Atmospheric Science at NASA: A History

• Epilogue
• Chapter
• Johns Hopkins University Press
• pp. 321-324
• View Citation
• Additional Information

Epilogue

In January 2006, on a sunny but chilly day in Manhattan, I did a couple of hours’ worth of oral history with James E. Hansen, the director of NASA’s Goddard Institute for Space Studies (GISS). At one point he asked me if I knew about a recent publication from one of the science teams associated with the Gravity Recovery and Climate Experiment (GRACE), the first of the Earth Systems Science Pathfinder (ESSP) missions. The paper concerned apparent, and unexpected, mass loss from the Greenland ice sheet, as inferred from slight changes in the local gravity field.¹ I hadn’t known about the paper and pointed out that ice sheets weren’t really in the scope of my mandate, a history of atmospheric science. He looked at me with surprise, and said, “Well, that depends on how you define atmospheric science.”²

Indeed. Despite the thickness of the tome you’re holding, I’ve drawn my disciplinary lines rather narrowly, focusing on the gaseous part of atmospheric science, if you will. I’ve generally ignored the atmosphere’s interaction with the Earth’s land, water, and cryospheric surfaces, all of which happen to be claimed by other scientific disciplines. I may have made a mistake in being so narrow. But if I have, I’m not alone.

Hansen had a larger point that took me a few months to see. In March 2002, Earth scientists of many disciplines were shocked by the sudden collapse of the Larsen B ice shelf, a chunk of floating Antarctic ice the size of Rhode Island. Imagery from the Moderate Resolution Imaging Spectrometer (MODIS) instrument showed what had happened. Meltwater has a much lower albedo than does ice; meltwater ponds on the surface of Larsen B had acted as thermal drills, rapidly absorbing incoming sunlight, warming, and boring their way through to the base of the ice shelf. An ice shelf that climate models suggested would survive for decades thus disintegrated in weeks. Albedo has long been considered the domain of atmospheric scientists (recall that Vern Suomi was interested in the albedos of both cornfields and of Earth) but this particular effect of albedo was unexpected by that community.

This error in understanding ice and albedo matters, because the world’s ice sheets hold enormous amounts of water. Glaciologist John H. Mercer had figured back in 1978 that the West Antarctic ice sheet held 5 to 6 meters’ worth of sea level; the Greenland ice sheet holds slightly more.³ If either of these ice sheets disintegrate, most of the world’s coastal cities will be inundated, and hundreds of millions of people will be forced to migrate inland. The cost will be in the trillions of dollars. Climate models predict that those ice sheets will remain intact for centuries, but the Larsen B event showed that the models’ treatment of ice sheet behavior is conceptually flawed. In a 2005 editorial spanning ten pages, Hansen took his own modeling community to task for this.⁴ Climate models build ice sheets by depositing and compressing snow in a dry process. Modelers simply run this process in reverse to predict their breakup. What actually happens is what happened to Larsen B: meltwater drills through the ice sheet, destroying it quickly. But neither Hansen nor anyone else in his community has figured out how to model this process adequately.

This ice problem was not seen as a problem by the leaders of the climate science community until the Larsen B event. One will peruse the Intergovernmental Panel on Climate Change’s (IPCC) 1995 and 2001 reports in vain for discussions of ice-enhanced rapid sea level rise. The IPCC, composed primarily of atmospheric scientists and climate modelers, laid out priority research areas that didn’t touch on this ice problem. Since NASA based its downsized EOS on the IPCC’s priorities, it didn’t favor this research area either. Yet the issue of meltwater destruction of ice has been known to glaciologists for decades at the least; when Hansen started his campaign to get his modeler colleagues to pay more attention to ice, he used a photograph of a crevasse in the ice with a meltwater stream pouring into it to make his point clearer.⁵ It was provided to him by Roger Braithwaite, University of Manchester, who took it during a field expedition. There have been field expeditions to glaciers and ice sheets since the nineteenth century; crevasses, and meltwater, have been known to science for quite a long time.⁶ To put not too fine a point on it, Mercer’s 1978 article was written specifically in the context of rapid Antarctic ice sheet melting induced by human carbon dioxide emissions.

There has clearly been a disconnect between the atmosphere and climate modeling community and glacier specialists over the past few decades. That’s not particularly surprising. The Earth sciences have grown so enormously in the postwar era that the Journal of Geophysical Research publishes more than forty thousand pages per year, and there are many other Earth science journals. No one could possibly read it all, and scientists don’t try. They read what’s directly relevant to their own specialty, and, in reality, only a select few papers in their specialty. So while Francis Bretherton’s 1986 committee formulated the concept of Earth System Science to try to break disciplinary boundaries down, the professional demands of science have the opposite tendency. Keeping a broad view is extraordinarily difficult.

The Larsen B event seems to have caused a number of scientists to broaden their view just a bit; the 2007 IPCC assessment released in February contained the following statement: “understanding of these effects is too limited to assess their likelihood or provide a best estimate or an upper bound for sea level rise.”⁷ They specifically excluded sea level contributions from future “rapid dynamical changes in ice flow” in their sea level forecasts.⁸ Because they cannot quantify the possibility of rapid deglaciation, they effectively assigned these possibilities zero value. In other words, after three decades of climate research by the world’s most advanced nations, we cannot yet put a ceiling on the potential impact of rising seas. That’s quite an assertion, given the potentially enormous cost of coastal inundation. Hansen seems to think that several meters of sea level rise may be possible this century.⁹

Much of the science discussed in the tome you’re holding was directed in nature. In other words, it resulted from either carefully formulated, long-term research programs that were advocated by proponents for years, or, like the Upper Atmosphere Research Program (UARP), were effectively commanded into being by law. Such programs have achieved important results (else this book would be much thinner!). But as the above story suggests, there is still a strong element of random walk to modern science. Even the best scientists are not all seeing; IPCC didn’t foresee the importance of ice dynamics in its first nineteen years of existence, so I was in excellent company.

From a science policy standpoint, then, it’s clear that maintaining flexibility in long-term research programs is necessary. Despite vast knowledge and best intentions, American scientific leaders didn’t quite see in 1988, or 1995, or 2001 what all the key scientific and policy challenges presented by global warming would turn out to be. Having resources to explore surprises seems essential. It isn’t at all clear that NASA’s budget outlook, at least, contains flexibility.

There’s a lesson for historians here, too. In her discussion of Big Science, Catherine Westfall warned historians against allowing scientists to define our rhetoric and our research programs.¹⁰ In a very real sense, I did exactly what she warned against in writing this book. I built it around a rather traditional, narrowly drawn definition of atmospheric science. In so doing, I’ve missed a great story about the belated discovery of ice-atmosphere interactions by atmospheric scientists. I hope some other historian will be able to do the topic justice in the future.