great Red Father back in Moscow, to make up their mind what to do. I don't think that happened during Technical Working Group 2. So I certainly, if there was any suggestions out of Washington, I certainly was oblivious to them at the — well, I can't say I was certainly oblivious to them at the time, but I am certainly oblivious to them now.
Barth:Sure. I think this covers our discussion of Technical Working Group 2, definitely for the moment, and maybe we can move on to the next technical meeting of experts you were involved in and the…
Romney:Well, can I say a little bit more about Technical Working Group 2?
Romney:Because the other major topic that we discussed was decoupling.
Barth:Okay. Yeah, sure.
Romney:And well, decoupling is not my field of expertise or anything. It was the theory of decoupling in a big hole had been developed by, primarily by Al Latter in the early 1959 period, and reviewed carefully by the Berkner Panel, carefully to the extent — I disclaim great knowledge in this area, so — but the Berkner Panel did review it, Hans Bethe and Frank Press who were both more theoretical than I looked into it carefully, and we believed that there was real substance there, and I believe the Berkner Panel report accepted the fact that this was a theoretical possibility that there could be an important degree of decoupling. I believe we were very conservative and said it needed to be tested, but I believe the Berkner Panel did point out that we had one event, namely the EVANS explosion, which appeared to be at least an order of magnitude worse coupled, or lower coupled, than AMALPAIS, which was about the same yield. And so there was empirical evidence that maybe small changes in the emplacement conditions could affect the signals in a major way. That conclusion, or the conclusion of Latter, or both, were presented at TWG 2 to the Russians. And it took quite a while I think for the Russian group to fully understand that, that in fact, that was a real possibility. In fact, Sadovsky and or Pasechnik I think pointed out or mentioned some experience they had had with firing explosives, high explosives, in cavities. And in their experience they said it went the other way, which they were better coupled. But eventually the Russian scientists did understand. Keilis-Borok, I seem to remember, understood it very quickly. The concept is fairly simple once you understand it, and I think he did, and passed on a simple view of it, and why it is probably true, to Federov. And everybody who was there when they [the Russians] finally understood this real possibility — there's different interpretations of what they saw, but it seemed to me that Federov turned visibly gray at the time. He took it; I thought he took it very hard when he finally did understand that and that this is a real thing. And so that was the other major contentious thing that we had to present. And then I want to say a little bit about the criteria for on-site inspection. We did discuss it, we did try very hard to define criteria, but it was very clear that the first motion, which was the major identification criteria, was a very weak criterion and it applied only to very large events in any practical system. And in the end we were completely unable to agree with the Russians on suitable criteria for on-site inspection. I do remember that the presentation that the Russians gave as to their ideas — I guess the better way to say it is when their ideas were really carefully summarized, maybe their preliminary presentation for the final report was done by Ustyumenko and we quickly pointed out to him that the criteria as he proposed it would have made the largest underground explosion up to that time, BLANCA, be interpreted as an earthquake, be misinterpreted as an earthquake. So, and they quickly pointed out that our criteria, as we proposed them, which were very safe and were intended never to let an explosion be missed, would identify only the very largest of earthquakes and leave literally thousands of unidentified events. So I've thought about this, written a little about it in a paper I wrote on on-site inspection at one time, then the lesson I think is that in a field like seismology where the uncertainties are so large and at the time when our understanding of the seismic signals from explosions was so weak and primitive, it was very dangerous to try to agree on criteria. It leaves no room for the improvements that could be expected and did in fact come out of the research program. So that was my, sort of my final word on TWG 2.
Barth:It's such an important story, so I'm glad we have it on tape. Okay. Maybe we can continue with the, very shortly to the reactions, the public reactions and the political reactions to TWG 2? I remember that Tsarapkin was very unhappy and very angry about the outcome, and his strong reactions created quite some reactions in the White House.
Romney:Gee, I do remember that Tsarapkin really blasted the Americans after that, after TWG 2. I guess my mind is blank on the general, the public reaction.
Barth:What then brings us from TWG 2 to the Seismic Research Program Advisory Group in 1960? What lies in between? So how do we end up with another Technical Working Group on seismology?
Romney:We during — well, it goes back to TWG 2 and there's another element that I neglected to mention earlier, but we did talk about possible improvements. And of course the grounds for improvement had been laid by the Berkner Panel's report, and so we did discuss with the Soviets during Technical Working Group 2 how to improve the Experts' system, and for example, and the thing that sticks most in my mind is the expansion of the seismic arrays from 10-element arrays to 100-element arrays, which would give you another factor of 3 or so in signal-to-noise ratio. The Russians suggested some ways to improve the system too, and as we did, during the meeting itself, and all suggestions for improvements were gladly woven into the Russians' final report. And so there was left a bit of unfinished business. Both sides had agreed that there were ways to improve the system. The Soviets of course rejected the fact that there was any need to do so at the time, but nevertheless there was this unfinished business. And that led to the idea, and I don't know the, I've forgotten the politics of how it developed, but that led to the idea that the three nations who were involved at that time, Britain and the Soviet Union and the United States, should maybe consider seriously what could be done to make things even better. And so the agreement was finally reached then to convene the Seismic Research Program Advisory Group in Geneva to agree on a joint research program, or to agree at least on a program, or programs, of research to improve the system. So we did. We met in May in Geneva. Our delegation was at a now substantially lower level than TWG 2 in that instead of having somebody from PSAC chairing it with great political guidance, Frank Press was the chairman for the U.S., Sadovsky was the chairman for the Soviet Union, and a man named Henry Hulme was the chairman for the UK. We met and exchanged ideas. At that time the Berkner Panel recommendations had been built into a concrete program of research and the authority to carry out that research had been given to ARPA, working with AFTAC and also working with the AEC, and the program that was in fact already under way in part was described in considerable detail to the conference.
Barth:The Vela Uniform Project?
Romney:The Vela Uniform Project, yes. And the Soviets presented there some ideas on research. It was clear they didn't have a research program in the sense, in as highly organized a sense, as the Vela program, but they presented ideas. The atmosphere was completely different than during TWG 2. Everybody was very friendly, and a great spirit of cooperation among the three, the delegates from the three, countries. And the Russians criticized our program, our LRSM program, Long Range Seismic Measurement program, as having an inadequate number of stations. For instance, well, we were proposing 20 seismic stations to measure data from large explosions, and they said, "Oh, you should have to at least double that," and we did. We cabled back to the United States and got approval that there would be more money put into the budget to do this work. The other interesting part of it was there was an agreement on the need for explosions and we talked at — to support this research — and we talked at length with the Soviets. They, the Soviet scientists' attitude was that…
Barth:...speaking. It is part number 4 of the interview with Dr. Carl Romney. Today is January 28th, 1998. It is about 20 minutes to noon, and we were just discussing the background of the Seismic Research Program Advisory Group in 1960 with Dr. Romney.
Romney:Yes. Well, we came to agreement that large nuclear, large explosions, excuse me, would be needed for the, in support of the research program that we were jointly trying to develop. The Russians had made it clear that they would not support nuclear explosions on Soviet territory. The negotiating countries were in a moratorium at the time. That was one of the conditions for the negotiations. But they said that they would propose very large chemical explosions. We're talking about explosions of the order of 1 to 10 kilotons at this time. And that if explosions were to be carried out by the — nuclear explosions were to be carried out by the United States — there would have to be a high level agreement that they were permitted, presumably as some kind of an exception to the moratorium, and that the Russians would want to be there to participate and share in the data. And we went on. We had extensive discussions of how we would exchange data and the fact that we would have observers, share observers back and forth among the countries participating in the program. Ustyumenko worked with us. We must have had — we must have formed — some kind of a small drafting committee to put this stuff together, and I believe Ustyumenko presented a preliminary version of this joint program sometime around, oh maybe like, the second week into the meetings. So all was sweetness and light and cordiality and cooperation, the kind of cooperation one would really like in an international program. I guess IGY [the International Geophysical Year] was a recent model that was fresh in our minds. Meanwhile, preparations were underway for a Summit Meeting in Paris with the Soviets. The Gary Powers [U-2] incident, however, which was at the beginning of May, around the 1st of May as I recall, was festering more and more in the public press. I believe Khrushchev did not even acknowledge the incident for a week or so, roughly a week. But when he did, that would have been just before the Seismic Research Program Advisory Group met. When he did, it was with lots of polemics and criticism of the United States for doing that. Well, the long and short of it was that the Summit Meeting was canceled as a result of that incident, and Eisenhower's refusal to apologize and, I believe, and punish those reasonable for it. And this event would have taken place very near the time that I mentioned when Ustyumenko, I can't remember whether it was a little before or shortly thereafter. But we were suddenly told privately, that is out in the hall, by Sadovsky, that they had been given orders to withdraw from the meetings, break up the meetings. And he asked for time to see if he could get this decision changed in, presumably back in Moscow. And Frank Press was under pressure to terminate the meeting. I guess he had expected that — he had allocated a certain amount of time like two weeks to complete this [meeting]. And this meant if we gave Sadovsky a few days to do this, this meant we'd run beyond the time that Frank had allocated. And so we tried to convey this to them — that it was very inconvenient for us to take time. But anyway, the long and the short of it was we realized that Sadovsky really badly wanted to come to an agreement on this and that we should give him the time and so we did. We went into a recess for several days. I'd have to refer to my notes to find out how long. But like four or five days. And when we reconvened again after word came back from Moscow. The word was unfavorable, and so we had a last meeting where we all said kind things about each other and closed the meeting. And so the possibility of a joint program had vanished. Tsarapkin very promptly, within a few days, disavowed all the agreement that had been reached, said that the Russians were not going to carry out any program, there was no program, and furthermore he insisted that if the United States was going to carry out any program that it would have to be with the agreement this Russian government. Which of course was totally unacceptable to us. I believe in retrospect that he was really focused on the nuclear explosions issue, but that isn't what he said. He was quite categorical that we could not have a research program without the Russian agreement. The Vela program did proceed and I guess the aftermath of that is that we did invite Soviet scientists to come and visit us and find out about what was being done, and it was on that occasion, I mean it was for that reason that Riznichenko came at one time. Pasechnik came, and in both cases we met them, made them very welcome, gave them VIP briefings on what was going on, and at least in the case of Riznichenko took him to various places around the United States where work was going on. I personally escorted him out to the so-called Geneva Array at the Uinta Basin Seismological Observatory near Vernal, Utah, and we had lots of time together to talk about old times, and that was when he told me he had been very upset about my comments on TWG 2. But I think we got very comfortable with him. Pasechnik was with, was accompanied by, a man named [Vladilen F.] Pisarenko, a scientist named Pisarenko. And they were guests, dinner guests at our house one night, and it turned out that Pisarenko was a great fan of American folk songs. And so we had, after dinner, we had a singing session where we sang "On zee Bank of zee Ohio" and other similar songs. I remember it was a very nice occasion. So that's kind of the end of TWG — I mean of the seismic program. But we did have at least one-sided cooperation with the Russians after that. We kept them informed. Not systematically or anything, but periodically.
Barth:Would you like to start with the Vela Uniform now, or shall we...
Romney:Okay. Let's eat.
Barth:Yeah. [tape turned off, then back on...] Okay. So we can continue our conversation about Vela Uniform. And would you like to tell us a little bit about the start of Vela?
Romney:Well, after the Berkner Panel report was submitted, there was a meeting between DOD and I believe [James R.] Killian's office and I'm a little hazy now, maybe AEC, and agreement was made that DOD would be the lead, that there would be a program, DOD would fund a program, and that DOD would take the lead, with assistance from AEC, particularly on shots, and as it turned out a little later also on the development of the Vela Satellite System. But that was quite a bit later. The early program, we were notified by Herbert Loper, who we always referred to as General Loper, he was the assistant to the Secretary of Defense for Atomic Energy at the time, a retired General, and we in AFTAC were notified this program was to go forward and that he was going to recommend it be in AFTAC. We then prepared a program plan which is very similar to the plan that I showed you. We developed a plan in late — or mid-summer to late 1959 it would have been — and the program that I showed you was as of about March or April of the next year. But it's very similar to that, and we notified, we in AFTAC, notified the people we thought would be interested in this program, namely the universities that do seismic research and a few of the companies that we knew that did seismic research and invited them to submit proposals and began collecting information. Shortly after that we heard from Herb York, who was then director of Defense Research and Engineering, that the program would in fact have to come, have to be managed by ARPA, and but that AFTAC, we in AFTAC, would continue to play an important part in the program. So that's how it developed. We were a bit disappointed at the time, of course, but soon a program was in fact funded, late '59 I believe, with about $9 million, something of that order, and we began writing contracts with some of the universities and some of the industrial firms that were to continue for some time. ARPA began rather slowly getting a little bit of a hold on the program. They hired and brought in a man from the Institute for Defense Analyses named Carlton Beyer who headed the program within ARPA and he in turn I think not very long after, oh probably early 1960, convened an advisory group. I think — I don't think it had a — I don't think it was a standing committee or anything of that nature, but he began seeking advice from other people and the program was off and running. So that was the beginning.
Barth:How did the academic seismologists react? Were they happy? Was there kind of euphoria about all the possibilities coming with Vela?
Romney:Oh, I think there was a fair amount of excitement and interest, and certainly the participation by the University of California at Berkeley and Caltech, the University of Michigan, Lamont Geological Observatory, Columbia, the participation was quite vigorous and enthusiastic. I think they were delighted that the program was off and running. The other element of the program, the other very early element of the program was the beginning of the WWSSN, the Worldwide Standard Seismic Network, and that was passed on to the then Coast and Geodetic Survey, I believe is what they were called, the seismology group there. And they, or they and ARPA jointly, convened a panel on seismology from the National Academy of Sciences. The panel on seismology heard about the ideas and actually wrote a report describing, giving their recommendations as to the kind of instruments, which is sensor instruments and recorders and timing systems and so on that would go into the stations of the WWSSN. And that also was a program of great interest to the academic community. I think they recognized that that was a very valuable resource that was being developed with standard responses and good calibration.
Barth:How much influence did AFTAC have on the selection of stations? I remember that the Committee of Seismological Stations of the National Academy, they apparently made a couple of suggestions like I think they suggested 125 stations, but AFTAC apparently had the last word to some extent in the selection of these sites. Do you know anything about this situation?
Romney:Well, I sat in on the meetings of the Academy, and I don't recall that AFTAC had the last word. Now possibly, we certainly had some influence because we did have experience with selecting good sites, but that is we had the knowledge of how you find a quiet site where the recordings will have maximum value and we knew a bit about collecting data from worldwide networks. But I don't think, I don't remember that we had any last word, I really don't. It was never viewed as a surveillance system, never viewed as a monitoring system for the Soviet Union by AFTAC. We already had what we considered a better system for monitoring the Soviet Union. Better in that, you know, much more thought go into site selection and geographical distribution with respect to the Soviet Union, and they had arrays of seismometers in most cases. So from the surveillance standpoint I think we viewed the WWSSN as a little bit of extra gravy, but not important. Much more important for things like studying seismicity. Also of course the data, we did think that the data would be very useful for improving the epicenter locations within the Soviet Union, and China of course, at that time. But I don't think we felt it was a contributor, a major contributor to monitoring the Soviet Union at all.
Barth:So beyond seismicity it wasn't of a big help for AFTAC 's mission in this respect.
Romney:It was helpful for location accuracy. If you had events large enough to be detected by these stations. We were detecting smaller events and then were detected by the WWSSN, but for those above the threshold of the WWSSN the data were very useful for improving location and obviously there are locations like in — areas like in — Southeast Asia where the very proximity of a station, say in Thailand, made it contribute some to surveillance of the Soviet Union. But…
Barth:While you mention this, I mean looking at the map of stations, the world map and then looking at the stations of the WWSSN, it's kind of striking to see that of course there is a high density of stations on the southern border of the Soviet Union. So I always asked people like Carl Kisslinger, there must have been more involved than just seismological interest. But he said, "No, that basically was it. We asked station managers around the globe to send in suggestions for research programs and also had the best ones receive the stations basically." But I mean, you know what I mean, this strong, strong density around the southern border.
Romney:Well, your eye sees that in part because the Soviets didn't participate, and so you have a big void above that zone. But yeah, I think he's right. We bent over backwards to avoid the appearance of surveillance, and I know that was a major concern to people in the Coast and Geodetic Survey, and they, when they went to contact people in person they went to great lengths to try to assure them that this was not a spy network, it was a research network. And there were countries that were skittish about it for a long time. India I think was one, and I don't know if India ever got a WWSSN station. I suppose I could look at the map and find out. But I don't know if they ever did participate.
Barth:The French did not for a long time.
Romney:The French didn't. The French had their own network. And the Canadians didn't, they had their own network. The Russians were given a set at a fairly early time, of equipment, which I believe they installed in Obninsk and I think it's still there. I believe I saw that the last time I was in the Soviet Union, the early '90s.
Barth:Maybe we can talk about some of the other system developments in which AFTAC was involved. For example, we have developments in deep hole seismographs, we have developments in array systems, these are two which come to mind, then we have of course ocean bottom systems and the unmanned systems. Would you like to comment on any of these?
Romney:Yeah. The Berkner Panel had identified the construction of a Geneva — so-called Geneva Array — as a very high priority item, and that was under AFTAC’s management. We in AFTAC already had a good site at Fort Sill, Oklahoma, and that had been installed possibly originally for, well, part of its function might have been to monitor the Greenhouse explosions, but it's a little fuzzy in my mind right at this time. Mainly that was a test facility at Fort Sill, Oklahoma, where we could try out ideas. But it was a good, quiet site for that locale and a good base of operations already existed so we expanded that into the first Geneva type array, meaning a 10-element array at that time. We went on and arranged for site surveys for other locations and began to select — oh, and we called it the Wichita Mountain Seismological Observatory. We named it very early on, and set up that observatory nomenclature as identifying these higher quality stations that had arrays as equipment. ARPA was becoming fairly active in the program. Charlie Bates had been hired as a program manager, and I believe it was Charlie who named the others that came along.
Romney:Cumberland Plateau, Blue Mountains I remember
Romney:Yeah, Uinta Basin.
Barth:And Tonto Forest.
Romney:Tonto Forest. And I think he, but he followed this pattern of calling them observatories, and I think AFTAC had a little say in it, but I believe Charlie kind of felt that that was really his prerogative and I guess it was, and he controlled the money. So that was an important program. The grand opening at Wichita Mountains was attended by [John A.] McCone I think who was then chairman of the AEC and other — we had at least letters of congratulation and things like that — from other people. It was quite a nice affair. And so were the openings of the other observatories by and large. They were done with a bit of fanfare and attention, and it was a good program. Each array was a bit different. We didn't make identical layouts of the instruments. And that was in part because the people involved in the design and development of these had different ideas. Different contractors had different ideas and suggested things, and so that seemed well within the spirit of the program, and so these arrays looked different; their maps looked a bit different.
Barth:What do you mean by different approaches by the contractors? How would you describe the difference between …?
Romney:Well, for instance I remember I think it was Texas Instruments people thinking that you should do a kind of logarithmic array. The spacing would increase with distance from the center of the array. And I don't remember the patterns of the array layouts anymore, but I think there were steps in that direction taken to test this idea from antenna theory that was exemplified in high measure with the design of the LASA [Large Aperture Seismic Array] which was, the rings, and the major rings were logarithmic. The program at Wichita Mountains proved to be a very, Wichita Mountains proved to be a very good place to carry out tests. It was fairly convenient to Geotech, who was the principal contractor from Dallas, Texas, and we were soon testing ideas about burying the seismometers and at greater depths. The original installation were in vaults that were 2 or 3 feet down into the bedrock, and even the construction of…
Romney:[We used shaped-] charges, for instance, to excavate the holes for these tanks and were trying to find economical ways of doing it. We found that the short period noise, the high frequency noise, was very dependent on wind. That was easy to demonstrate there. And so the idea came, well let's bury the seismometers deeper, and I think that was the first buried array, deeply buried array, 50 feet maybe, or maybe it was 50 meters. It's been a long time and I don't remember, but going down to that depth made a great deal of difference when there was any wind at all, and we could easily demonstrate it by leaving the shallow ones in for a length of time after the deeper sensors were installed. And so we showed the importance of getting down away from the direct influence of wind. I don't think the wind was coupled directly into the seismometer because the tanks were well sealed, but the energy from the wind was coupled into the superficial layers of the rock quite well, and apparently trapped there, pretty much there. And that was demonstrated. So that was a productive place, and I think the whole program was productive. We also determined the effective threshold of detection of a network of that nature, and we could show, you could — while you couldn't do high precision epicenter locations with that network — you could pretty well show that the threshold of detection was down around magnitude 4 or even a little bit lower at teleseismic ranges. So it paid off in a lot of ways. And Fort Sill also is where we evaluated the Russian seismometers. Two parts of it, and I don't remember the chronology exactly. I believe that the SVKM instrument that became a contender — or was what at least Pasechnik advocated for the detector at TWG 2 — we had one constructed for use during TWG 2 and were able to show the Russians that in fact the Benioff had a much better response as far as detection of the first motion is concerned, high frequency signals at long range.
Barth:Was this an actual SVKM or...?
Romney:I think this was simulated in this case, but at one stage we had a captured SVKM.
Barth:A captured one.
Romney:Yeah. One that we were able to buy from an Eastern European country, that the CIA was able to buy for us from an Eastern European country. So we had a real one, and then I believe the one that played a part in TWG 2 was simulated. And this, by the way, is the way you should display the response curves [Romney points to a graph], and not the way Pasechnik did when he had this, two curves. Not this way, that's not the way. Because what controls the magnification you can get is the microseisms at longer periods. So it was no surprise to us, since the gain is set out here [at periods of 5-10 seconds], it was no surprise to us that the Benioff was superior to the SVK in detecting small signals at long ranges and the first motion. I'm not sure the order and the sequence in which we did this, but we had also tested, at Fort Sill we also tested an actual SVKM instrument. So it was a first-class test bed.
Barth:What was the origin of the deep hole seismometer research?
Romney:That was my idea. I think I can fairly claim that. In the working hypothesis we had at that time was that the short-period noise was mostly Rayleigh waves, and the idea came from work that had been done during World War II to try to track hurricanes. This work, a lot of it was done at St. Louis University and they built what they called tripartite stations. They were stations with seismometers laid out in a triangular pattern about 2 or 3 kilometers apart. And they correlated the microseisms as they traveled across the array and got a bearing to the source, and they used that to try to advise the Navy of the existence and the location of hurricanes. It was not a very successful program. As we learned more about microseisms we began to understand why. They are not generated with the storm is over deep water. They are generated in shallower water. But anyway, from that work it was pretty well established that, at least at that 6, 8, 10 second period, that the microseisms were Rayleigh waves. I mean they traveled at Rayleigh-wave speeds. And so our working hypothesis was that even at shorter periods there was an awful lot of Rayleigh-wave noise. Now we knew about wind noise also, but that is very locally generated motion in the ground induced by local winds. We knew about that, but we thought once you got past that, the underlying noise even down around 1 second was very probably Rayleigh waves. So knowing that Rayleigh waves decrease in amplitude exponentially, sort of, with depth, it was a natural hypothesis that if we put the instrument down a sufficient depth, on order of several thousand feet, considering the wavelengths that we expected, that you could improve the signal-to-noise ratio. The first experiments were well before Vela came along — well-meaning a few years before Vela came along — but they were not long [before], they were probably right around 1956, not very long after I went to work at AFTAC. And the work was carried out I believe by Texas Instruments, the original work. They used a low frequency geophone, prospecting-type geophone and found an open, well, a hole that they could use, a dry oil well that they could use, and did some tests which we felt were not definitive. We didn't get much signal-to-noise improvement, but I believe the…when it was really looked into, our belief — that is mine and I think Ben Melton who was the principal engineer working in the seismic field at AFTAC — we believed that the instrument's internal noise level was too high, that the amplifiers were not adequate to make measurements at the levels that we needed to really investigate this. It was fairly expensive. I think that program might have been around $200,000, which was a lot of money for those days. And the major program did take place under Vela funding and was carried out by, I believe by United Electrodynamics, and subsequently by Geotech. And the results are astonishingly variable, really surprisingly variable. We found examples of more than 20 dB improvement in signal-to-noise ratio, and we found other examples where there was very little improvement in signal-to-noise ratio. And we finally developed the idea, or at least this is my recollection of what we developed, the idea was that there is a component of the noise that is Rayleigh-like trapped energy near the surface, and the deep well is good in those areas. But there are probably other areas where the ambient noise is — behaves like P-waves and like old worn out P-waves from very distant earthquakes or generated by wind over large areas or something like that. And so nothing seemed to work very well for that noise. The array work had similar kinds of results. The arrays were effective for some kinds of noise, but there seemed to be an underlying component that just coming in maybe vertically and hitting all the sensors simultaneously, and therefore you couldn't do anything about it. So our noise model was consistent that there was some, that the noise is more complicated than just Rayleigh waves and that there is some other component that…
Barth:So the high frequency noise is not the problem. The low frequency noise is the problem, the long period.
Romney:Well, no. Even around 1 second we found this result. That we could make a great improvement at Hobart, Oklahoma, is one of the places I remember, and there was certainly 20 dB improvement in signal-to-noise ratio going down that well.
Barth:For 10,000 feet.
Romney:Ten thousand feet, yeah.
Barth:And later on for the larger array, like LASA, 10,000 feet was not used. I think it was just a couple of hundred feet or?
Romney:Yeah. We were just trying to get away from the wind noise really.
Barth:Because there was no way to beat the other Rayleigh noise or...?
Romney:Well, also expense, but I think that the objective, the hope of that array was that you could do this elaborate signal processing and wipe the noise out anyway, so you didn't have to go deep. You didn't want to be so shallow that you were inviting trouble from wind, but you get away from the wind, the very local noise, and then you would cope with the rest of it by signal processing.
Barth:How successful do you think the deep hole research program was, in retrospect?
Romney:Not very successful. The fact is that we did drill deep holes at one or two, maybe a few, of the AEDS stations. One of them is at Pinedale, Wyoming. I think the hole is still there, and could be used for experimentation, but it was difficult to operate and it was dangerous, I think the Air Force felt it was dangerous to handle, winches and the seismometer, the Benioff seismometer, the stretched-out-Benioff seismometer made to go down a deep hole weighed a couple hundred kilograms and by the time you have 2 miles of cable and…
Barth:It's very heavy.
Romney:Yeah, it's very heavy and winches are notorious for catching fingers and things like that. So I don't think they liked to do it, and the signal-to-noise gain we were getting in locations like that, frankly, was not very much.
Barth:So not worth the effort in the long run.
Romney:Not worth the effort in the end, yeah.
Barth:So this whole program of 10,000 feet deep holes is basically nonexistent anymore.
Romney:That's right, yeah. Now I'm not sure that that's entirely a wise end result, because I think it's quite clear that you can do things at longer periods. And how important that is I don't know, but I think you could, I think you could probably do things at, oh, 2 seconds to 10 seconds it would make a difference. We have tended not to do very much research in that range because the way the microseismic noise spectra are shaped, we have a hole, a dent in microseismic noise level at 10 to 50 seconds. We look there and we look around 1 second because we have noise spectra drop off greatly toward 1 hertz, and attenuation wipes out the signal beyond a few hertz, so we tend to focus in these two bands where we can do better, and we can, but I wouldn't be surprised if there's information in the intermediate period range that could be made worthwhile. Now an example is the Rg wave that seems to be a fairly good discriminant at regional distances and Rg tends to have a period of, oh 2 seconds is a reasonable number to draw out of the hat, and I believe you could enhance that with deep-well techniques. You probably could with array techniques also if you made the arrays bigger than what we tend to do these days. But so anyway, we have not done any major amount of research in that period range from 2 to a few seconds so I don't know. I think if we had a real good reason to want to get 2-second or 4-second information, I think it would be worth rethinking the idea of deep wells or deep-well arrays or that kind of thing.
Barth:What about other projects which were geared towards improvement of signal-to-noise and worldwide coverage? I think of course about the ocean bottoms.
Romney:Yeah. You mention the ocean bottom, and we in AFTAC did some work on that. One project was to put an instrument on the bottom, essentially connected to a buoy at the surface which would radio information back. And I can't tell you whether we actually improved signal-to-noise ratio with that or not. I tend to think it didn't show very much gain, much improvement. What I remember most vividly was that that was a terribly unsuccessful project from the engineering standpoint. Cables dangling from buoys and tossing sea waves have a very bad habit of wrapping around themselves, tying themselves into knots and breaking and doing all kinds of things, so that was…
Barth:So it's an engineering nightmare basically?
Romney:That was an engineering nightmare. And I've talked with others who have had the experience mostly in the ASW Anti-Submarine Warfare] world and they'll tell you the same thing, that's a hell of a way to try to operate.
Barth:The major hope was of course to get global coverage, because 70 percent of the globe is covered by oceans. Has this been still done after Vela? So did Vela there provide something like a model for later systems?
Romney:Not as far as I know. I think there were some successful operations, one by Columbia University I believe, off the west coast of the United States with cable coming onshore, and it produced some promising results, but I think it must have been very costly to operate. And we in AFTAC had tested some ocean bottom seismometers I believe made by Texas Instruments but I won't swear to it. You dropped them in the ocean, they sank to the bottom, recorded data, and then upon receipt of a sonic signal, coded sonic signal, and they cut loose their underpinnings which were iron feet or some such thing, iron, heavy cast iron cradle or however they were. It was something like that. And then they would pop to the surface and start blinking lights and road-casting direction-finding signals for you to pick 'em up. And we did, we dropped those off the Kurile Islands in an experiment, and I believe there were some explosions that were conducted along with that, some depth-charge type things, a few hundred pounds probably, and one of the objectives was to look at the question of location bias, which we thought might be associated. I think [this] was after LONGSHOT here we discovered an enormous bias in the location, and we took that to relate to the complicated structure associated with the dipping plate down under the island arc. And so we were looking for that in [the] Kurile Islands, but also trying to look at the earthquakes from the seaward side instead of the landward side for discrimination purposes and location — improving location accuracy. That was a, we got some information out of it. I couldn't tell you what it was anymore. It was not important enough that we continued the program, whatever it was. We had an Air Force officer on board the ship, and the ship I imagine was a rented ship, and a heavy storm came up and they had to take shelter behind the — we notified the Russians that we were going to carry out this experiment. A heavy storm came up and they had to take shelter behind one of [the] Kurile [Islands], inside the Sea of Okhotsk, which the Russians claimed was their internal lake at that time. And there was a quasi-diplomatic incident — well a minor diplomatic incident associated with that. The Russians knew about it and didn't like it. And I can't tell you the story anymore. It was not a firsthand story. Colonel Ridenour, who was the man, worked for me at the time and I know it must have been one of his great adventures in life, because the Russians were not very friendly. Although it was after SPRAG and there was some cooperation with it at the time. But by and large I think AFTAC’s involvement in that was simply managing some programs where someone else was the source of the enthusiasm for it. And ARPA was enthusiastic enough to put up the money. Not much came out of that, and I don't remember regarding that work as being mainline as far as our efforts in systems improvement.
Barth:Maybe we should clarify the interaction of ARPA, AFTAC and industrial contractors. ARPA was the funding agency.
Barth:AFTAC was one of the major
Romney:Program agents, as they call them.
Barth:Program agents. So there was AFTAC, AFOSR [Air Force Office of Scientific Research] and AFCRL [Air Force Cambridge Research Laboratories] as the three major ones?
Romney:Those were the major ones, yeah.
Barth:And then there was also DASA others I guess. ONR maybe with some countries.
Barth:But AFTAC was probably the biggest in terms of contract money as far as I can
Romney:Yeah, I believe that's right.
Barth:So how would this work? ARPA would write an ARPA order, send it to AFTAC, and AFTAC would then be able to choose the contractor to work on this project?
Romney:That's the way we would have liked to work, and we did to some degree, but by and large as I remember it we put together a program package and a paragraph long description of the work and identified contractors and funding amounts and sent that to ARPA. They would approve it or part of it, and then initiate an ARPA order authorizing that work to be carried out. That's the way it worked for the most part. I believe there were some times when they sent us an ARPA order with some money to do something that we didn't want to do, and I believe there were occasions when we — I don't know if the word is refused, but in essence refused. We probably persuaded them to withdraw it or send it to someone else. And I really can't think of a specific incident, but I know there were cases where they did ask us to do some things that we didn't much want to do and maybe did to be nice guys or keep up good relations with them but didn't think it would be very productive. And probably we were wrong in some cases.
Barth:So ARPA had at this time a staff of I think only six people working on Vela Uniform with Charles Bates. There was Donald Clements, this name comes to mind. How many people were working on the AFTAC side?
Romney:Rudy Black [with ARPA], oh, probably a dozen [in AFTAC].
Barth:Okay. So even more than ARPA's own staff.
Yeah, that's right. But we did some data analysis, which is something they never did in ARPA, which gave us an advantage in program management because we really knew, we had people who really understood the meaning in seismological terms of what was coming out of these programs. So we had not only people who were program managers, people with let's say a college degree and some management experience, technical experience, but we also had some more junior people who could read seismograms, identify the phases, understood noise and noise measurements, and so we kept involved in things like putting together the information. Well, let's take the Long Range Seismic Measurement program as a good example. In that program data were collected on individual explosions and a shot report was produced containing the measured data from that explosion. Now we in AFTAC then would take, let's say the magnitudes, and we would put together the data that gave us the magnitude versus yield curve from lots of shots. So we did a data integration function over these separate projects in many projects. And part of our data would come not out of the Long Range Seismic Measurement program but out of the AEDS itself, and we'd compare results, combine results, and we kept ourselves in a position, I think, of knowing more than the contractors did about some ac-[abrupt end of Tape 4 of 4]
Session I | Session II