I was a friend of Alvin Weinberg who was the Director of Oak Ridge for many years. He's a great physicist who had... a very close friend of Wigner – they worked together on the early reactor work – and he stayed at Oak Ridge for 40 years. He's still there. He's now retired, he's 83 and he still plays a vicious game of tennis. He's an excellent leader and Oak Ridge has always been first rate scientifically. They've had their problems with nuclear energy like everybody else, but in the meantime they've done a lot of excellent science. One of the things that Weinberg did long before it was fashionable, was worry about carbon dioxide in the atmosphere and what it was doing to the climate. And so he invited me to go down to Oak Ridge and work on this 20 years ago, and I went there regularly for quite some years until it became fashionable and then I gave up because obviously I wasn't going to compete with the huge influx of people who came into it later. But in those early years Oak Ridge was really the only place that was worrying about carbon dioxide, and we had a very good group there: Ralph Rotty, who was collecting the information from all over the world about what was actually going on, and Greg Marland who is interested in vegetation in particular. And so my interest has always been in... in finding out what was actually happening in the real world, as opposed to doing computer models. So we fought very hard to get observations, and that's remained a central concern of mine ever since; that the huge industry which has grown up is doing computer models of the climate and trying to determine effects of carbon dioxide from computing what's going to happen, and this is a very dubious business if you don't have good inputs. The out... output of a climate model looks very impressive to the non-expert; the experts know that it's no better than the input. And in this case we simply don't yet know what's going to happen to the carbon in the atmosphere because we don't know what already has happened. We don't know what is happening, and the only way to find out is by observing.

So anyway we were at Oak Ridge, this little group put together a programme, and the Department... the Department of Energy who runs Oak Ridge didn't pay much attention to us, instead they put all their money into computer modelling. And that remains true even today. But, so what I was mostly doing at Oak Ridge was actually just looking at the balance between the vegetation and the atmosphere, which to me has always seemed to be the central problem; that there's more carbon in the vegetation on the earth than there is in the atmosphere, so that the atmosphere is the tail and the ground is the dog in this case. I mean, it's a... vegetation is really controlling what happens rather than the atmosphere. In fact, what one needs in order to understand the problem is to understand the vegetation first, whereas the emphasis in the climate models, of course, has always been on the atmosphere. But you can't understand the atmosphere by itself; the vegetation is absolutely essential. Well, what do we know about the vegetation? Not very much, and so the elementary questions are: How much carbon dioxide is going into the vegetation through... the photosynthesis, and how much is coming out through respiration, and what is the balance? Well the first good measurements of this were done only a couple of years ago by Wofsy at Harvard, who has a very wonderful technique called eddy flux measurement I think he calls it; it's a figure for eddy flux measurement and it's a very clever trick. He puts a tower up above the forest, if we're talking about trees; it doesn't have to be trees, it could be any sort of vegetation. He puts a tower, and at top of the tower you have instruments which measure accurately the speed of motion of the air, second by second or even 1/10th of a second at a time, and also measure carbon dioxide abundance, 10 times per second. This can now be done with modern instruments very precisely. And the remarkable thing is that there's a very high correlation, second by second, between the movement of the air up and down, and the carbon concentration. So the... you can actually see the flux going into the ground just by measuring what's in the air above the forest. It's a very direct measurement of the flux. And this has had amazingly little notice. I mean, I think it's a revolution in the whole state of the art. He's finally done what we proposed 20 years ago and it's... been done in Brazil in one place and it's been done in Canada in one place, and there's another group in, I think in Duke University, who are doing some measurements, but it's on a tiny scale. I mean it's tiny compared with what's going into the computer models. But this is what's really needed, so one will know whether the carbon dioxide is actually going into the ground or whether it's coming out. And the evidence is that in Massachusetts it's going in, much... much larger quantity than was estimated by the experts; in Brazil it's going in more or less to the extent that was estimated; in Canada it's coming out. Well the quantities, of course, are enormous and what should happen is, in the next few years, is we should have similar measurements in a hundred different places, then we should begin to know what is going on on the global scale. But until now we don't have it, and until you have that sort of information it makes very little sense just to believe the... the output of the climate models.