Clouds are one of the biggest wild cards of climate change: They’re crucial to forecasting the future but devilishly hard to model. Worse, computer models don’t seem to agree with what satellites see. Now, researchers are bridging this gap, finding that, since the 1980s, storm clouds have gotten taller and have shifted toward the poles. And, they say, just as models predicted, these trends may be making climate change worse.
The group’s findings are “good news for models, but not such good news for the planet,” says climatologist Veerabhadran Ramanathan of the Scripps Institution of Oceanography in San Diego, California, who was not involved in the study. Cloud systems associated with storm tracks currently exert a cooling effect at lower latitudes. The migration of those clouds to the poles is “problematic for our future,” he notes. “The good comparison between models and observed cloud trends has made mitigation even more urgent.”
Ramanathan says clouds are the “Gordian knot” of the climate problem. They both influence the climate and are influenced by it. They may act as reflectors, bouncing incoming sunlight back into space, or like blankets, absorbing heat emitted from the surface and then radiating it back down. It’s a complex interplay that makes it hard for scientists to tease out the effect of clouds on climate and vice versa.
Scientists have been working on untangling that knot for decades. To help refine their models, many researchers look to data from weather satellites collected since the 1980s. But the “satellites were not set up to monitor clouds,” says Joel Norris, an atmospheric scientist also at Scripps and lead author on the new study. For example, geostationary satellites look directly down at the planet, but clouds are easier to detect on a slanted path, so the instruments might underestimate cloudiness. And polar-orbiting satellites are designed to cross the same spot at the same time each day, which is important for tracking changes in cloud cover—but over time, the satellites may drift in their orbits as they run low on fuel, arriving a bit later each day.
To get around these issues, Norris and his team first performed a series of corrections to account for those systematic problems in cloud data records for the period from 1983 to 2009. Then, they hunted through the corrected data for clear, long-term patterns. Several patterns soon emerged: Since the 1980s, the world has gotten cloudier toward the poles and less cloudy in the midlatitudes; towering thunderclouds have also gotten a bit taller.
Those patterns matched three rather dire climate model predictions: that storm tracks—the paths along which cyclones travel in the Northern and Southern hemispheres—would shift poleward; that subtropical dry regions would expand, and that the tops of the highest clouds would get even higher. All of these changes can worsen global warming, Norris says. Storm clouds play a big role in keeping the planet cool by reflecting heat back into space—but they’re not as effective farther north or south, where there’s less solar radiation anyway. Dry, cloud-free areas at lower latitudes mean more absorbed radiation on Earth. And higher cloud tops create even more of a blanketing greenhouse effect.
So how did those patterns stack up against the climate models? The researchers compared simulations of climate change from 1983 to 2009 with the new patterns found in the satellite data. And this time, the team reports online today in Nature, the cloud observations and the models told the same story. The study is the first to use satellite data to confirm that models have been correctly predicting cloud patterns, Norris says.
Norris notes one caveat: There were two major volcanic eruptions in the time period they studied—Mexico’s El Chichon in 1982 and the Philippines’s Pinatubo in 1991. Both cooled then warmed the climate. In the model simulations, that warming produced broadly similar cloud patterns to warming caused by greenhouse gases. Resolving how much each contributed to ongoing cloud cover changes is an area for future work, he says.
The study is a “painstaking analysis” of the fragmented satellite record and shows some consistency between models and observations of clouds, says meteorologist Bjorn Stevens of the Max Planck Institute for Meteorology in Hamburg, Germany. But there are a lot of lingering questions, says Stevens, who also was a lead author on the “Clouds and Aerosols” chapter of the Intergovernmental Panel on Climate Change’s Fifth Assessment Report. For one thing—as Norris noted—it is difficult to distinguish between greenhouse gas–driven cloud changes and volcano-driven changes in the study’s time period. And, Stevens says, the study doesn’t discuss the types of clouds that are thought to be the most crucial for future warming: low-lying clouds over the subtropical oceans, which have a strong cooling effect but may be dissipating as the world warms.
Perhaps most chillingly, the study reveals how inadequate our present observing systems still are when it comes to certain fundamental climate questions—such as whether the world is getting more or less cloudy, Stevens adds. “This work reminds us that if we really want to understand our changing climate … we need to do a much, much better job of watching clouds.”