James Hansen - Session I

Weart:

If you would, please identify yourself for the microphone, sir.

Hansen:

Yes. I’m Jim Hansen, the Director of Goddard Institute for Space Studies.

Weart:

Right, and let me just say, I’m doing this mainly for my own research but I’ll give you a transcript afterwards and you can decide if you want to let me share it with other scholars, especially Paul Edwards. I think you’ve talked with him. He’s interested in this.

Hansen:

Oh, Paul Edwards. Where’s he?

Weart:

Oh where is he at the moment? Stanford? And he’s interested in the history of GCMs in particular. I’m more broadly interested in the history of global warming.

Hansen:

Yes, I hope I sent you the paper that was just published in the Arakawa [volume] edited by David Randall.1

Weart:

Yes, and that was very useful. In fact, that saved me asking some questions, and the biographical material you sent in general, means we can short-cut things a bit. In fact, I see from what you sent, just to very briefly cover your start, your father was a tenant farmer and your mother is a waitress. Is that right?

Hansen:

Well, not at the same time. My father, for the first few decades of his life, lived on a farm until the time that I was four years old. So this was…gosh, let’s see. I was born in ’41 and I have five sisters, four of whom are older. So they must have got married about 1930. So they lived at the time when farms were becoming very un-economic.

Weart:

Well were you brought up as a farm boy, or?

Hansen:

Only until the age of four, so I can barely remember the last year on the farm. My father never owned a farm. He went to school only through the eighth grade. But at that time, it was fairly easy to farm somebody else’s farm, because they were becoming un-economic. People either were dying or leaving the farm, and so it was possible to farm a farm and give the owner a portion of the crops–sharecropping, you know. But that was very hard to live on. The last year we lived right near Denison [Iowa], and on that farm just one year. It was right across the road from the country club. So my father went and served as a bartender part time. He actually made more money being a bartender than he did farming.

Weart:

And you still can nowadays.

Hansen:

So we moved into town, actually bought a little house and moved it onto a piece of land near Denison, and lived there for many years. But he was a bartender in town for five or eight years, something like that, and then became a janitor. That was his last profession. In the meantime, while he was a bartender, we had a pretty big family with seven kids, so it was a little hard to support that based on just his earnings, so my mother got a job as a waitress, and continued to work at that.

Weart:

At what time did you first become interested in science, do you think?

Hansen:

Well, math and science were the things that were easiest for me in school. I wasn’t academic, spend my time playing pool or playing baseball…

Weart:

As people who are in a small Iowa town do.

Hansen:

Yes. And I had a newspaper route, and in high school I became the distributor for the Omaha newspaper in Denison with about a dozen newspaper boys. I saved enough money to go to college that way. But you know, on national tests, I always got the highest scores in math and science in our school, but I didn’t get particularly good grades.

Weart:

So when you went to college, it wasn’t with the idea, “I will be a scientist.” It was just, “I will go to college.”

Hansen:

Yes, but I assumed that my major would be math or science of some sort just because that’s what I could do well. But I was very lucky because of the fact that the state university was the University of Iowa where they had a really good department of physics and astronomy. I was really too shy to try to approach [James] Van Allen.

Weart:

Was he already the Great Van Allen at that point?

Hansen:

Yes. And you couldn’t ask for a more gentle man than him. He’s a pipe-smoking guy who walks slowly and is very thoughtful and is very nice. Not aggressive at all as good scientists often are, and have to be, or whatever. But nevertheless, I was too unconfident of myself to really say, “Oh, I want to take a course under this great scientist.” But I was very lucky because I did give a talk in an undergraduate seminar course. You had to read a paper and then discuss that paper and it was about the atmosphere of the outer planet.

Weart:

This was already as an undergraduate.

Hansen:

Yes.

Weart:

Did you pick that or it just happened to you?

Hansen:

Planets seemed interesting to me.

Weart:

Were you a science fiction reader or any of that?

Hansen:

No. No, not at all. But as a junior I took an astronomy course. The astronomy professor, Satoshi Matsushima, was interested in getting students, and both Andy Lacis, who is now here at the Institute, and I were…

Weart:

Oh, he was a fellow undergraduate, then.

Hansen:

Yes, as was Larry Travis who’s also here. Then there was a seminar course which Satoshi had. Van Allen had this side interest in astronomy, and so he came to the seminar, he came to the talk. And then knowing my interest on planets, he suggested to me later, “Oh, there’s some new data on Venus.”

Weart:

Was this after you graduated?

Hansen:

Yes. That must have been the first year after I graduated. Actually, Andy and I were the first undergraduates to take the graduate exam, the Ph.D. qualifying exam. Usually at the end of your first graduate year, you would take this exam, a written exam, problem-solving. And for some reason, Satoshi, who wanted to keep us, we were undergraduates, and he wanted to keep us as graduate students–he suggested we take this exam even though we were undergraduates. Andy and I both passed it, and I did actually quite well on that exam.

Weart:

This is the point at which Van Allen becomes—

Hansen:

Oh yeah. So then he suggested this problem…

Weart:

Okay. Let’s back off for a minute, before we get onto Venus and so on. You mentioned some interesting things. You admire Van Allen’s non-aggressive ways, and describe yourself as a shy person, and yet you’re also known for being quite frank in your scientific statements. So the question is, what formed your personality? Where does this sort of mixture come from in your early experience?

Hansen:

That I’m not sure. But one other perhaps relevant comment about Van Allen. I didn’t take a class under him. He once gave a talk for the students and everybody. He came in with this pad of paper and just wanted to discuss the numbers, the measurements that they had made. They had previously made measurements discovering the radiation belts, and this was another instrument. It was still fairly early in that process of satellite measurements, but a later instrument where they were measuring the energy of different particles. These were new measurements and he was just thinking about possible interpretations of them. It was just an ideal example of the way a scientist thinks and tries to formulate a solution of the data. So that kind of thing may have a role in shaping at least your approach to scientific problems.

Weart:

You’ve been willing to speak more frankly than many scientists have. You’ve been willing to sort of put yourself on the line a little more than many of them have been.

Hansen:

Yes. But again, this does go back to the scientific method. I have this favorite quotation of Feynman’s about (I don’t remember the exact words) in science you learn a certain integrity. You have to report both sides of the story. That’s exactly what Van Allen did – he didn’t have any particular thing that he was trying to prove. He may have had some preconceived ideas, but nevertheless, he reports the data in a very objective way. That’s the nature of science, to be objective.

Weart:

Set it out as frankly and as clearly as you can.

Hansen:

Yes. It is a very natural scientific thing. Now, it is true, and it’s something that I puzzle over, the fact that there is this very strong pressure. Maybe not very strong; maybe it’s subtle in some ways. You know, in my old age, I will rethink this situation.

Weart:

We’ll get back to it in 20 years. Let’s go on to Venus and I’m not going to spend too much time on it, but I would like to cover it. There had been these measurements that they found that Venus was very hot. I guess you were aware at this point of [Carl] Sagan and his carbon dioxide paper?

Hansen:

Yes, this is again a good point, because again, it does relate back to Iowa. The procedure at that time was that a student, after first taking this quantitative exam, the second thing that you had to do to qualify for doing a thesis was, at the end of two years in graduate school, you would take another exam. Part of that exam would be, in addition to having a written exam, to make an original proposition and defend that before a committee.

Weart:

Is it like a Master’s thesis kind of a thing?

Hansen:

No. This is before you’re allowed to do your Ph.D. thesis.

Weart:

So it’s part of the qualifying exam.

Hansen:

So it’s part of the qualifying exam.

Weart:

That’s interesting.

Hansen:

And if you didn’t do well on the combination of the written and this oral exam, then they might say, “Well, you just get a Master’s degree,” after a year or two. But the idea was to just show that you understood the way that science works. So you could make an outlandish proposition, and it may be just as well to make an outlandish one. But then you have to try to defend this and they’re going to judge you on the basis of your understanding of science and physics. So there were these various theories for why Venus was hot, including the greenhouse effect, and I needed an original proposition. I had previously, on my Master’s paper, written about the Earth’s atmosphere to try to understand the effect of the volcano, Mount Agung, on the brightness of the lunar eclipse. I tried to understand how the aerosols in the Earth’s atmosphere affected the light getting into the shadow of the Earth.

Weart:

I see. Had that been measured?

Hansen:

Yes. When I was a senior, and taking Satoshi’s course, then he had the idea to measure the eclipse of the moon. He knew an eclipse of the moon was coming up and he wanted just to make photometric measurements of that.

Weart:

And that would have been about the time of Mount Agung.

Hansen:

Mount Agung had gone off six months earlier. When we went out there, it must have been one of the coldest days on record there. It was like 25 degrees below zero on this hill outside of Iowa City. Andy and I and Satoshi and a different graduate student, an older graduate student who knew more about the telescope up on the hill, went out there and measured. And that eclipse was invisible. You couldn’t see the moon at all.

Weart:

It was gone because there was so much stuff in the air.

Hansen:

There was so much dust in the air. Actually, the moon went through the southern part of the Earth’s shadow, and the southern hemisphere had all this dust in the atmosphere. Most of the aerosols went into the southern hemisphere. But it made the atmosphere so opaque that almost no light was getting into the shadow.

Weart:

It’s interesting how I can see some of the pieces of your career coming in that early on.

Hansen:

Yes. So anyway, I wrote a paper on that topic. I probably still have it. Well, by December 1963, I was in my first couple months of graduate school when the eclipse occurred. Then I did calculations of the brightness of the light getting into the Earth’s shadow. Basically, I read this paper in German by Link on eclipse phenomena. I had to translate from German. Then I just followed his method, except I put it on a computer for doing the calculations. But anyway, I learned enough about scattering and radiation and like that.

Weart:

I see. It was published in JGR. I see.2

Hansen:

Right. So therefore, I was trying to suggest the idea of how aerosols might affect the temperature [of Venus]. But to make it physically different, I was saying that the energy, instead of coming from the sun, I was going to trap the energy coming from the internal heat source of the planet. And actually, then it’s a little hard to think of how you’re going to keep these aerosols suspended if they’re dust. So, it was a little difficult to believe that this was very likely. Nevertheless, it was good for the discussion for this oral exam, and the professors actually liked it so they said, “Well, why don’t you work on that idea.”

Weart:

And so it became your thesis.

Hansen:

Right.

Weart:

Now, I haven’t seen your thesis, but I assume it’s based on some kind of one-dimensional radiative model?

Hansen:

Yes. We were in Japan then. Our professor, Satoshi Matsushima, was Japanese. He received a grant to go to Japan for a year, and he took us with him. And it so happened his professor at the University of Kyoto, was Sueo Ueno (who’s still alive, I visited in Japan about a year ago). The Institute for Astrophysics was in an old building there on the Kyoto campus, with a coal burning stove. Anyway, he was doing these calculations for multiple scattering with this invariant embedding method with Bellman of UCLA. They would apply a technique where you add a thin layer to the atmosphere and calculate how that would act, add a thin layer, and it gives you…

Weart:

Right. And add them up together for a thick layer.

Hansen:

Yes. I sort of wanted to try to use it. He was such a nice, polite gentleman, and he really wanted somebody to use it. Bellman wrote 500 papers and many books, and likewise, Professor Ueno who was working with him, they were doing very many papers. But people were not generally using their method for very many practical computations. So I really thought I’d like to try to use that in my thesis. I did include a chapter on it, and I spent a lot of time in programming the invariant and embedding method. I had a chapter in the thesis on the infrared radiative transfer on Venus. If you carried it through to the logical conclusion and looked at the infrared spectra of Venus, then it could tell you something about what the clouds were made of. This would be relevant to the issue of to what degree the clouds or aerosols, or whatever.

Weart:

You came out a little later with a paper saying that saying they were one-micron aerosols, more or less.

Hansen:

Right. That was a different thing. And that would have been nice, if at that time in my thesis we had been able to identify what this haze was around Venus. Then it would have been relevant to our interpretation. So that was a chapter in the thesis, but it really wasn’t central to the question of what the temperature is, and the radiative transfer problem of calculating surface temperature. So for the surface temperature, I read these papers of Suki Manabe’s. Manabe and Muller, in particular, and Manabe and Strickler. I remember that at the University of Kyoto, Andy and I would go in after hours. If had a paper we wanted to Xerox, at that time, it was a photographic process where the paper went through this liquid, and it came out and you’ve got this purple paper…

Weart:

They had some strange ways of doing it.

Hansen:

And anyway, we got a copy of this Manabe and Muller paper. So my idea was that I would use Manabe’s one-dimensional radiative-convective type of calculation.

Weart:

I have a general question about that, I haven’t been quite able to figure out. Manabe and people before him and a lot of these planetary people were of developing their own methods, and I never see them referring to Chandrasekar’s work.

Hansen:

Yes. See, Chandrasekar’s work is very relevant to what Sueo Ueno was doing.

Weart:

These people knew about Chandra’s work? They read it?

Hansen:

Well, see, Chandrasekar’s work in the same category of being very mathematical and theoretical, but what you want is a calculation that will give you a number that you can use and try to understand what’s happening on the planet. So Chandrasekar’s work, he has this beautiful mathematics and does an exact solution to some hypothetical problem. It was almost always for isotropic scattering and then the great crowning triumph was to do Rayleigh scattering. That’s a more complicated phase function than isotropic, it’s only got two terms in a Fourier expansion. But in addition, you could even include polarization in the Rayleigh scattering. But that’s only a fraction of the scattering in the real atmosphere. Also, it’s relevant mainly to solar radiation. To calculate the temperature of the atmosphere and at the surface, you do have to take account of the energy from the sun, but you also have to take count of the thermal radiation, the real atmosphere with aerosols and clouds and things, it becomes very complicated, and these highly mathematical treatments were nowhere near reaching that degree of complexity. What I did later was follow Van de Hulst’s suggestion– “Let’s approach this problem numerically,”– he suggested this doubling method. Well, he also did some things analytically or semi-analytically, but he was willing to use a computer. And of course, that was just the thing that was beginning to be available then. When I came here in 1967, just the next year we got the [IBM] 360-95, which was the fastest computer in the world.

Weart:

And so the idea was that instead of this beautiful formalism, you had to base what equations you would use on what your computer would do. You started with the idea of the computer and went from there to what equations you would use.

Hansen:

Yes. And Ueno’s and Bellma’s approach was really, like Van de Hulst’s, well suited for computers. But still they were just doing the light scattering problem. And that’s great for remote sensing problems, and I later used that in trying to understand what the clouds were made of. But back to the question of trying to interpret these observations of Venus, why it seemed that the surface was so hot, then you have to figure out not only where the energy is getting deposited, but how it can get back out. While I was in Japan, I saw this advertisement for post-doc positions at NASA, at the Goddard Institute in particular. I wrote to Van Allen, saying I thought I was making some progress on trying to do some calculations related to Venus, and that I hoped that in a year I could finish and I would like to apply there. Satoshi went through the roof because his idea was that graduate students should stay several years and write papers with their professor, and so he tried to stop me from getting my Ph.D. But fortunately, although Satoshi was my advisor, Van Allen was then the Chairman of my thesis committee.

Weart:

And he was, of course, sympathetic to NASA and all that.

Hansen:

Yes, and also— Well the point was that when we got back from Japan, Satoshi didn’t encourage me to write my thesis, because he didn’t want me to finish. But, when I belatedly realized there was a deadline, I rushed to give Van Allen half of my thesis, and he said I had another week to get the other half done. I worked very hard for a while. Van Allen thought it was a good thesis and I got out. Okay, back to the method. What do we do to calculate the temperature? Well, I couldn’t use the mathematical approach of invariant embedding. As I say, I made a chapter on it just so I would have something to show. But then I decided, well, I could use Manabe’s approach. But it was too complicated. I didn’t have time. I wouldn’t have been able to finish that year. All I really needed to know was how much optical depth do you need in the infrared in order to block the heat radiation, to trap heat radiation and to keep the surface temperature warm enough. So that would yield an optical depth for aerosols, and then the question is, is that amount of aerosols plausible?

Weart:

That is still difficult to answer today.

Hansen:

Yes, right. And actually, we wrote a paper, which was published in Astrophysical Journal.3

Weart:

Now you also did a paper that used, I don’t think it’s the same paper, where you used polarization measurements?

Hansen:

No, that’s later as a post-doc. When I came here to New York, they had this big computer so I could start programming methods of doing radiative transfer and multiple scattering with a realistic phase matrix for cloud particles. Actually, I used a realistic phase function and neglected polarization when I was first a post-doc here. But then I got another post-doc, an NSF post-doc, to go to the Netherlands. The idea was, when I went there, I was going to work on either one of two problems. One was to do a phase matrix to introduce polarization and do this doubling method with polarization, or else do what I started to do with Sueo Ueno’s invariant embedding, but in the infrared. Because if you looked at the thermal part of the spectrum, and if you did high spectral resolution, then you could look at absorption lines. There was a dispute between Pollack and Sagan with Don Rei of JPL about what are the absorption bands on Venus that we see in the infrared. Sagan and Pollack said they were ice bands – which you would get absorption at about 1.5, and two microns, and near three microns due to ice. But also, CO2 gas is absorbing there. So I wanted to do this invariant embedding with a high spectral resolution and try to interpret that infrared spectrum. Or, alternatively, do polarization and to interpret the visual polarization observation of Bernard Lyot.

Weart:

And still with the thought that it might be dust?

Hansen:

Well, that was the idea, to determine what information could we get on the microphysics of the scatterers. Because when the sunlight is scattered, in general, the first scattering introduces very strong polarization.

Weart:

Yes, you can learn about the size if nothing else.

Hansen:

The size and the refractive index and the shape of the particles. It’s called microphysics. And the thing was that when people first observed the polarization of planets as the phase angle of the planet changes in going around the sun, then it looked like the light of the scattering was basically un-polarized. However, Lyot developed this very precise polarimeter which revealed polarizations that were ten times larger than his error.

Weart:

Okay, well then to finish with the Venus thing, so you came with this paper finding there are small round things.4 Do you remember at what point you heard that it was sulfuric acid and how that affected you?

Hansen:

We had these observations of polarization as a function of phase angle, and they showed us two features in particular at near 20 and 160 degree phase angles, which were related to the light. And they changed with wavelength in such a way that we could determine that it was scattering by spherical particles. In fact, we could get the refractive index index very precisely. Because one of these features is the rainbow, caused by light rays that go inside the particle and are reflected once, so it depends on the index of refraction. But the angle shifted in with wavelength going from UV to near IR. We could say that the refractive index changed from about 1.46 in the UV to about 1.43 at one micron. So I looked in the Handbook of Physics…

Weart:

To see what could it be?

Hansen:

Yes. The refractive index was a little too large to be hydrochloric acid, but it was consistent with sulfuric acid. So, I talked with John Lewis of MIT, the expert on planetary atmospheres compositions, and I foolishly let him convince me that “it’s very unlikely, implausible, that it could be sulfuric acid.” So I decided, “Well let’s publish this thing without interpretation of what the chemical composition of the clouds is, just publish exactly what we had learned about the partial microphysics.” Of course I later regretted that. It was a good lesson. When you have found some really hard data, that you are very confident of, you should follow through on it. Still, I’m not sure it was the wrong approach. I really like to be able to write a paper in where the conclusions are quantitative and very solid. So maybe it was all right to let someone else identify the composition. Andy Young was the one who did that. He relied mainly on our refractive index data, but then checked that the Venus near infrared spent was consistent with sulfuric acid. How he was let to that was a funny story. Jim Pollack is perhaps the one who should have discovered the identity of the Venus cloud because he had all the data, refractive index and the IR data. He was a friend, so I have him and his colleague Carl Sagan the first draft of my paper with the microphysical data. Andy Young sort of beat John Pollack to the punch after he found out that Pollack was having somebody in Kansas measure the optical constants of sulfuric acid. Andy Young said, “Why is Jim Pollack having this guy do these sulfuric acid measurements? He must think the clouds of Venus are made of sulfuric acid. What a stupid idea!” He was aware of Lewis’ chemical equilibration arguments, but then he started to thing about it and decided to check the infrared spectrum of Venus to see if it fit sulfuric acid, which it did. The funny thing was that Him Pollack was having the sulfuric acid measurements made for an entirely different reason. He was interested in volcanic aerosols on Earth. He never thought to connect the two planets.

Weart:

It’s always been a question to me, how much first of all volcanoes made people think of sulfuric acid on Venus and how much was the other way around? Do you have any idea? It just sort of went back and forth?

Hansen:

Well, volcanoes on Earth should have made people think of sulfuric acid on Venus, but the connection was just the odd chance I just mentioned, until it was realized that the Venus cloud was sulfuric acid. There were aircraft and balloon measurements several years earlier that found sulfuric aerosols in th Earth’s stratosphere. Well, the sulfuric acid on Earth was measured by the sampler, these aircraft or balloon measurements…

Weart:

But people didn’t seem to recognize the importance of it at that point yet.

Hansen:

Well, they realized there were sulfates, and perhaps weren’t sure to what degree they were ammonium sulfate versus sulfuric acid. It was realized in the early 1960s that the stratospheric aerosols on Earth came especially from volcanoes and that they did contain a lot of sulfate.

Weart:

At what point did people realize that played a very important role in the radiation field?

Hansen:

That had already been known was know for a long time. Look, for example, at this book, Physics of the Air by Humphreys, which is from the first half of the century.

Weart:

I see, even Humphreys’ Physics of the Air. [1940]

Hansen:

Even way back when Benjamin Franklin was talking about aerosols and a dry fog, he was thinking about the effect on the radiation field but I don’t think he knew what the aerosol composition was.

Weart:

At that point, I don’t think he knew it was sulfates. Anyway, I can look that up. We haven’t time for that.

Hansen:

Yes, this is older than that. So regardless of composition, they knew that these aerosols were reflecting sunlight, and therefore tending to have a cooling effect. But that provided the optical data which Andy Young could then check in the infrared to see if that was also consistent with this inference about composition from observations. He is the one who identified the composition.

Weart:

Okay, now we’re almost up to the greenhouse work, but not quite, because I still wanted to ask you, just because it would be wrong not to ask you a little bit about your early years here and your relations with [Robert] Jastrow and you became a PI of the Pioneer Venus orbiter cloud polarimeter. What was going on here?

Hansen:

Well, I was a post-doc here from ’69 to ’71. In my second year, I had applied for an NSF post-doc to go to Holland, but before I left, Jastrow called me in and asked if I was interested, after that post-doc, to come back and take a position here.

Weart:

On the NASA payroll, not Columbia?

Hansen:

He was suggesting I would be hired as a NASA employee. Now, when I came back, I came in to see him right when I got back, and at the time, he was getting a little tired of the job or something. So I started asking him questions and he said, “Oh, go see Ichtiaque Rasool. He’s going to be deputy director,” or something. So he was a little annoyed about my asking. Then he said I should work with Al Arking. But as usual, I just went off and did my own calculations. Actually I came back after just six months in Holland, even though I had an NSF fellowship for a year. I got this computer program working to do multiple scattering polarization while in Holland, and I was so excited about that because then I knew I could interpret these Venus observations. However, I had to have a computer, and the computer here was more than an order of magnitude faster. So I came back early. That was when I had to see Jastrow, because I had to get paid somehow. So he said, “Go see Al Arking,” or something like that. I didn’t get paid for a couple of months, but eventually I started to get paid by Columbia, a Columbia grant. Then I started getting these nice results on the Venus polarization curves, and Arking knew about that, so he showed them to Jastrow.

Weart:

From there to the Pioneer, what’s the path?

Hansen:

Well, we had people here who were interested in planetary studies, Ichtiaque Rasool in particular, and young scientists, Joe Hogan and Dick Stewart. I got to go to Kitt Peak conferences, so I was aware of what they talked about there, people like Sagan and Pollack. Don Hunten and Richard Goody wrote an article in Science in which they recommended that NASA should try to do inexpensive planetary missions. So I was aware of those planned missions, and made, together with the other people here, Andy Lacis and Larry Travis and David Coffeen, a proposal for an instrument on that was accepted. I guess partly because we had shown that polarization was a useful tool for studying the clouds. That occupied me for a while.

Weart:

Let me cut you off on that though because I do want to get to the global warming I guess you’ve explained before that what really got you started was when Yuk Yung was a post-doc here from Harvard. Is that the story?

Hansen:

It does relate back a little bit to Jastrow. In the early 1970s, my interest was in planets, and the editor of Science invited me to write a review paper on the clouds of Venus. I think Joe Chamberlain, the director of Kitt Peak National Observatory, and Don Hunten suggested it to the editor. And I struggled over that. I wrote the title, “The Veil of Venus,” and an abstract, but I couldn’t finish the paper because I had to write for a general audience. It’s really hard, I struggled with that. Simultaneously, I was working on a review paper on light scattering and this Pioneer Venus proposal, and I never did get that paper done for Science. So I missed that opportunity. But at the same time, Jastrow had realized that there wasn’t much money left in planetary studies. He was trying to get the Institute more directed towards practical applications.

Weart:

Terrestrial applications.

Hansen:

Yes. Terrestrial applications, and particularly using satellite measurements to improve weather forecasts. Milt Halem sort of ran the project for Jastrow. Jule Churney and Peter Stone provided the high level scientific advice, and Richard Somerville was perhaps the main person putting the model together. They needed someone to do the radiation for the model.

Weart:

Just sort of a side job here? Radiation transfer?

Hansen:

Yes, even though it really could be a full time job.

Weart:

So was this Manabe’s radiation transfer that you took again?

Hansen:

No. I was able to hire Andy Lacis to work with me, and we did it pretty much from scratch. But all we did was the solar part. Joe Hogan did the infrared part. We were supposed to provide calculations for the deposition of sunlight in the atmosphere and on the surface. We did that.

Weart:

Lots of people had done that, but you just started from scratch again.

Hansen:

Well, lots of people do radiation transfer, but having the calculation be fast enough to be used in a 6CM isn’t so common. Andy Lacis and I wrote a paper that is actually my second most referenced paper. It was first for a long time.

Weart:

Which is the first?

Hansen:

The first is Hansen and Travis, which we wrote in 1974.5 It’s just a review of multiple scattering. But this one with Andy was called the Parameterization for the Absorption of Solar Radiation in the Earth’s Atmosphere.

Weart:

This is Lacis and Hansen, Journal of Atmospheric Sciences, 1974. I see.

Hansen:

Yes. We really were just fitting formulas to some multiple scattering calculations. [Look at paper.] Oh, okay. So now I remember. The calculations were the doubling method. Back to the doubling method. But you can’t do a doubling method in a weather prediction model because you would be taking more computing time than all the other parts of the model put together. So we just did calculations, and then we parameterized results in terms of the amount of ozone in the atmosphere and the amount of water vapor, and we integrated over the infrared spectrum with five points. We used a K distribution for the absorption coefficient, a probability distribution function, and integrated over the near infrared spectrum, the solar spectrum, and just found some analytic formulas which would fit the results.

Weart:

Is this what eventually went into the GISS?

Hansen:

Yes. So this then formed the solar radiation package in the GISS weather model.

Weart:

Have other models taken it up from you since?

Hansen:

Yes. Lots of models used it. Although as we mentioned in the review paper in David Randall’s book, we never used it in our climate model. We used it only in the weather model. But for the climate model, when we decided to start with a weather model but to make a climate model out of it, then we…

Weart:

Are we talking about a GCM or are we talking about a one-dimensional kind of model?

Hansen:

A GCM. The one dimensional model is just a single grid point in the 3-D model. But when we decided to work on a climate model in the middle of the 1970s, about the time of the Yuk Yung thing, we realized that we wanted to do the radiation more accurately than it was in the weather model so we never used these analytic formulas again. Instead we did the multiple scattering explicitly, but did it with what we call a Two Gauss Point Method. It’s like a two stream method, but we try to weight the different angles optimally.

Weart:

Back to the classic two stream.

Hansen:

Yes, something like the two stream.

Weart:

But you still built it up on your own, not derived from somebody else’s model.

Hansen:

Yes. Right. And eventually, at about that time, we decided that we had to do the infrared also.

Weart:

I see. So the weather model is actually quite different from the climate model.

Hansen:

Well, it depends on what you mean by quite different. The dynamical core was the same. It’s using Arakawa’s finite difference methods for the fundamental equations. But one term in the energy equation is the radiation. That, for climate, becomes a very important part of the problem.

Weart:

Because that’s where all the questions are on this.

Hansen:

Yes.

Weart:

So you had to do them more thoroughly.

Hansen:

Yes. So for the weather, you’re interested in the pattern, where the highs and lows are, and you’re trying to simulate those well. The momentum equation is key to getting the winds and weather patterns right. But for climate, you really are just averaging over those weather patterns anyway.

Weart:

Right, and you need to be much more careful because it’s the radiation that determines what happens in the long run.

Hansen:

Right.

Weart:

Let’s get back to Yuk Yung. I think we’re about to that point now.

Hansen:

Yes.

Weart:

You were doing this weather model…

Hansen:

Yes. We were doing a weather model, and at that time, the chlorofluorocarbon issue had come up. But the reason it came up had nothing to do with the greenhouse effect of chlorofluorocarbons which Ramanathan pointed out. It has to do with ozone. So there was this workshop in Washington, in which Don Hunten was the chairman. He had as much influence on me as anybody. His point was that if the United States and other governments are going to be able to deal effectively with issues like the effect of chlorofluorocarbons on ozone, and therefore on human health, they need objective scientific analyses of the problem. This was the second time something like this had happened within the period of several years. Prior to this, it was the SSTs, the issues of the [supersonic transport] aircraft effect on ozone. He said, You have to have a basic understanding of the science as a background and then you can respond to these problems.” So they organized, and they really did a very good job, the upper atmospheric research community. At that time, I went to this meeting. It was basically a chemistry problem, however, there was a climate connection. There’s this issue of whether the changes in ozone will change the climate. And so they thought, “We should probably have a climate component of this research program.” At the same time, Ramanathan realized that the chlorofluorocarbons had a direct influence, not just through ozone, but they directly affected the radiation field. Yung, was a post-doc at Harvard of Mike McElroy’s. Then, because I had decided we would try to propose to have a model, a 3-D GCM that we’d apply to this ozone problem, we needed a chemist. So Don Hunten recommended, “Well, the three best young guys are so and so, so and so, and so and so,” of Yuk Yung was one. So I tried to get him to come to GISS and be our chemist in putting together this global climate chemistry model. We failed to do that. We set up a dinner. I was supposed to go to the Moon Palace with Jastrow and Yuk Yung. We went there and Jastrow never showed up, so Yuk Yung was put out. But nevertheless, he is a very enthusiastic guy and he said, “We really have to look at these other greenhouse gases and see how much they’re contributing.” So we worked on this paper together.

Weart:

Wang was the first author so what was…?

Hansen:

Yes, he was the first author. He was a little more aggressive Andy. Andy did the solar part of the calculation, and Wei-Chung Wang did the infrared part, and I wrote the paper. But it was all of us talking.6

Weart:

This was a radiative convective model, so where’s the convective part come in. Again, are you using somebody else’s…

Hansen:

That’s trivial. You just put in…

Weart:

... a lapse rate...

Hansen:

Yes. So it’s a fudge. That’s why you have to have a 3-D model to do it properly. In the 1-D model, it’s just a fudge, and you can choose different lapse rates and you get somewhat different answers. So you try to pick something that has some physical justification. But the best justification is probably trying to put in the fundamental equations into a 3-D model.

Weart:

Right. Let me go back again to these one dimensional models you were working on at the time. Were you communicating with Manabe or [Warren] Washington, or any of these other people on these types of things?

Hansen:

Only by way of reading this purple paper that was photocopied in Kyoto. No. I didn’t come into contact…

Weart:

Because Manabe’s right down in Princeton.

Hansen:

Right. But things are clearer in the paper, probably then going to talk to him anyway. In any case, the first contact with Manabe was when Charney had this study group on CO2 and climate.

Weart:

Yes. Before we do that, there’s one thing that comes in between and that’s the Pinatubo thing.

Hansen:

No. Pinatubo is later. There is something that comes in between and that’s Agung.

Weart:

Oh, Agung. Okay, I’m sorry. Agung.

Hansen:

Yes, because we still remembered this nice test case.

Weart:

A natural experiment.

Hansen:

Yes, it’s a natural experiment. You see, this was a long time after the experiment. The experiment was in ’63.

Weart:

Computers have got to be better then.

Hansen:

Yes, computers were better and we had this one dimensional model then, and it was a nice, logical radiative forcing to look at.

Weart:

I’m just curious, because to me it seems like a very pure, clean paper,7 and I wonder what kind of reactions you got to it.

Hansen:

In that case, I got good reactions. In fact, the NASA administrator read the paper, because it was in Science and I guess he looks at that journal.

Weart:

At least read Science and Nature, yeah.

Hansen:

And yes, what was his name? He sent a note down the chain of command and implied that, “Yes, NASA should be ready to look at the next volcano.” Because that was one of our recommendations. And he sent the message to the chief scientist, who at that time was Ichtiaque Rasool. So they started a program. Which we didn’t get anything out of, but Jim Pollack did, because you had to have somebody who had access to aircraft or whatever, and he was at Ames Research Center. They were in fact ready when El Chichon went off. They kind of got egg on their face because as soon as El Chichon went off, and they were predicting cooling, actually the Earth warmed up after El Chichon because we got the El Niño of the century. I didn’t make…

Weart:

You didn’t make that prediction.

Hansen:

No. We weren’t’ ready to make any. We didn’t have a model ready. We were working like mad trying to make this three dimensional model and it just wasn’t ready yet. We didn’t publish it until ’83.

Weart:

Okay, why don’t we get back to that, then. You were talking about Charney. You mentioned earlier that his study was one of the things that got you started on it.

Hansen:

Yes. He was a great scientist. It seems like in that era we had so many really impressive scientists. He is again a perfect example of how you should think about problems. And he liked to think about things that were different from his main study. He did think about some things which were radiation, and proposed that the drought in the Sahel was due to goats eating the grass and therefore increasing albedo from 14 percent to 35 percent. That was the first experiment that we put into our model.

Weart:

I think it’s still regarded as part of the answer, right?

Hansen:

It may be part of the answer, however, I think the goats had more effect on getting aerosols in the air. When you denude you get all that dust coming up and that affects the stability of the atmosphere and things. But anyway…

Weart:

So anyway, we skipped a little bit, which is you decided to do a GCM, which is a big decision.

Hansen:

Right.

Weart:

Whose decision was that? Is Jastrow still here?

Hansen:

In that case, it was more an influence Don Hunten. Again, I had the pressure of being a research manager as well as being a researcher, in that I had to support a few people. I realized I couldn’t get any money out of planetary studies. But also, it was a very interesting problem. So planetary people like Hunten, Pollack, and other people were at least dabbling in Earth studies. Pollack never really left planetary. He remained primary planetary. But other people got really interested in the ozone problem. So I wanted to try to make a model which could be used for that purpose. But then as a first step, before you get chemistry into the model, you first have to get the model to work for climate. The idea was to get a very efficient version of a weather model, which then could be used for long time scales. About that time, Charney was asked to write a report on this CO2 climate thing and to provide some advice to the government as to how important this was.8 He had a summer study at Woods Hole, and had really excellent people: Manabe, and Arakawa, and Carl Wunsch, I believe. So Charney heard, probably from Peter Stone, that we were doing a CO2 experiment.

Weart:

You already had your GCM up.

Hansen:

Yes. GCM was working. That’s what we call Model Zero…

Weart:

This is the one that took Arakawa’s finite differences?

Hansen:

Yes. The model was not published until 1983 but already at that time, ’79, we had Model Zero running.

Weart:

Is this the one with the thousand kilometer resolution and so on?

Hansen:

Yes. And we did a doubled CO2 experiment analogous to Manabe’s and got a rather different answer. Our model warmed up four degrees, while his most recent one warmed up two degrees. He had one earlier than that which warmed up 2.9 degrees or about three degrees.

Weart:

Because of differences in the sea ice and the…

Hansen:

And the clouds.

Weart:

And the clouds. Let me ask you a minute about that. You say you developed pretty much all of the physics from the start. Where did you get your clouds from? That’s always one of the interesting questions.

Hansen:

The clouds initially were mainly based on Arakawa, except that his model was two layers. We had to generalize it to nine layers, but still it was basically his cloud physics. It was then we tried to get a graduate student at UCLA, not an Arakawa student but a Schubert student, and that was Tony Del Genio, to work on the clouds. Eventually, we got him to come here, and the clouds that we now have, those have been his speciality.

Weart:

So, he learned it from Arakawa?

Hansen:

No. Tony worked on planetary atmospheres under Jerry Schubert. He was working on the clouds even as a student. He has done his own cloud physics, not using Arakawa.

Weart:

How do you differentiate your philosophy from Arakawa’s approach to modeling?

Hansen:

In the book. I was just looking at the introduction. Manabe made a comment that, “Arakawa is working on the perfect model. It’s not quite ready yet.” Well, it will always be that way.

Weart:

He said that he was the design and you were the production, or something like that.

Hansen:

Yes. He will always be in the design, I think. You know, if you want to do real applications, then you really have to just be willing to go ahead and do something.

Weart:

You’re more willing to put in fudge factors, so to speak?

Hansen:

Well, no. See, there are still very fundamental problems with the basic fundamental equations in turning them into numerical solutions. That’s part of our difficulty. And Gary Russell emphasizes that the momentum equation is still not done very well. Arakawa develops this numerical technique that conserves certain quantities very well, and better than in the previous competing methods. That was very crucial for being able to use this coarse resolution. It’s a little bit like Professor Sueo Ueno and his invariant embedding method. His love of the subject is in the mathematics and he takes it so far. Either he’s just not experienced in the applications or whatever. At any rate, he’ll leave it to a student or somebody else to try and make use of it. Similarly, Arakawa develops a model, and there’s great challenges in that. He had this nice collaboration with Yale Mintz, who went to use the model for weather prediction. We’re taking the model and using it for climate applications. It’s hard to have enough time to work on the basic structure of the model and also use it.

Weart:

Another question– Seasonality, that was just starting to come in to the models at this time, I guess.

Hansen:

That comes through naturally, in that the radiation just drives the seasons. There’s no problem of putting in the geometry right and looking at that. One thing that was different about our paper, which we published in ’83, was we focused equally on the different seasons, in particular on January and July, while models had always been looking at January and to some degree tuned to give good results then.

Weart:

Another feature is your ocean slab. You seemed to have a particular interest in the ocean uptake and release of heat.

Hansen:

That problem is kind of ignored in climate models, so the only reason to put in the slab was just for the sake of having heat capacity so that you would get a reasonable amplitude for the seasonal cycle.

Weart:

But heat comes back and bites you when you start to look at the global warming itself.

Hansen:

It doesn’t bite you, but you’re missing an opportunity to learn something if you don’t look at the heat storage on longer time scales.

Weart:

Right. So I’m wondering where this interest came from. It’s one of the interesting features of this paper.

Hansen:

Oh, that came because in 1983, there was this big study which was kind of the answer to –the interest in CO2 which was 1981 helped to heighten a little bit. Boy, this is a perfect example of where science goes awry from political pressures, and where it becomes less objective than it should be. The Charney report was done under the Carter administration, and DOE who got interested in it at that time and David Slade, the DOE program manager was starting to look at social implications of climate change. And when there was the election in 1980 and the administration changed, then [Fred] Koomanoff was given the job. David Slade had promised we would get support for our modeling effort here, but that promise disappeared with the change of people.

Weart:

We’ll get to ask more about that later.

Hansen:

You know, and we had written a paper in 1981 about CO2 impacts on climate10, which got a lot of attention because…

Weart:

Yes, we’ll have to get back to that.

Hansen:

That was because Walter Sullivan reported on the front page of the New York Times and the Department of Energy didn’t like that at all because they were just trying to downplay that topic, and here it got up-played. So as a result, we had no chance of getting any money from them. But not only that, then they had a conference near Washington. They were going to have a DOE CO2 climate program, and it seemed like basically the point of the conference was to criticize our paper. But anyway, they went ahead and prepared this big fat report by 1983, and it concluded that climate sensitivity was relatively low. The reason was that they had a fundamental error in their thinking. That was because of this ocean heat uptake, because what they assumed was that the time constant for the ocean to warm up is 15 years, and they took that as independent of climate sensitivity. But if you look at the ocean heat uptake and the response time, you see that in fact it is strongly dependent on climate sensitivity.11

Weart:

Right. You explained that very well. You have to take into account the feedbacks involved, that the thing is actually heating up and that delays everything.

Hansen:

Right, and that they didn’t realize.

Weart:

Now, I noticed in the Charney Panel report itself, there were a couple sentences that mentioned the ocean heat uptake. Was that just sort of intuitively? They just sort of guessed it?

Hansen:

No. One of the nice things about this Academy report was that you got to talk with Charney. Charney called us two or three times and sometimes we had a speaker phone…

Weart:

You met with the panel?

Hansen:

A speaker phone on this and so, we met with the chairman, at least. Maybe once with the panel, but I talked two, three, or four times with Charney. His thinking with regard to this ocean heat uptake, he was thinking of the dynamics of the ocean overturning, and how deeply the ocean circulation goes. His argument was that it’s the upper layers of the ocean, and he estimated that the overturning time was a few decades.

Weart:

For the mixed layer.

Hansen:

Not for the mixed layer. The mixed layer is actually mixed very rapidly and its depth varies seasonally, so it basically has a one year overturn to get to the full depth of the mixed layers.

Weart:

You said a couple of decades for the whole ocean?

Hansen:

No, not for the whole ocean. But he argued that it’s down to the thermocline that’s important. So, the upper several hundred meters. So he’s saying, the time constant for that is several decades, a few decades. Now at that time, he didn’t realize and I didn’t realize, that the response time is a function of climate sensitivity. We were just thinking of overturning time in the ocean. It was only a couple of years later that that became clear.

Weart:

I see. So it may have been in the Zero Model but you didn’t understand it until you sort of worked with it.

Hansen:

No. See, at the time of the Charney report, there was only the slab model. So then it was just a qualitative discussion of what the time constant should be. Then over the next few we started to do this diffusive mixing into the deeper ocean, and sort of approximating the effect of the full ocean heat capacity. Then as soon as we started looking at numbers, I realized this was more interesting than that.

Weart:

I want to ask you a couple more things about this ’83, ’84 work. You mentioned that it took you a long time to convince the referees that the low resolution worked. In fact, when I first looked at it, I said, “How do you get a jet stream out of such a huge grid size.”

Hansen:

We still have that problem [with referees].

Weart:

You just look at it and it is intuitively hard to understand how you can get anything out of that?

Hansen:

Well, no. I went in 1976 to NASA headquarters and made a presentation to the people who were in charge of the chemistry climate problem. We wanted to try and get support for the 3-D model. I took a graph of the wind structure, zonal mean which was from a paper by Elmar Reiter of Colorado, I believe, Colorado State. It was showing the exchange between the stratosphere and troposphere. The argument was that you had to be able to resolve the large scale eddies, and therefore you needed a resolution only of about a thousand kilometers. So that was the argument. It seemed to us that you didn’t need to do mesoscales to get the general circulation. If you’re not interested in weather itself, but just the effect on the circulation, then you should be able to use a relatively coarse resolution. However, in order to get the model to work, you needed good numerical methods, and that was what Arakawa had [unlike other models]. Some of the other models that had concluded that you needed finer resolution were making that assumption. It was a model-dependent result. [A few seconds missing on tape.] If you had artificial viscosity in your model for the sake of numerical stability, then that could spoil your results at a coarse resolution, because the viscosity was a function of resolution and became very strong with the coarse resolution. And we still have the problem. We’ve been trying to get support for our present model which is…

Weart:

Is this coarse resolution?

Hansen:

Well, it’s coarse resolution compared with other people. I just try to help convince NASA that they should continue to support it. We had an advisory committee meeting a few months ago, and they’re really the best people in the country, one of them from England. I’m still disappointed that even they, who I think are the best people in the field, are still very critical of our coarse resolution. I think I at least have to pay lip service to that, but the truth is that even at this time, the biggest problems in the model are associated with the physics, not with the resolution. We really have to be able to get the stratosphere in properly. We do need higher vertical resolution. We’ve got to get the proper wave interaction with the stratosphere.

Weart:

The absorption of the aerosols and basic things like that. Let me ask one other thing about this ’84 paper. You also tried to fit it to the CLIMAP results. Eighteen thousand years ago.

Hansen:

Oh yeah.

Weart:

And you got a bad fit. And Rind wrote a paper12 saying, “It must be the CLIMAP data,” which was correct.

Hansen:

I recall it very well, however…

Weart:

I’m not describing it right?

Hansen:

Yes. I don’t think that’s quite right in that at the time we did these calculations for 18,000 years ago, there were at least three major uncertainties in the CLIMAP data. One of them was how big the ice sheets were. There was the large ice sheet hypothesis and the much smaller one. There was also a question about the southern hemisphere sea ice. Different interpretations of the sediments they find [reflecting conditions] on the ocean surface led to disagreements about how much sea ice there was. And then there was this question about how cold it really got in the tropics. Therefore, for all these reasons, we had then very large error bars on the interpretation of that. And there was also, to some degree, uncertainty about what the forcings were in trying to compare the Ice Age with the Interglacial Period. The one thing you mentioned, is, “How much colder was it?” We assumed it was somewhere between three and a half and five and a half degrees colder. If CLIMAP is right, then it would only be about three and a half. To be more precise, when we put in CLIMAP SST’s, then we used the GCM to calculate that the world would be 3.7 degrees cooler during the ice age. In that case, that would tend to favor a climate sensitivity of about two and a half degrees for doubled CO2, we argued. But if the climate were really five degrees colder, which we thought was more likely and which David Rind and Dorothy Peteet and Wally Broecker argued that favored a higher sensitivity. They argued that it was not really as warm in the tropics as CLIMAP had claimed. I think that story has been going along, and still there are always arguments, but the weight has shifted in their favor. It’s now clear that in a lot of places in the tropics at least, it was pretty cool during the ice age. So the real temperature change was probably at least five degrees, and that tends to favor the higher climate sensitivity.

Weart:

You had been arguing that there was a higher climate sensitivity, so in a way, that made you more suspicious of the CLIMAP results? Since other reasons you thought there was a high sensitivity.

Hansen:

Yes. It would make it more consistent. However, you know, frankly I was suspicious both ways. I kept thinking, “What is it in our model that makes it so damn sensitive?” Because it just seemed like it was too sensitive in the sense that it made it very hard to believe that the Earth never froze over, if the climate was as sensitive as our model said. So we were always suspicious of our cloud scheme, that it was pushing us towards too high sensitivities. And the final thing in this story is still just playing out, and that is this snowball Earth.

Weart:

It does freeze over sometimes.

Hansen:

It does freeze over. And we just stupidly thought, “Well, if it freezes over, then you’re done for. It’s going to stay frozen over.” But then we didn’t think that if it does freeze over, the CO2…

Weart:

The CO2 [from volcanoes will rewarm the Earth]… I know. Everybody missed that. I felt dumb.

Hansen:

[Laughs]

Weart:

By the way, you mentioned Wally Broecker. He’s right near here at Lamont. Have you had much dealings with Wally or with the Lamont people. Is that…?

Hansen:

Yes. It’s a fruitful interaction. In the case of this CLIMAP thing, well, he’s been particularly concerned about paleo-climate.

Weart:

The things I’ve seen in your papers where you refer to his work are on the possibility that the oceans will not respond nicely. Now I’m asking about, just in general, does he come up here, do you go up there, is there much interaction between the people?

Hansen:

Yes. I think we have a lot to learn by looking at these longer time scales. It’s a kind of thing which NASA does not support in general. But I think that you need to apply the models to extreme cases in order to get a fair measure of how well they’re representative of the real world. Therefore, we hired Dorothy Peteet as a NASA civil servant, even though what she does is go out in the bogs and take these cores. They let her spend half the week at Lamont and half the week here, so that provides support.

Weart:

I see. That’s close interaction. Actually, I should back off because we’re up in the ‘80s and in 1981, you became the head here, so maybe I should ask you about that. How you became the head here and how your life changed.

Hansen:

You know, it’s a fairly small institute, and I was already the head of the planetary group and the modeling groups, so it was mostly…

Weart:

You were already heading a lot.

Hansen:

I was already heading a lot of it.

Weart:

Jastrow was backing out. You said he was losing interest?

Hansen:

Well, yeah. He was teaching at Dartmouth, so he was away most of the time, really. And the IGs didn’t like that.

Weart:

Who?

Hansen:

Inspector General.

Weart:

Oh, okay.

Hansen:

Anyway, he decided to go full time to the other stuff. He was really more interested in writing than anything, so he just sort of left things to Pat Thaddeus, for the astronomy, and to me for the planetary and the climate modeling. So it was logical that Goddard would either pick Thaddeus or me.

Weart:

Now, one other thing that started just about the same time… well, several things started at this time. One was that you started looking at the recent temperature trends, you started going back looking at the 20th century record13. How did that come about?

Hansen:

You know, it’s just, again, part of the way you do science. You have to spend about equal emphasis on data as on models, I think.

Weart:

Well, not everybody does. In fact, you hadn’t up to this point been doing any data work, right?

Hansen:

That’s true. I mean, we did observations of the Moon and things, in a sense. And we were trying to observe Venus also, but it was mainly other people’s observations. That’s true here, too. But the thing is, that if you’re trying to model the Earth’s temperature, you need some data to compare it with. Murray Mitchell had done these estimates of northern hemisphere temperature change based on meteorological station data. If you really want to compare, you have to have the data. Say you want to look at Agung effects or something, you need to have the detailed data and you want to have the geographical distribution of things. So I had started to work, first with an NYU student to look at this meteorological station data, and with the idea that you could probably do a little better. You’d like to be able to keep up with real time. Murray Mitchell and his data would be a few years behind, and by doing the analysis ourselves, we could keep up to near real time. Also, Murray Mitchell had only looked at the northern hemisphere. But there were stations in the southern hemisphere. Our idea was that a station should be representative of some region, and the question is, how big a region? Well, you can just look at the data and do the temporal correlations. See how well the neighboring stations are correlating with each other.

Weart:

That was the interesting thing to me. Instead of taking the data and applying it to the model, you’re taking a coarse resolution grid and applying that to the data. That’s a very interesting philosophical thing in the sense that one is trying to get the data to conform to your calculational scheme.

Hansen:

Well, no. Say for Mars or for Venus, if you’re going to send these spacecraft there, how many do you have to send, or how dense do your observations have to be?

Weart:

Right, that’s something you already had to think about for the Pioneer.

Hansen:

Yes. So the Russians sent in a Venera spacecraft and made a measurement and it maybe went in a very special place, and not very representative. But you draw a lot of conclusions about these planets based on very sparse measurements.

Weart:

Right. It’s the joke about it falling and landing in somebody’s pot, right?

Hansen:

Right. So the station distribution in the southern hemisphere may be very crummy, but on the other hand, we should be able to say something about how things are changing there. The question is, “How much can you say?” So if you try to optimize [to study] change, how should you weight these different stations? So that was the problem.

Weart:

How wide an area can represent...

Hansen:

And it did make a difference in this case because this conclusion that the Earth was getting cooler was a northern hemisphere perspective, and the southern hemisphere didn’t really agree with that. We were able to get a somewhat different conclusion from looking at station data ourselves and doing this analysis.

Weart:

In this paper, was the first place I noticed that you said we need to do more monitoring. We need to monitor aerosols and so on. There is the first place where you said that. Was it this temperature work, or is there anything particular that made you feel the need for better climate monitoring?

Hansen:

So they’re different things. In regard to interpreting the Agung effect on climate, we needed to measure the aerosol characteristics if you’re going to have the right forcing that you put in the model, but also you’ve got to know what happened in the real world. So you need observations of the climate change. So you have to measure both the forcings and the response, and that is a logical conclusion. And in working for an agency that does measurements, it’s very logical that I should be thinking about what needs to be measured.

Weart:

Okay. We’ve done enough for today. We’ll set up another one to cover some of these later things because you are a busy person and I’ve kept you for quite a while already.