Simulations also mask exact year-to-year changes in moisture, fail to register El Niño events, and fail to capture precipitation trends on decade scales.
According to Wentz, it's not surprising that the models lag in the precipitation department.
"A lot of attention has been focused on their ability to predict global warming," he said.
Most computer models say that as the globe heats up, "the wetter areas on the planet are going to get wetter and the drier areas are going to get drier, which is a gloomy prediction," Wentz said.
The satellite data showed slight support for that prediction, but not enough to convince the authors that it will continue to hold true.
If the models are wrong about regional effects, that could be good news for places like the U.S. Southwest, which has withered under varying degrees of drought since the turn of the century.
But for other regions already prone to heavy rains, more precipitation could spell disaster.
"If you lived in Bangladesh"—where large areas flood annually—"you might not want to be reading this paper," Wentz said.
(Related news: "India Monsoons Intensifying, Hazard Risks Increasing, Study Says" [December 1, 2006].)
Back to the Drawing Board
Brian Soden, a professor at the University of Miami who was not involved with the study, said the new paper is likely to be controversial.
It suggests that "all models used to predict global warming underestimate the rate at which precipitation increases in response to surface warming," he said.
"If these models turn out to underpredict the global mean precipitation response, it seems plausible that they would also underpredict the increased frequency of heavy rain events," Soden said.
RSS's Wentz believes the real work begins now, helping computer experts reconcile the differences between their models and what the satellites show.
"I hope," he said, "we've put a little doubt in some modelers' minds."
Free Email News Updates
Sign up for our Inside National Geographic newsletter. Every two weeks we'll send you our top stories and pictures (see sample).
SOURCES AND RELATED WEB SITES