I have to backtrack for a second to answer that. After I completed my Ph.D. in 1987, I wound up doing a post-doc at the Max Planck Institute of Meteorology (MPI) in Hamburg. The Institute was led by Professor Klaus Hasselmann, and the focus of MPI was on developing a coupled atmosphere-ocean climate model. At that time, the U.S. had a number of different climate modeling centers—in Princeton, in Boulder, and in New York—the British had a climate modeling branch at the Meteorological Office in Bracknell, but the Germans were just in the process of trying to develop a coupled atmospheric-ocean model with which to perform greenhouse warming studies. When I came to Hamburg, I soon became involved in these efforts to use the climate model they were developing for global warming experiments. My interest was not so much in developing models, but in looking at results of experiments and comparing the model with nature, with the actual data, and trying to figure out whether the model projections of climate change matched the actual observations of, say, temperature change at the Earth's surface.

And so the motivation of that 1992 study, the first author of which was Uli Cubasch, was to take a slightly different look at this issue of global warming. The way in which modelers made the comparison between their model predictions and the observed climate variables had not been all that systematic up to that point in time. What we tried to do was put this comparison with observations on firmer statistical footing. This involved using signal-processing techniques. Essentially, we were trying to look at the model and figure out what the climate signal was in response to the imposed changes in atmospheric carbon dioxide. We wanted to find out where the signal’s largest, and in which climate variables it’s most obvious. Was it in temperature, pressure, rainfall, winds? We asked those kinds of questions and tried to address them in the paper. The other climate modeling groups had not really focused on the statistical aspect of this issue: how do you extract a climate change signal from the model data that (like in the real world itself) contains both signal and a background "noise"? That was one thing that really distinguished this paper from its predecessors: the focus on the statistical aspect of the problem.

  Was there another distinguishing feature, as well?

Yes. A second thing that was rather innovative at the time was the focus on sea level changes. Not many modeling studies had looked at sea level changes in response to global warming. These sea level changes result in part from the warming of the oceans and the thermal expansion of seawater. We tried to quantify that effect and understand how much sea level might change in response to changes in atmospheric carbon dioxide. That's an issue that remains of considerable practical concern to many people.

  What prompted you to look at the change in sea level?

I think much of the credit has to go to Professor Hasselmann. He wanted to do something different, something that distinguished the Max Planck Institute and the German climate modeling effort from the other international efforts going on. He didn’t want it to be just the same old thing with a different climate model. So it was Professor's Hasselmann's vision that made this paper rather special.

  Were you surprised by the influence this paper has had?

Yes, because it was not published in a high-impact journal like Nature or Science. And Climate Dynamics at the time was a journal that was more or less in its infancy. I'm gratified that it’s done so well.

What do you consider the biggest obstacles at the moment to global warming research and climate modeling?

There are many. I think it's tough to put your finger on any one and say this is the neuralgic point—like finding one pain point in the body and saying this is where it hurts the most. There are a number of obstacles that limit our ability to understand the nature and causes of climate change. One is clearly the climate models. We have an imperfect understanding of clouds and how they interact with incoming sunlight; how they modulate climate change. We have an imperfect understanding of important physical processes. Modeling rainfall, for instance, is very tricky, and clouds and rain are important for both the Earth's radiation budget and the hydrological cycle. What limits modeling is often observations. Or to put it this way, the limitations in climate modeling go hand-in-hand with limitations in the observations themselves: how well we can measure and understand very basic physical processes? Inadequacies in key observational records are another serious obstacle to continued progress in climate modeling.

  What would you like to achieve in your research in the next decade?

Many scientists around the world believe that there is indeed a discernible human effect on global climate, and that this conclusion is based on a solid scientific underpinning. The problem is quantifying the size of the human effect. Is it large or small in historical climate records? How large is it likely to be in the future? There are many efforts directed towards answering these questions, and improving our estimates of the size of the human contribution to climate change. I'd like to contribute to these efforts. I’m particularly interested in looking at a whole range of climate variables—not just temperature, which is the focus of most climate change detection studies. Many people are a little uneasy with studies that look at temperature only. They feel, quite rightly, that a human effect on global climate should be manifest in more variables than temperature alone. We should also see it in pressure and rainfall. The climate system should be telling us an internally consistent story. Some of my own future research will consider whether the changes in a number of different climate variables are indeed internally consistent. And I’d also like to study new and innovative climate change "fingerprints," like the rapid increase in the height of the tropopause (the boundary between the troposphere and the stratosphere). The warming of the oceans is another promising "fingerprint" of human effects on climate. Unraveling the causes of climate change is the ultimate detective story. I hope I’ll be involved in this story for a long time to come.

Dr. Benjamin D. Santer
Lawrence Livermore National Laboratory
Program for Climate Model Diagnosis and Intercomparison
Livermore, California, USA

In this ESI Special Topics feature, Professor W. Lawrence Gates, one of the executive editors of Climate Dynamics, provides a brief commentary about the status of this journal in global warming research.