Cloud confusion vexes global warming predictions

By Will Parker on November 24, 2010 in News

Estimates of global warming vary widely in large part due to the difficulty of modeling clouds and their effects in computer simulations. Now, as climatologists race to compile better cloud atlases, new research shows that previous worst case predictions of global temperature rise may be dramatically off-target.

Clouds are the least-understood variable in the sky, according to Texas A&M University geoscientist Shaima Nasiri, who notes that scientists do not even have a common naming system for mid-level clouds. “We do not have a unified definition, so the scientific community can’t look at the statistics with a shared level of understanding,” she explained.

Nasiri gives new NASA satellite technology the credit for helping accurately define and measure middle layers of cloud. “NASA satellites launched over the last few years have helped us identify height and base, and temperature and pressure of mid-level clouds. This has revolutionized atmospheric studies,” she says. But the amount of data received from the satellites is so enormous that a great deal of her work to date has simply been number crunching and developing algorithms just so she and other scientists could process the information.

Meanwhile, researchers from the University of Hawaii Manoa have been assessing the performance of current models in simulating clouds. Some existing models predict that global mean cloud cover will increase in a warmer climate and the increased reflection of solar radiation will limit the predicted global warming. Other models predict reduced cloudiness and magnified warming. Writing in the Journal of Climate, the researchers present a new – and apparently much more accurate – approach to determining the expected cloud feedbacks in a warmer climate.

“All the global climate models we analyzed have serious deficiencies in simulating the properties of clouds in present-day climate. It is unfortunate that the global models’ greatest weakness may be in the one aspect that is most critical for predicting the magnitude of global warming,” said lead author Axel Lauer.

Lauer and his co-researchers applied a model representing only a limited region of the atmosphere over the eastern Pacific Ocean and adjacent land areas. The clouds in this region are known to greatly influence present climate yet current global models do poorly in representing them. The new regional model, however, successfully simulates key features of the region’s present-day cloud fields – including the observed response of clouds to El Nino.

Applying their model a century in the future, the researchers found a tendency for clouds to thin and cloud cover to reduce – dramatically more so than in any of the current global models. “If our model results prove to be representative of the real global climate, then climate is actually more sensitive to perturbations by greenhouse gases than current global models predict, and even the highest warming predictions would underestimate the real change we could see,” warned co-researcher Kevin Hamilton.