Many people are afraid that the global warming will turn many areas into deserts, besides flooding low areas and melting the mountain glaciers and polar ice caps.

But a new analysis of 20 years of satellite data shows that rising temperatures will bring in fact more rainfall, challenging classic climate model concepts and could help researchers forecast with greater accuracy dramatic weather events like El Niño.

It has been known for a long time that with each additional degree Celsius added by the global warming, the planet gets an extra 7% of water into its atmosphere.

Previous computer models forecast that an increase in the atmospheric wetness will only increase rainfall by 1% to 3% for each supplementary degree with which the temperature rises. This was assumed because, even if there is more water in the atmosphere, the rainfall amount and evaporation slows down. But these models were found to underestimate rainfall and neglect great weather events like the 1998 El Niño in simulating weather patterns over the past 2 decades.

That's why a team at Remote Sensing Systems, a satellite analysis company in Santa Rosa, California, fixed the models. The researchers employed real historical data gathered from six satellites to assess the relations between total atmospheric water, precipitation, evaporation, and global temperature.

Precipitation and evaporation shifted in connection with total atmospheric water: a rise of 6.5% for each additional degree Celsius for Earth's temperature, thus evaporation and precipitation levels are directly connected with the level of humidity in the atmosphere.
The problem is that this is an overall data, and the annual rainfall level can vary greatly around the globe, and some regions can experience drought. "In the tropics, you would get as much as 65 millimeters of water, whereas in the northern latitudes, it might only be a few millimeters," said physicist Frank Wentz from the research team.

"The study is the first to question the accuracy of precipitation in current climate models. There are dozens of different climate models out there, and every single one of them predicts that precipitation will increase more slowly than this study suggests. Plus, they all get the historical record wrong. Improving these models could help climatologists better predict future storms," said climatologist Brian Soden of the University of Miami in Florida.