Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
We encourage researchers to utilize the full-text search on this page to navigate our oral histories or to use our catalog to locate oral history interviews by keyword.
Please contact [email protected] with any feedback.
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.
In footnotes or endnotes please cite AIP interviews like this:
Interview of Syukuro Manabe by Spencer Weart on 1989 December 20, Niels Bohr Library & Archives, American Institute of Physics, College Park, MD USA, www.aip.org/history-programs/niels-bohr-library/oral-histories/5040
For multiple citations, "AIP" is the preferred abbreviation for the location.
Informal, unedited, exploratory interview. Covers Syukuro Manabe's career, and the work of other groups, in computer numerical modeling of the earth's atmosphere and climate.
Try identifying yourself again.
My name is Suki Manabe. I work at Geophysical Free Dynamics Laboratory of the National Oceanic Atmospheric Administrations… What shall we talk about first?
Well, let's start just by, if I can ask you just to tell me what some of the main events and people and institutions have been in your atmospheric field, and let's start with the tree when you got into it first, the tree is when you started doing your radiation work, up to let's say the famous Manabe and Wetherald paper.
Now, originally, the first person who talked that, who said that the carbon dioxide increase, particularly from industrial activity, may increase the atmospheric temperature, I think was Callendar. I think Callendar.
Callendar, yes.
But there was another one, Arrhenius.
Arrhenius, yes.
You know these.
I'm asking about from the time when you first came into this field.
Yes. So I continue that, because that, Callendar's and — he had an excellent absorption (?) and so forth, but Callendar and then there's the people like Plass, you know about Gilbert Plass, and Lou Kaplan, and Kondratiev, of the Soviet Union — they came up with all of these series of greenhouse warming, the so-called back of the envelope calculations. And what they do is, you have more carbon dioxide in the air, and then this will absorb more radiation from the surface, and send back more radiation downward, and —
Was this a concern already in the fifties? When you started atmospheric modeling, you didn't put in carbon dioxide in your early models?
Yes, I did, yes. From the beginning.
Right from the beginning?
Right from the beginning, yes. And so now what happened is that because these theories all flowed — so my friend, very close friend of mine, Fritz Muller who is pioneer of calculation of radiative transfer in the atmosphere, they evaluated this, and he included this effect of water vapor. When the temperature goes up, more water vapor in the atmosphere. And he put this effect in, and he studied it in all kinds of crazy results. Now, one of the basic problems about these studies is that, you see, you know the Kirchoff's law.
Yes.
Now, Kirchoff's law tells you that absorptivity and emissivity (?) have to be equal in order to maintain thermodynamic equilibrium in the black body box. So you know, the fallacy becomes very clear if you think about it. If the atmosphere absorbs more radiation from below, you have to emit more. So how you can say that just because it absorbs more, that it's emitting more too, so how you can say that that will make it warmer? And the problem is, people until my time were all looking at radiation balance of the ground, rather than the atmosphere, yeah, the atmosphere as a whole. And so when you think about this whole system, the radiation balance, it becomes evident that just because you absorb the radiation from below, that doesn't mean that that will make it warmer, because it will emit more as well. Right?
Well, you know, Tyndall back in the 19th century had what I thought was a good analogy. He said that carbon dioxide was like a dam, and the water had to pile up behind the dam.
Yes. Now, the thing is that people tend to get better explanation of carbon dioxide warming when they think about Venus. So the base calculation is the radiation which coming from the bottom of the atmosphere is re-absorbed, so the more greenhouse gasses you have in the atmosphere, if that is the source of emission, go to higher altitude, where the temperature is lower.
Right.
So that the more opaque the atmosphere is, the higher the effective source of emission, therefore emission into space would be reduced. Therefore in order to get the same thing you have to have an increase of temperature. That's a much better explanation. And then Venus has a very thick atmosphere, so it's (?) down, then you get a very warm temperature.
Right. And this way of putting it, you say, wasn't really understood until around the time of Plass and so forth.
No, Plass and all his people did it all wrong, because they didn't start with heat balance. Until Fritz Muller. So what I did was basically try to straighten this out by radiative convective equilibrium including this convection and emission and all these things. So I was the first one to develop combined radiation and convection, and then clarify this effective source of emission issue, and also discuss what water vapor does to the outgoing radiation emissions. The water vapor will have an amplifying effect.
This was after you had joined Smagorinsky.
Yes. So that is what basically the 1967 paper. See, the Fritz Muller extrapolated the same emissibility as Callendar, and then discussing surface radiation values, and he said, he can get anything he likes depending upon temperature. He can get a cooling, warming, he can even get a cooling, and so Fritz Muller and I discussed, we decided, the only way to go about it is to go to radiative convective equilibrium, and then change carbon dioxide and water vapor, let water vapor change according to cloud (?), and then just see what's the effect of that on the temperature of the planet.
So this was —well, let's see, this was the groundwork for your paper with Wetherald.
Yes. '67.
'67, the famous paper. Now, up until that period, was there any work like this going on anywhere else?
No.
Other than within, well, Smagorinsky's group I guess.
No, at that time, you know, so the people there, Kaplan is still — you know, he's active, and Plass was very active, and Fritz Muller, so that at that time, but everybody doing surface heat balance. Now, the thing is, if you do surface radiation balance, you have to discuss how heat is exchanged between surface and atmosphere. Right? And the way to see it is not only exchange through radiative transfer, but boundary exchange. And so then, in order to discuss how much heat is exchanged between ground and atmosphere, you have to discuss heat balance of atmosphere. Right?
Sure.
And in order to discuss heat balance of atmosphere, you have to discuss the radiational heat exchange between atmosphere and space.
And between atmosphere and ground.
Yes, ground. Now, fortunately ground and atmosphere are very closely summary coupled, so that's why my simplistic explanation of Venus atmosphere works very well, because surface coupled by boundary exchange, so all you have to see is look at the exchange between atmosphere into space, and move this temperature back and forth, and that comes pretty close to what I did in radiative convective equilibrium.
Right. Now, if you wanted to go back to the early sixties and ask what work was being done on climate —
— yes —
— aside from models, what would you say was the main work that was being done?
Now, what is happening here is that at that time, now, von Neumann had this weather prediction group, headed by Jules Charney. And one of the activists (?), Norman Phillips, they worked very closely with Jules Charney, and I think Smagorinsky was a very young scientist at that time. He had written some nice recollection of it. George Platzman also wrote a very nice essay, so you can go — George Platzman can send to you.
I don't think I've seen the Platzman —
— yes, Platzman is very good on that, so you can ask George Platzman and Smagorinsky to send you.
Yes, Smagorinsky's I've seen, I haven't seen Platzman's.
George Platzman also put a paper on the (???) of AMS. So that's Chicago, get him to send you that historical essay. Or he can tell you which volume so you can get it, either way. So that started from late 1940's, I think, and Norman Phillips published this historical paper in a 1956 or something which described the first general circulation model experiment, which have a two layer model, and demonstrated he can get the jet stream, just like real atmosphere, and so that is, I think —
— this was still very much within the modeling world. Was there any interaction between that and the old school of climate people who just studied —?
No, there's very, very little interaction. I think this is a very interesting thing, that even weather, weather forecasting, there was a lot gradually, you know, getting to 1950, there was a lot of strain between the people who are doing forecasting by traditional methods, and by using the numerical weather — so that in a sense that, we cannot deny that the people who are doing forecasting made a tremendous contribution to our appreciation how atmosphere looks like. However, there was sort of a competitive relationship.
A philosophical difference.
A difference, fundamental, unreconciled, so that weather, and the same thing applied to climate, I mean, Norman Phillips is interested in the jet stream and so forth, and so that, there wasn't much interaction between traditional climatologist who was more interested in statistics, climate, and this... — (?) doesn't try to understand why climate behaves that way.
We feel that this may be a general characteristic of geophysics, that reached in this case not only meteorology but in many cases with the old geologists.
That's right.
They had very similar problems.
Yes. So you know, the idea is, always sort of new things have to sort of break ground. So that Norman Phillips 1956 paper is truly pioneering paper. And then, Smagorinsky, I think von Neumann thought that this weather forecasting model which he sort of instigated under Jules Charney, that can be applicable for the study of climate, as Norman Phillips nicely demonstrated. So he thought, there must be somewhere, there is a new find needed that can see all these and these two gentlemen's articles. And so he established, Smagorinsky in the article described very clearly, and so he established a so-called at that time called Weather Bureau, this small climate simulation group.
Right, the ancestor of this laboratory.
Yes. So that's what happened. And then we also had, at the same time we had a competitor. UCLA Yale Mintz. Yale Mintz and NCAR, at that time he was at Lawrence Livermore Laboratory under Edward Teller, and Cecil Leith. So we had this nice competitive relationship.
Beginning about when?
That was about the time I came, so it's about 1958.
Already then they had models at Lawrence Livermore.
Yes. Now, Smagorinsky, one of the important things is that the Lawrence Livermore Laboratory's model, it was very nice but the results were never published at that time.
Yes, because I haven't seen the —
— no, were never published. There was some instability in the system and he was suppressing with very large viscosity, and the chap realized it later, but it's —
— he was suppressing the viscosity so that it tended to oscillate too much?
No, damped his wave too much, too much viscosity, swamped with viscosity. Because the instability, so that you know that, it's like chemotherapy, you put too much chemos and you may kill the —
— the whole thing.
Yes.
I see, it was over-damped.
But one thing which comes out of this competition is that we came up with the, Smagorinsky came up with this moderal primitive equation, we call it. It's sort of more like Navier Stokes equation, which sort of doesn't rely on this geostrophic approximation, we call it, and because of this, it was very easy to go to the putting hydrologic cycle, water cycle, in, and about that time, I was collaborating with Smagorinsky, and we had the first (?) saturation (?) model with hydrologic cycle.
This was in the mid-sixties?
Yes, '65. We published '65. First with hydrologic cycle.
There was a paper you did with Strickler also.
That one is a one dimensional one with Strickler, that is a follow-on of this Manabe-Wetherald paper. But I think when it finally quit was water feedback. But —
— we don't have time to go into all of that now.
But 1965 paper, we gave, Manabe-Smagorinsky-Strickler paper, we put the first hydrologic cycle in. And I think this is a milestone. On the other hand, our competitor, UCLA, people are able to put the realistic geographies.
Who was that again?
Leo Mintz and Arakawa.
Mintz and (?).
Arakawa. So these are the early stage of the thing, and after that, of course, we started putting all kinds of things. And 1969, Bryan and myself were the first to combine the ocean and the atmosphere, and that was at that time, it looked a little bit like Don Quixote, but they told, a good thing we did, and it's a pretty good paper. It is Manabe and Bryan 1969, in the JOURNAL OF ATMOSPHERIC SCIENCE.
It's widely cited. In fact, I have it here.
Yes, that's I think one of the best papers we wrote.
Right, it even had the ice, cryosphere.
— yes, we got a little bit muddled when I thought (or, I thought ice wasn't working?) recorder? wasn't working but anyway we had it. Anyway it was very interesting. So that was the first one, and then gradually go on and —
When did other competitors begin to appear?
At that time, still these three groups, that is, UCLA and Livermore and here. We are competing with each other. And I look back, and Chuck went to National, NCAR, National Center for Atmospheric Research.
Who went there, you say?
Chuck Leith. Left Lawrence Livermore, went to NCAR. But then Chuck about that time, he quit doing it, so he got out of competition.
I see. Is the current CCM descended from that?
No. NCAR didn't have, I shouldn't say too much, but they didn't have too much — now, after Chuck moved to NCAR, Kasahara did a run-through (or, Kasahara did too?) so it's now a sort of three way competition between, among the NCAR, the GFDL, and the UCLA. I don't know competition, I shouldn't call competition, but we have a nice relation.
When did NCAR get started would you say?
I think middle sixties, '65.
There's also the UK Meteorological.
That's much, much later, much, much later. So —
So for a long time it was only the —
— yes, so, then, coursing along, all we are writing, we all started writing lengthier, lengthier papers, and to show how well we can simulate the climate, and until 1975, then we published first the three dimensional model applied to carbon dioxide issue.
Right, this was again Manabe and Wetherald.
Yes. That's 1975. That's another milestone paper, because we got into three dimensional modeling rather than radiative convective models. Now, we started doing it —
— also you see the differences between the Poles and the Equator.
That's right. That's 1975. So that is more or less the — and then, the coupled GCM model, (?) and I have a lot of problems.
The what model?
Combined Ocean Atmosphere, GCM. We had a lot of problems, even though we started from 1969. As a matter of fact, that '69 paper is the only jewel, and we keep producing some, in our own opinion, kind of lousy papers about coupled GCM model.
I want to ask you about that when we get to talking about your latest papers.
And then, so that until early — there was a paper 1982 where we first finally we decided to apply a coupled model to the study of Greenhouse warming, and before that, Jim Hansen's — of course, 1984, he had an '84 paper, but he was discussing some kind of a — Jim Hansen and other people like Marty Hoffert and various people, I think we can look at them more carefully, but they are looking at sort of putting always in saturation in, some kind diffusion ocean downward, so using diffusion equation, and the heat is diffused down.
I guess I should ask, talking about Hansen, when did the Goddard people start coming into the picture?
I think somewhere around 1970, they started working on the Greenhouse warming, maybe 1977 or '78.
Now, you still say Greenhouse warming, but you make it equivalent, working on climate models and working on Greenhouse warming?
No, it's not necessarily, because you see originally, for example, the only original people thought that it was stupid to try to study this Greenhouse warming issue by three dimensional model, because it cost so much computer time.
I see, and you can find out all you need with the two dimensional.
One dimensional one, one dimensional model, or —
— all you care about is the average temperature.
Yes, why bother with, go to these three dimensional?
So people really didn't have a feeling for the importance of the feedback.
Yes, they kept telling me, how dare you do this three dimensional model? Oh, you can do it, you have so much computer time.
But the importance of ice albedo for example was known from a long time back?
Yes. Now, this ice albedo feedback is another interesting thing. So as we developed this three dimensional model, also there is sort of one dimensional model going one, one dimensional model mine, that is the radiative convective model. However, the scientist Budyko, my friend, he came up with one dimensional model in the in the north (?) and then, atmospheric heat transport effect, he expressed it simple Newtonian damping or something, and then he discussed the interaction between you know, diffusions of heat by atmospheric eddies, and ice albedo feedback, and he was the first one to discover that you may have a, more than one climatic (?) area (?) such thing as, you may have present comfortable climate as well as ice covered climate.
Back in the sixties, I guess.
Yes, '69 paper, famous '69 paper, so Budyko came out, so, when I first talked about this '75 paper in the meeting in Stockholm or something, and there was sort of a competitive relationship between the simple one dimensional model, no (?) one dimensional model, and the three dimensional model.
And that persists to some extent to this day, the competition between Budyko's approach and —
Yes. Now, Budyko's approach is a little bit different. He, actually even though he's a pioneer of this beautiful simple one dimensional model which elucidates essential mechanism involved in albedo feedback, that he is also have advocate another methodology.
Climate history.
Yes, and then try to sort of predict the future climate change based upon anecdote method, so that you look at past the warm climate and try to see say that similar, if it's same warm climate, why shouldn't we have the same thing.
We were starting to talk about this a little bit when I first came in.
Yes.
What do you think have been, not just Budyko but also American and so forth climate history people that have had some effect on your work, or that you at least felt that you had to pay some attention to? Have there been particular groups in that?
Climate history?
Yes, people who have done, call it climatology, people who have —
— so what happened in that, in 1970 Joe Imbrie who is a professor at Brown University realized that meteorologists are doing all kinds of interesting things. Maybe they are picking in the back but anyway we are trying all kinds and this climate model, modeling approach to the climate seemed to be a very promising one, and so I in 1970 — we had, there is a National Academy Report on what's called understanding of climate change, and at that time we get together and we map out how we can study these climates, past and present and future, by using these models, and how we can cooperate between geologists and ourselves. And one of the persons who pushed this the hardest was John Imbrie, and in the atmosphere there's a guy named Larry (Lawrence) Gates (?), and John Imbrie approached me first. But I didn't jump up and down, so he thought that it would be better to — and so they collaborated with each other, and then sort of, geologists traditionally tend to look at the deep sea core or something like this and look at the time history of the core and so forth, but they introduced a sort of new concept that is sort of, why not map the world climate? (CLIMAP Project.) For example, distribution of world climate. So that they model is three dimensional, and so that they, can we understand this snapshot of climate? That means 18,000 years ago, when the Ice Age was at the maximum. Look at the snapshot, and then look at this, and then see, apply that boundary condition to the modern. Can we understand that consistently with the performance of modern with model?
Right, and another planet too.
That's right, to look at. So they have this marvelous technique, regression analysis of John Imbrie's, and to determine the temperature distribution from the taxonomic composition of fossils, cells, fossil plankton, I guess. And so that — so there was John Imbrie's personal charm, and he was able to assemble very diverse group of geologists and egos, but he was successful sort of you know, people made an effort, in the common objective, and the idea is that always they have model, three dimensional model in mind, so that instead of sort of just getting a piecemeal thing, but always try to map the thing, and so model was very useful, had very useful influences in planning climate, even if we didn't use it that well maybe, but at that time of course, what they did was, they put ocean temperature as a lower boundary function, and then put... this... So what they did was, at that time, they called Sea (?) Temperature, the boundary concern, and then university man George Denton and Hugh, they had this nice ice sheet distribution, which made a little bit too much but anyway, too much buoyant there now in the prospect but they put this ice sheet in there, and then they also look at land pollen, from the pollen they determined the surface, the air temperature, things like that. So then what they did was, give ice sheet and sea surface temperature, can you use GCM to reproduce land temperature data? Now, so this was really the beginning of interaction between climate models and paleoclimatologists.
I see. Before that there just hadn't been enough to really — had there been any interaction in the sense that some of the early paleoclimate people had showed an interest in very rapid changes and the ability of the climate to shift from one form to another?
Yes, now, this is — so, originally, so that it's much more like, originally we — what we are doing at that time is more of the testing of atmosphere model. However, the best data you can get about the (?) for example or the climate change from that time is, what are the deep sea cores in the ocean? So, you know, you can't just play games just looking at the atmosphere, general circulation model.
You have to include the ocean too.
Yes. Unfortunately, at that time when John Imbrie approached us, that our ocean atmosphere model wasn't good enough, so, the idea of, instead of giving his average temperature, as a lower boundary function, what we would like to do is to put the ice (isotopes?) (ICP?) in and then maybe see how the ocean disposed to that icy (ICP?) distribution.
Because you need for the isotope measurements, you need to know —
Yes, and then you see, so that the climate temperature, instead of an input, it should be a (?) tool, the temperature should have come out of water.
I see, because so many other things are involved. The usual problem.
Yes. But unfortunately, of course, since that time, the formal chemical isotopic analysis of deep oceans are rapidly accelerating.
Yes, let's talk a little bit about the late seventies and eighties, up to the last —
— so the biggest excitement is that the people begin to see in for how ocean was behaving during last glacial maximum, or Younger Dryas at that time, how ocean was changing or behaving at that time, by looking at doing a chemical analysis. And of course they can also look at not only plankton foram, they can look at benthic foram. Or they even now begin to say by looking carbon 14 analysis of planktic and benthic, compare these two, at the same time, and then you can see that what is the age of ocean water at that time, and they begin to do that, because of accelerating mass spectrometer available, they can do these things.
What are the names that you associated with that?
They call it AMS.
No, I mean the people.
You mean, I'm telling you the Wally Broecker, and of course, the — so Wally is a sort of leader of the chemistry, isotropic analysis, and the U.K., a guy named Shackleton. And he had been in the forefront of this, to try to study the past history of the state of ocean, not only just the surface which climate originally went into. And so Wally is at Lamont and Shackleton is in the U.K., and in France, this guy Duplessy, who also has been doing —
What's his institution?
The center for something. Claude Duplessy. He's a very interesting fellow. So French is getting very, very active, and then, there are two, in addition there are two, these are deep sea cores, and then there's another two very active group in ice core and one is Claude Lavier and Hans Deshger (?), Laviers is in France, Grenoble, Deshger is (?) Switzerland. And these are some of the key people who, leaders who sort of stimulate younger people doing work.
Right, what about in modeling? What are the new players in the eighties?
In modeling, I think that a person, I think that the leaders of paleoclimate make studies in modeling is I think that John Kutzbach, and —
Where?
He's at University of Wisconsin.
He's a pollen guy?
No, he is more like meteorologist, but he is with Reid Bryson, and he became a sort of a paleoclimatic model, Mr. Paleoclimatic Model, with John Kutzbach, and John's biggest contribution so far is to debate (date?) the change in African date models, with Milankovitch, chronology, and — (crosstalk) monsoon variations with Milankovitch again. Monsoon circulations.
Oh, monsoon variations. Not just paleoclimate but GCMs in general, what are the new GCMs?
So now in 1980 then the number increased. So the middle 1970's GISS started doing GCM, Hansen's group. Before actually there was Somerville, so before Hansen, anyway, GISS — yes, Somerville was now in Scripps but he was one of the leaders started it and then Jim Hansen took over. So middle 1970's GISS got in. And then Lalie Gates, early 1970, Oregon State came in. Adopting UCLA models, Oregon State.
Where did the GISS model come from, by the way?
GISS model also adaptation of a UCLA.
That's right.
Yes, UCLA model.
Has your model been adapted anywhere?
No, the, you know, the, no, I don't think so. Of course, a piece of our model, I used, but we didn't —
One of the things I want to do eventually is trace the genealogy of all these models.
Models, yes. It's very complicated. We didn't, Smagorinsky didn't encourage too much to exposing our model.
Now, you mentioned NCAR —
NCAR models are originally made by themselves. But didn't, I shouldn't say, but didn't work that well. And then eventually they got, there is a model, what you call, instead of finite difference, there's the spectral model. The Australian group, and you gave the Australian group, guy named Bill Bourke, and Brian Hoskins, and headed by Tony Borden. They all made the spectral model, semi-spectral model. You go back and forth between (?) and spectral space. But anyway, we actually exported our model to Australia for their weather forecasting service.
Oh, the weather forecasting.
Yes. And then, Australian group came dynamics into spectrum, but all other physics parameterization, like moist convection, and so, radiation, they had from us. And then NCAR imported from Australia, very complicated, so that they have a spectral model and the dynamics and all the rest of it, the GFDL model. They have a hybrid over there. Things like that. It's very complicated.
The parameterizations would be very interesting.
Yes, very complicated. So anyway —
OK, what other models do we have?
Then the U.K. Met Office came in maybe around 1980 or so, started getting into climate modeling. The French —
What name would you associate with the U.K.?
Meteorological Office.
No, what person?
Maybe, right now the guy in charge is John Mitchell. And then French also have models, which is probably started from late seventies. Sadourni or something. And so in Paris there's a French group. And now I think the Japanese have started, the Japanese —
— only recently?
Yes. They have been on and off. They may have also several ideas now, so there's a whole bunch of group there.
West Germans?
Yes, the Max Planck group, Klaus Hasselman's group. At the Max Planck Institute. At Hamburg.
What about the Soviets?
Soviets have some model, but because of lack of computers, they never took off the ground very much. They run a little bit and fizzle a little bit and fizzle.
Not competitive?
No. Because of computer technology. That's their problem always. The same thing applies to Chinese. And so, oh, the Canadian, in Toronto, Canadian Environment of Canada, they have another group. So you know, there's — it's a pretty decent model.
I'm surprised that the Japanese are so late in this.
Yes. You know, I think that there are many bright people there, but the way Meteorological Research Institute, which probably suitable organization to do this type of thing, that the way (?) management and so forth, people don't stay very long.
I see.
They have been on and off, just like Russians, in a sense. But when he wrote it out to some other management force then start from scratch again. So that's a more or less history. So in summary, now all over the world you have models.
It's become very important.
Yes.
Now I want to go back and try and see if you can give me some help on thinking where we can find written documents for this. Now, obviously I have to talk with a lot of people, but I'd also like to know what documents might be — so let's start by just talking about what the communication patterns have been. How much have you relied on letters? How much have you relied on preprints? Let's start before the — you know, the fifties and sixties period. How much did you rely on phone conversations or meeting people personally?
'50 to '60?
Up to 1970, let's say.
I think that the '50 to '60, I think this essay by Smagorinsky and Platzman and I don't know whether Jules himself wrote something.
I'm thinking about, I want to know how you communicated with people back then.
Back then?
Yes. Were there a lot of letters sent back and forth, or did you mostly rely on meeting people at conferences? Or was it mostly just people who were right here, and you didn't really need to deal much with people outside?
It's a very interesting period, because at the beginning, of course, this Norman Phillips was doing it himself, and following it Smagorinsky started doing this, and then it became a sort of city group competition. So we have very nice competitions, which sort of enhanced each other's problems more or less.
So were you writing letters back and forth?
We didn't write too much letters because we are competing so we don't want to spill too much beans. So we go to meetings and present papers.
You present a paper, then you have a useful discussion.
Discussions, yes. And so that is the way. And we thought that to try to talk too much to each other is not very good idea, because individuality of these three groups getting lost. So what happened is, actually the degree of communication or lack of it was just right, because when finally paper comes out, that our papers all seem to be very complementary.
I understand. You don't want to all be working with exactly the same model.
Yes. If you know, you know, what next door the guy is doing every day, always green on the other side of fence looks greener. And so we tend to — it turned out that it was a very nice competition going, and usually the contribution is Arakawa's genius in the finite differences, so that they are able to map the entire globe, spherical harmonics and calculate very successfully. He developed what we call energy conserving system, which prevents nonlinear instability. And we, I think that our contribution is, Smagorinsky started this primitive equation, which enabled us to go to anywhere in the world, including tropics very nicely, and which benefits the group too. But when I started collaborating, we got into hydrologic cycle. We also first model which takes care of the stratosphere-troposphere interaction. We got into stratosphere.
Right. When you're doing all this internally, are there any internal memoranda or papers generated? You have notes. I'm wondering what may survive that would document beyond — after all you know, the scientific paper, everything is beaten out of it. You don't see the struggles that went in, when the paper appears. Is there anything that survives that would document the struggles and the difficulties?
You know, there may be some. The best person is Joe Smagorinsky. He may have some of these. I usually throw them away.
You don't have any of your old letters or notes?
No. Amazingly not. I feel, at that time I felt like I was obsessed to get the development and I always felt that the only thing I publish in the publication is the one which is left behind.
Yes, this is what many scientists feel. We often go and ask the scientists about their papers and they say they're all published. That's not what an archivist or historian means by papers. He means the unpublished.
But you know, if you are sort of very super scientist, like Einstein or something, then people come at you and then look at everything. But if you are an ordinary scientist, I felt that you know, if I left anything, nobody would look at. Only thing, if I did anything good, is what's published. Everything else I presume will perish.
This is very close to the ideal of science, which is that what really matters is the final truth and not how you got to it. (crosstalk)
Yes. You know (crosstalk) if you know you are like Beethoven or Mozart or something, and then, you know, everybody is kind enough, fight for you.
But the interesting thing about geophysics is that so much of it is a collective endeavor, and we need to understand how the community works.
Yes.
So let me take another approach. You had grants and so on.
Yes.
So you had to have grant applications and you had to have reports to the granting agencies. Do you think those have been preserved anywhere?
No, this is one of the fascinating things about this laboratory. We never, I never in my whole life — I never wrote grant proposal for my own research.
Aha, not even internally? Not even just to the director?
No, only thing is that there is the annual report. We have annual report over so many years, and there is one issue which we discuss some of the history of this laboratory, for example. The 20th anniversary report.
Aha, I'll have to get that.
So that that have some section which discuss this. Annual reports come very year, but that was a special report, so there's a special section which one of the scientists wrote.
In the meantime, when it comes time for the annual report, you will write up, what, a few pages on what you've done?
I write maybe ten page for my whole group every year.
Then that goes to the director and it gets boiled down?
Yes, and send all over the place, every year.
That's interesting, do you know whether those are —?
Yes, all available.
OK, that would be interesting.
But the most interesting one is 20th year anniversary volume.
It would be interesting to see the annual reports too, because those show what's going on right —
That's right, so I'll give you one. I'll give you one. (off mike)
OK, I could get one from the director's office probably. (Manabe off mike) ...
Some of our articles go to — CO2 review or something — before all group came in, I have some review of our work.
OK, that would be interesting.
Now it's becoming impossible to write a review because so many groups come in.
Well, this is another problem.
In old days it was very simple, because I just go out and talk. Since there was no other competition being done.
You knew everything that was happening.
Yes, and then nobody questioned the inaccuracy of my results. Now, they say, "You know, you say this, but look at this result and this one," and then we have a discussion, and the interesting thing is that it definitely causes modeller constituted differently, so that the more people come in, then the more range of results.
Because more different types of modellers, yes.
Yes, and also there is some group who is a little bit careless, very careless in design of experiment, very careless set up, and then create a number which is way out of whack sometimes.
And then gets quoted in the newspapers.
Yes. So then, so that that is why uncertainty has been increasing with time. Rather than the decreasing.
I understand.
I think, so — you know, there were so many — this is a difficult problem — I think, though, that when people get sort of loss of self-confidence or something, a loss of their confidence, they always go back to sort of most fundamental considerations. That is, how much solar radiation absorbed by planet? Then if you have a Stefan-Boltzmann rule, sigma T4, and then you calculate sigma T4 is equal to incoming solar radiation, now, if you do that, you get 250, six, seven degree, (?) (temperature?) All right. And so that would be a temperature (???) if our planet doesn't have any greenhouse gasses, that is, if you have only nitrogen and oxygen, —
— right —
— that's the temperature. And so we know that this nice climate we are now in is thanks to greenhouse gasses, including water vapor.
Right.
And so, we always felt that you need acrobatic move on the part of the atmosphere, to cool it, if you put more greenhouse gasses for some other reason, doesn't warm up and cool. You know, you know that if little greenhouse gasses, your tempered 255, and you have this much greenhouse effect, and you now got 288 —
You expect by extrapolating that it will keep on going up.
Keep on going. And so I think, when you really come to, what is the bottom line? And that probably gets back to, because you say, people say, "How about microphysics and the cloud cover?" If you change, if you have more ice crystal nucleii, condensation nucleii, size distribution of cloud droplets changes. Or dimethyl sulfide will change this. And it's defenseless. See, the reason why these models have been successful in simulating hydrologic cycles is because — because we simplify. Like if it's supersaturated, supersaturated particles of water drops the heat of condensation, and our model doesn't have any cloud microphysics. And that's why it's so successful. Imagine if you tried to put these complicated — the chain reaction, cascade showers, rain droplets —
— a lot of room for errors.
Yes. So you know, how the rain drop fall on your head, and if you try to model that, the entire process, starting from ice crystal nucleus, condensation nucleus, it's a monumental task!
Yes.
And so, and also there are the scale gaps, between the scale of the individual cumulus, the larger scale, greater scale, tremendous differences. So to be honest with you, that I don't think that time will come we can ever prove that our model is correct.
Yes, how can one prove something? That's one of the real problems.
Yes. I don't think we can ever do it. And even if you have a perfect model which mimic the climate system, you don't know it, and you have no way of proving it, and your model is always a simplified margin of reality, and since you cannot evaluate the consequence of simplification, onto your project, because you know, to evaluate that, you have to have a perfect model and you compare simplified model and then you say how your two model give you different forecasts, therefore what it is. Since it is impossible to make perfect model, as complicated as reality, you always have to do simplification. Since you are incapable of assessing the consequence of that simplification on your forecast —
Yes, I understand. Yes.
Right?
Yes.
So there's absolutely no time we'll have perfect forecast.
Nevertheless I've been impressed at how stable the numbers have been in order of magnitude over the last —
— yes, you know, it's three degrees, plus, minus, 1.5. I think that's probably all right. But you know —
— you can't prove it.
The trouble is, between 1.5 and 4.5 — 4.5, you're pretty sure, aren't you, if you add the 4.5 to summer temperature of India or Japan, oh my God, or even Arizona, you put the 4.5 degrees Centigrade to Arizona's temperature —! But if it's 1.5 you say, uh, we can do it. Yeah.
So if you get back to talking about these models, to the question of documentation, you know if you had been doing accelerator experiments or something, then one would want to take a piece of the apparatus and put it in the Smithsonian or something like that.
Yes.
Now, I wonder if there's any way to do that with a computer model? For example, take the model of Manabe and Wetherald in 1967, does that still exist? Is there some way that that could be preserved?
I don't think that anybody interested in the thing. Now, again —
— in the first place, is there a printout of the code in the closet somewhere?
I think it is, but it's all yellow. Wetherald is a monumental person to save everything. He may have it.
So he might have it.
He might have it, but you know —
— what was that written in, by the way? Was that in Fortran?
I think it's still Fortran. Because I remember, I was writing some program at that time, it was Fortran.
Let me just propose something to you. I'm not at all sure, but would it be possible, I'm not saying that there's any sense in doing this, but if you wanted to preserve it so that, you know, future people could look at it, the way to preserve it would be to adapt it so that it could run on a mini or something like that, so anybody who wanted to could have this model, old Manabe and Wetherald's model, as a specimen. Is that conceivable?
Yes, but, you know, the computer models different from the instrument you are talking about.
I know, it's a real problem.
Yes, and I don't — the people don't have that — I don't think, not many people have that sentimental attitude. Because they want to — also, it's a complicated mess, because you are struggling at that time, and so —
In those days the code was so, it was code in those days was always so complicated.
Yes. And when you are struggling, you try to, graduate student's program, they struggle, struggle, struggle, and then you improve the system as you go by. And in some of the older systems, still bungling there. You don't clean up.
Right, it may even be there, but with the GOTO running around there —
— yes, and all that mess.
But you frequently find it interesting, it could be just as interesting as having Galileo's telescope or the first cloud chamber. You know, at the Smithsonian they keep some of the early cloud chamber models.
Now, it is an interesting thing that this, what I'm talking about, this confusion of Callendar's method to evolving into mine, it was also very nicely written in the paper by Steven Schneider. It's called "The CO2 Confusion." And he discusses somewhat about this issue I begin to discuss, why some of our pioneering works are all flawed. You may be amused by looking at them. I think that kind of thing is much more interesting than these computer listings which put everybody to sleep.
I was just thinking if you could put it up on a personal computer. I mean, nowadays you could probably run that thing in the background on an AT 386 or whatever.
I know, because you know, I was always proud that I had opportunity to use these top of the line computers, at each time, but now look back, some of the present powerful, powerful computer is, you know, equivalent with all these gigantic computers I was so proud of using.
I was just thinking coming down here, it would be fun you know if one could just have that model, that old model, to run on a personal computer. It would be just like owning an old microscope or something like that.
Now, there would be some. When you go back, when you go out of this place, you may walk in front some part of computer or something which was used in the Institute, there, in front, you know, there is glass case stuff. One of the staff members made it. So there is something like that. Also there is some document which von Neumann wrote when they established our group. Things like that, something interesting.
Those have a certain importance.
What von Neumann thought when he thought that these radar models could be useful for climate models is an interesting thing. There's a letter there.
Yes, that's the kind of thing that would be interesting. Do you know? I guess I should talk to someone here about what kinds of old files they have — director's files and so forth going back I guess to von Neumann's day.
Yes. I think that the best thing though is Joe Smagorinsky. He's a man. And — but you know, make sure to call George Platzman also, because he is — he is, if you read his article which describes this, you can see that he could have become science historian, because he had a knack of it. You notice it. He documents carefully, and then, the very careful choice of word and you know very factual. He's a real perfectionist, so that he don't write something without thinking carefully, and George Platzman really have the knack of it.
OK, I hadn't run across that, I'll make sure to look it up.
In this sense this is interesting, because in the sense that, the weather model is a form of climate model but it's highly variant (?)
Yes. OK. Let me now just ask you, since it would be a pity not to, about this paper of Stouffer, Manabe and Bryan that you just published in NATURE 7 December (1989).
Yes.
It's very interesting, and there are various things I could ask about it, but I guess, this is essentially one more step in this program that you've been doing all along.
Yes.
Who's Stouffer?
Stouffer is a sort of — he is excellent when it comes to computers. He's a computer resource for us. But he understands the science very well.
How does it happen that his name is first?
Because he worked so hard. I mean, he learned this experiment and so forth, and so, we thought that he deserves —
I see. Your model and Bryan's model and he's the one who put this together.
Put it together, and so we worked very hard, how to set up this experiment, and in discussing with him, but he worked so hard, so we thought, he deserves the senior position occasionally.
I see, so you trade around.
Yes. And I think this case, you know, he did really work hardest among all for this paper, and name ought to be — the ideas maybe, yes, but we thought that he deserved sometimes to come first.
So the paper was written jointly, everybody had a hand in writing it?
I mostly write it. But you know, everybody have a different type of contribution.
I understand. Now, I noticed, you were talking earlier about spectral models. I noticed that this one is in fact done with the functions of the atmosphere, and finite (?) differences for the oceans.
Yes.
This I guess is partly a historical thing, that's the way the models?
No, I think there's a reason, because atmosphere is spherically continuous, so it's easy to note, you see, so it drives you nuts if you try to duplicate —
— you can do the oceans that way, but you can't do the atmosphere with finite differences. (Or, you can't do the oceans that way but you can do the atmosphere with finite differences.)
Yes. For example, GISS model, finite differences. And see, the usual way, the usual way so successful in finite differences, they conserve not only energy but prevent abnormal growth of the irregularity. They preserve square vorticity. All kinds of things. So that's how the difference model is very nice. Now, we didn't, we aren't that ingenious in building our finite differences, so when Stephen Orsag who's our mathematician came up with this, what we call transform method, which is a crossover between spectral and finite differences, we jumped over, because this is result of doing finite differencing. This is one of the cleanest ways to come up with the whole planet? Which probably better than —
Do you have difficulty matching the atmospheric model to the ocean model because they're so (???)
Yes, we made a — the — the problem, the most fascinating thing is of course, we had to make it communication, so that it had to be energy conserving, everything, water conserving —
— over the boundary —
— yes, everything consistent, so that airtight. So that any mistake there we always can find.
If the ocean loses heat, then the atmosphere has to gain it.
When you talk about warming over a hundred year period or something, you have to make —
— you can't lose a single —
— yes, and then you don't know whether it's a Greenhouse Warming or some heat is getting lost or —
— or it's somewhere out in the sixth decimal point.
Yes. Yes. And so, that consistency of the system is very important. But then there's another interesting problem. That is, maybe Kirk Bryan explained to you, but the — what it is that the atmosphere is faster system. That is, if you have one square centimeter air, it has 240 calories, therefore 2.4 is the heat capacity, therefore 2.4 calories per gram of air heat capacity. Ocean — so 240 calories to warm up 1 degree one square centimeter. Now, ocean column, two meters, 40 centimeters, (?) one calorie per gram of water. So if you have a full centimeter of ocean, if ocean mix that, heat down, like this, then no matter how much greenhouse gas you pump in, ocean —
— totally absorbs it all.
So what we are finding, some place, ocean totally absorb it all, and other place, it doesn't. So the ocean absorb very fast in second power? Ocean for example, or Wally Broecker's thing, this conveyor belt going around, and that's slowed down in the Atlantic.
I wanted to ask you specifically about that. To what extent was your work on this model driven by an interest in this thermohaline circulation?
I think we discovered that this coupled ocean-atmosphere model seemed to have two stable equilibria, similar to what Wally was talking about. And we have a paper on that. So it is our very strong interest to watch them and see whether such a thermohaline circulation may or may not happen.
Yes, now, does this affect your model work at all? That is, do you do anything to the model to make this particular?
— yes, we something sneaky here. This was one of the problems that — you see, you look at a Pacific and Atlantic, and surface salinity, which one is larger? Pacific is very fresh in high latitudes. Atlantic is much saltier and warmer. And that's why Western Europe is warmer than Alaska, for example. So it's very warm water comes in and sinks in the low regions (low ? sea?) or something. This is called thermohaline cycle, warm saline water comes and sinks. And Wally calls it conveyor belt. Now, this seems that, because of this thing, Pacific is more saline, no, Atlantic is more saline than Atlantic. (???) Because it's saline, it's heavier, so it sinks.
Right, right, I understand.
So let's keep on going, yes, so —
— no, I understand, (crosstalk)
— so this is very fundamental problem. Unfortunately, our coupled air-sea model is always biased towards killing this thermohaline cycle. So we always get larger fresh cold salty way of northern North Atlantic.
So you have a hard time keeping it going.
Yes, it's drop dead solution. We call it halocline catastrophe. (?) And so, why we had first paper in 1969 didn't use this couple model until the eighties is because when it looks at salinity Atlantic is fresher than Pacific, and then say, oh my God, this is — so we never use it.
So you had to keep putting them on.
We actually put some kind of flux adjustment —
What adjustment?
Flux, water flux adjustment, to give you realistic surface salinity. But what happen is, even with flux adjustment, same flux adjustment, depending upon initial conditions, one of them give you this over time another one drop dead solution. And so that was the paper we wrote recently. So even when we get this very interesting slowdown of thermohaline circulation, and warm up climate, we also back of our minds, there may be, you know, we don't know up to where our model is robustly working. Because we did some cheating here.
I see. So it provides a very sensitive test of your model, this particular effect.
Yes. But you see the, your own credibility of your own model are constrained because the cheating you are doing.
I understand.
And so that still coupled models are being in use by improvising, like your plumbing (?) system, your home.
(crosstalk) — parameters... you're never sure which parameter to take.
Yes. Here all we did was, we added some water flux systematically.
Instead of changing something else to get the circulation going.
Yes, but in this case — well, of course, this is rather straightforward because we are talking about salinity drift away, so we just change water flux at the surface. Yes, you could do other things. But we did it in such a straightforward way so it's very clear and clean, what we do, but still it's annoying. And so when we get this slowdown of thermohaline circulation, there is some doubt that, how genuine this result is.
Right. Tell me, are there any other things that also provide sensitive phases that force you to?
This is I think the main thing we are doing, which we have to —
— for example, the circumpolar, or the deep gyres off the Atlantic?
Yes, that is a thing that — what happened is that, one of the first major things happening in the Southern Hemisphere circumpolar ocean is very deep mixing of water which delays that warming over there. And the question is, is that really genuine? Now, we do know though, (?) ocean also mix, because any —
— how much the mixing is —
Yes, nutrients and everything mix. So probably it is right. But the Drake Passage is very narrow. The Drake Passage is the key for creating this deep mixing. If you close Drake Passage, then Southern Hemisphere warming is just like Northern Hemisphere.
I see, because it depends on the currents going around —
— yes, sir, that is the most important thing.
I didn't realize that.
Yes. And so —
— it really has to be completely three dimensional to get all that.
So this Drake Passage dissolution (resolution?) — there are more than two (?) to deserve it? Now, we do have the higher resolution model, and so we have a security blanket. But so, —
— you could close it artificially.
We did that. And it happened, there was some catastrophe maybe, something like that, and that may be the pre (?) of formation of Antarctic Ice Sheet.
When the Drake Passage opened.
Yes, once it's opened, then that created — because it cut the heat exchange between —
I understand, and then you get surface — uh huh.
Now. And so the more, the thing is temperature, but you know, it's a very interesting issue, what the role of Drake Passage in the formation of Antarctic Ice Sheet.
Right. But in terms of looking at the real world and saying, we have to adjust our model to the real world, that would be more the North Atlantic.
Yes.
The Antarctic is not such a constraint on your model. You get these gyres.
Yes. Oh, some herring circulations. I don't think it — the thing is, though, you have this almost bridge going through Iceland, and obviously that infrastructure, of bottom of topography (?) chemically, resolved by our cross- (?) model. Now, Kirk can explain to you more. So that's —
That makes it very sensitive, what might be changed, I understand, and we need —
So that when next the supercomputer comes, we'll go to higher resolution models and so, you know, the people think that if we took anything we see from the window and put in the model and then run it, without carefully validating these things, then model would be realistic the more complicated it is, but that is not true. So that I think we have to do a lot more homework in individual basic physics or dynamics of the things. We have to dig much deeper, just like putting a whole bunch of if statement algorithms in the computer, and then grind through, and hopefully because it's complicated, you see, we get more, get a better answer. And it cannot work that way. In a sense, we are talking about statistical mechanics of a natural system, working with statistical mechanics of a (?) go to 500 kilometers. If you have many cumulus moist convections, what is the statistical mechanics of that cumulus convection, on that scale? Or, what if you know that all these individual surface, when abrogate this to 200 by 200 kilometers, can you come up with the equivalent of something like statistical mechanics?
Yes, obviously you're very dependent on the data there.
Yes. The trouble is, we don't even know — see, when you develop, a physicist develops statistical mechanics, they have a classical physics to aim at.
Yes and you have a very simple model.
Yes. And so at least you know what to explain, what law of classical physics you explain by statistical mechanics. So you have some simple kind of molecules or something, and then you define the entropy or something, and then you have a partition function or something, and then you get into thermodynamics.
Right, and you don't do that with —
Yes. We don't even have a microscale (macroscale?) practical physics to guide you, to try to explain.
Yes, you have to observe it. You have to rely on —
— yes, have to observe is boggle your mind.
That's why satellites are very important.
Yes.
But that doesn't solve all of your problems.
You can't see everything from a satellite.
Yes, there's probably cloud cover.
Yes, or soil moisture. You try to make a soil moisture. Every point, you got a — so it's like opinion polls. How to do it is the question.
Right. Right. Now, one last thing I'm interested in is the interaction between all of this and the press and the public.
Yes.
For example, you're quoted in NEWSWEEK. When did you start finding that the press and the public was taking an interest in all of these things, in your work in particular? When did reporters start dealing with you?
I think, you know, the reporter always interested... but you know, the biggest thing change occurs is when Jim — Jim Hansen testified in 1988, when there was summer drought. I was also testifying in the same — you know, I remember, I was showing you the trans (?) of the Senators — and I was testifying on the summer drought, because I testified on summer drought, mid-continental summer drought, preceding November and then maybe one year after. So they said, oh my gosh, this fellow predicted everything. That's what they said. OK, invite him again, now it's so hot. So I was testifying, but I gave so many caveats on this thing. I told them that that drought had very little to do with — because I can't, I know how much, when I doubled the carbon dioxide, how much drought I got, and then if I pro-rate, warming is a bit small as compared to total (crosstalk ) — yes, total response is so slow when I pro-rate, the smaller warming, so I said, you know, so far, only half degree centigrade warming. It's too small. I can't explain this dramatic drought to this Greenhouse warming. Right. But Jim said, "We discovered Greenhouse warming because it's going up and this century, the highest in history," and I think he tried to sort of draw attention of Senators, so that they use the common analogy (???) which you may not use it in a science journal. I don't feel like blaming him that much, but some people criticize him. The thing is that — but one thing it accomplished is tremendous interest, and it's become, it went from purely scientists debating to really a political issue, and all this Green movement and so forth. It's become, make politician uneasy unless they are doing something on it. And so that is a thing. Then another thing is, shortly before that, there was the issue of ozone, ozone layer, and then these freon created catalytic reaction to this whole ozone, and you remember, there's a Montreal Protocol.
Sure, yes, that's —
— so these are the various factor which drove very strong interest.
How much do reporters in fact get in touch with you, how often?
I don't think I get as much as him. (or, as Zim?)
I'm sure you don't.
But, oh, last summer, for about a week or two or three week I was continuously getting continuous telephone calls, and after a while, get so tired answering same thing.
The message, when you deal with a reporter you try to get some message across. What message were you?
My, I try not to give any message. That is my —
What do you want them to know? What do you want them to understand?
Yes, that I try to do. I try to honestly say that we, that there are uncertainties in the warming, and this is the kind of range, and unfortunately we cannot narrow down, because of particularly cloud cover on the ocean. That's (crosstalk) — yes, and I didn't, even at that time I didn't say, say 2 degrees this (?) four degrees. And (2 degrees this might be or 4 degrees?) Because of cloud cover uncertainty, I can't say whether 2 or 4 degrees or doubling, which may be the latter half of the next century, may be realized. So I just told them more or less official line, not just my model but official lines of uncertainty at that time, and —
Did they accurately reproduce this?
Some TV crew or something come from Japan or something or some other in this country, they say, "Oh, you have to say it more strongly, otherwise nobody interested in what you say." There was something of that kind. But I do feel that — you know, for example, I just got a letter from Carl Sagan, and Carl want to have sort of a few scientists, several scientists, to support — he wrote a statement that our environment, global environment, is in danger and we have to protect it, and he said that he would like to present this to various leaders, and then have a united front and influence politicians, decision making. I am not very anxious to get into that myself, because I — particularly in working for government, and I don't want to get into political actions or anything. And I myself sort of continuously sort of doubting that it's — I am sort of, want to be open minded, and so I would like to reserve my right to change my mind, when results come out, like circumpolar ocean, you know. I'm the one who always said, high latitude warming rather than low latitude. All of a sudden I have to swallow my word here. And so, —
Will you save this letter from Sagan or will that get thrown out?
I have to respond to him, but I gave to — the director has it on his chair, I have to recoup.
Eventually all these letters get thrown out?
Yes, I always throw them out.
Sagan probably saved his side of it.
Oh, I'm sure. He worked very hard, so, to write it in such a way so that everybody can agree. But I felt, even probably for me it's much more effective — I really don't have that much talent to try to influence politicians. So I would be, it's much better using my talent, staying as anonymous as possible here, and try to publish a paper. I get much better credibility. I get much better influences. Because once you start getting in political arena, when you don't have that much talent in there, and then immediately when they ask you about your scientific results, you lose credibility, because they say, "Manabe's an environmentalist, he's not a scientist."
Right. I understand. Let me ask you about the other way, though, the fact that there is now so much interest and concern, does this affect what kinds of questions you want to ask your models?
Yes. What kind of question?
Well, obviously you're always asking your models questions. You're asking them about CO2 doubling. Does the public interest in this affect what you'll run next on your model?
No, the thing I am working very hard is, I have been worried about for several years, that I haven't wrote a paper on, is what happened to frequency and intensity of tropical storm generation. Something like that. Or I am interested in what happened to structure of first miles of atmosphere, when, and so that water pollution from the city, what it changes, in (?) and so forth, and some of these problems are never addressed anywhere. So I'm interested in these issues, so that influence of global warming on the regional pollution problem, or the problem like hurricanes, which we never addressed, and more, we go more to — we would like to study more of the (?) runoff. We'd like to get, we have a hydrologist here, and we would like get him to — these are some of the problems which we never wrote papers about. We may get into, along with some fundamental questions, like effect of ocean on the distribution of warmings, or things like that. Bryan is now studying how sea level, this warming affect the sea level rise. In this case of course west Antarctic ice sheet, what happen to that thing, which may be relevant, this paper may be relevant to that issue.
Still a live issue, you feel.
I think it's much more remote, until, you know, several hundred years from now, when finally deep oceans warm up, then no matter how much you mix, it will warm up anyway.
Sooner or later it will have to warm up.
Yes, so, you know.
Eventually the west Antarctic ice sheet will be a problem.
Yes, but push way into the future. If our model can be trusted.
So suppose that the public had no interest in any of this. Suppose that politicians had no interest in any of this. Would that affect what you would choose to work on?
No. Because you know, when I first worked on this, there was nobody interested in this, because in 1960, and it was really my curiosity. I realized that how much greenhouse gas is doing to the atmosphere, so I was just interested in the role of greenhouse gasses in maintaining our present climate, and then we did a little bit of perturbation study.
I see. So the fact that you're now getting more interest in regional studies just reflects the increasing sophistication of the models, the ability of the models to answer those questions?
I think, though, now I think the kind of question I just mentioned is my response to the various people's — and I think, I think that if — that one thing that's happening now is, I'm getting much more competition now. So that this, and so, and also I think people are asking these questions, and so that naturally I would like to answer these. So it's a case of research getting faster, and we can start doing more of the questions people ask, whereas originally I didn't care less.
Doesn't it also mean that your case can get faster because you probably have more funding now, more people to work with?
Oh, funding. You know, I think the effect of it is, maybe it's hard to prove, our funding may have increased because of this. Maybe make a (???) of computer. It may be due to greenhouse gas initiative. You can ask, for example Smagorinsky or Gerry Morgan, or, yeah, because I think it's…