The problem is that conventional projections for how warm things will get come out of a calculation everyone knows is wrong. Called the Charney sensitivity, it estimates how much the global mean temperature will rise if atmospheric CO2 is doubled from its preindustrial levels, before people began burning coal and oil on a grand scale. In the mid-1800s carbon dioxide concentrations stood at about 280 ppm. Double that to 560 ppm, and the Charney sensitivity calculation tells you that temperatures should rise about three degrees C.
But the Charney sensitivity, though not quite as stripped down as the billiard ball model, is still an oversimplification. The calculation does take into account some feedback mechanisms that can modify the effects of increasing temperatures on short timescales—changes in water vapor, clouds and sea ice, for example. But for the sake of simplicity, it assumes no change in other, longer-term factors, including changes in glaciation and vegetation; in particulates, such as dust; and in the ability of the ocean to absorb carbon dioxide, which diminishes as sea temperature rises.
Climate Models Struggle with Reality
“Many people, ourselves included, have tended to take [the Charney sensitivity] and apply it to the real world,” says Gavin Schmidt, who is also a climatologist at Goddard (though not a co-author on the new paper). “But the real world isn’t a model where a few things can change while others stay fixed.” At some point, Schmidt says, “we have to talk about the real climate.”
That’s what Hansen has attempted to do. He isn’t the first: other scientists, including Stephen H. Schneider of Stanford University, have talked for years about bringing additional real-world factors to standard climate models. The difficulty is that to add those factors, you have to come up with a reasonable way to weight them.
Like other climate scientists, Hansen and his co-authors use evidence from the deep past to sort out these feedback mechanisms. Over the past 800,000 years, for example, we know that the climate has oscillated between long ice ages and much shorter periods of interglacial warmth—much like the conditions we are in now. The relation between air temperature and CO2 is pretty well understood for that period, thanks largely to air bubbles trapped in ancient ice cores that have been drilled in Greenland and Antarctica (the CO2 concentration inside them can be measured directly; the global mean temperature can be calculated from the relative abundances of two different oxygen isotopes, which vary with how warm it gets).
But Hansen points out that the record contains other clues. “We also know how sea level changed over that period,” he says, from studies tracing the height of ancient shorelines. Because sea level rises and falls as continental ice sheets retreat and advance, you have a measure of what fraction of the earth was covered with a bright white, heat-reflecting coating. As ice retreats in a warming world, more dark surface is exposed to absorb solar radiation, which makes the world even warmer, melting even more of the ice. Conversely, a cooling world gets cold faster as ice sheets advance. This is one of the key feedback mechanisms left out of the Charney sensitivity calculation, partly because it is thought to happen only over hundreds of years, and, Hansen says, partly because “it just hadn’t sunk in that the paleoclimate record is a remarkable source of info on climate sensitivity.”
Using that record, for example, Hansen concludes that even if the human race could maintain today’s level of atmospheric CO2, which stands at 385 ppm—not even halfway to the atmospheric doubling we are headed for—sea level would rise several meters thanks to the disintegration of continental ice sheets. Moreover, he thinks disintegration may happen much faster than one might naively expect. “We didn’t have convincing data on this until we had the gravity satellites,” he says, referring to GRACE, a pair of orbiters that can detect tiny local changes in the earth’s gravitational field. “Greenland has gone from stable mass in 1990 to increasing ice loss. Another big surprise is West Antarctica, where despite little actual warming, the ice shelves are melting.” As those partially floating ice shelves melt, land-based glaciers are free to slide more rapidly to the sea. In Greenland, meltwater from the top of the glaciers is evidently pouring down through cracks to lubricate the underside of the ice sheets, easing their flow out to the ocean.