Modeling Future Climate

Understanding our planet's climate, and how it is changing, is growing increasingly urgent. But the task is so enormously complex that today's best computer models are still insufficient, almost primitive. At NYU Abu Dhabi's Center for Prototype Climate Modeling (CPCM), mathematicians and scientists with several specialties are working together on the problem.

The Center's principal investigator is Andrew Majda, Samuel Morse Professor of Arts and Science at the Courant Institute of Mathematical Sciences at NYU New York; his co-principal investigators are Shafer Smith and Olivier Pauluis, both associate professors of Mathematics. They work with the Center's senior scientist Ajaya Ravindran and a growing number of postdoctoral associates.

Majda, Smith, and Pauluis are based at the Center for Atmosphere Ocean Science (CAOS), part of the Courant Institute, while Ravindran is based in Abu Dhabi.

The CAOS and other institutions around the world are developing new mathematical theories to improve climate modeling. CPCM's role is to convert those pieces of theory into practical mathematical tools, and to demonstrate that these work in the real world. "In New York [at CAOS] we're thinking about the tools, and here in Abu Dhabi we're putting them together," Pauluis explained.

Mathematicians and scientists have been building climate models for a long time, but in recent decades the work has been increasingly dedicated to accurate predictions of climate change, as a basis for decisions by policy-makers.

But the globe's climate is so spectacularly complicated that despite all the advances in computing, "we still don't know very well what is the best way to represent climate and analyze data. We have to find the way to best represent the physics," Pauluis said.

"The current approach tends to be too restrictive," he continued. "The tools we have don't reproduce the natural fluctuations we find in weather. So we're trying to put some randomness in the mathematical model. But the problem is to make the randomness accurate."

People tend to focus on warming, but one of the more drastic effects of that is what it will do to rainfall.

Olivier Pauluis, NYUNY associate professor of Mathematics

There are two sources of complexity in the global climate. One is related to the problem of turbulence in the dynamics of fluids, not only ocean water but air as well. "In a swimming pool, for example, you create simple turbulence where your own swimming motions generate a few small eddies. But on the scale of the weather, changes generate a lot of smaller motions" and these interact. "Trying to model this," he added, "you soon run out of computing power and computer time."

The people at the CPCM, he said, are "big users" of NYUAD's High-performance computing cluster. Yet, as admirable as that computer is, "it will probably take another century" before the world's scientists get the computing resources they need for this problem."

Meanwhile, additional complications arise because climate is more than just fluid dynamics. There are aspects of chemistry, biology, and more to consider, such as when plants convert carbon dioxide to oxygen. All in all, said Pauluis, "we're still in the process of identifying the equations we have to solve."

As an example of the Center's work, Pauluis cited a recent project on rain. "A huge amount of energy is dissipated in the atmosphere as rain falls, because friction with the air slows the raindrops." The CPCM has used satellite data to measure the amount of energy released this way. That's the kind of research that will make future climate models more robust and reliable than the current ones.

Pauluis, born in Canada but raised in Belgium and educated there before earning a Ph.D. at Princeton, focuses mainly on clouds. "This is very good for me, because on a nice day I can go outside and look at the clouds," he joked.

Clouds are in fact an important part of the climate change puzzle. "People tend to focus on warming, but one of the more drastic effects of that is what it will do to rainfall. The distortion of precipitation patterns is one of the most difficult parts to forecast, but it's also one of the most important," he said.

And it is a topic of special interest in the Arabian Peninsula. "We believe that this region is going to warm faster than other areas," Pauluis noted, "and that the subtropics are going to get drier."

The pace of research on climate modeling can seem painfully slow, considering the urgency of the climate change issue. But there are no magical answers, Pauluis said, and unfortunately "there is a lack of urgency at the political level.

"Maybe my generation won't be able to do much about that," he continued, "but by teaching students, I hope we can help the next generation find the courage to do what must be done."

This article originally appeared in NYUAD's 2013-14 Research Report (13MB PDF).