Increased rainfall and droughts as a result of global temperature rise could be two to three times more severe than climate models predict. That's according to satellite data analysis by researchers at the University of Reading, UK, and University of Miami, US.

"The discrepancy we found between the models and the satellite data adds to an emerging body of evidence suggesting that models may underestimate the increases in intense rainfall in response to human-caused warming," Richard Allan of the University of Reading told environmentalresearchweb.

Based on standard climate models and current rates of global temperature change, the Intergovernmental Panel on Climate Change (IPCC) report of 2007 forecasts a 10 to 20% increase in equatorial and high-latitude precipitation, and a 10 to 20% reduction in the dry sub-tropics on 1990 levels by 2095.

However, after analysing their satellite data, Allan working with Brian Soden from the University of Miami, suggest these extremes could be two to three times wider than the models predict.

"When we looked more closely at the results, the models appeared to underestimate the response of the heaviest rainfall frequency to warming," said Allan.

The research was inspired after the IPCC identified the serious ramifications of climate model predictions of increasingly intense rainfall and greater incidence of drought on human health, agriculture and ecosystems.

"It is crucial to test the fidelity of such projections using observations," said Allan. "Our strategy was to use the natural periodic changes in global temperature, associated with El Nino Southern Oscillation (ENSO), to examine the response of extreme precipitation events to changes in climate."

By comparing daily satellite observations with model data during a 20 year period, the researchers confirmed that heavy precipitation events are more frequent during warm, moist periods and less frequent during cold, dry periods. The surprising result was the extent to which temperature rise widens precipitation extremes.

"Warmer air holds more moisture - about seven per cent more for each degree centigrade rise in temperature," explained Allan. "Warmer air is therefore likely to also enhance the intensity of convective rainfall at a similar rate."

On a global scale there is a balance between radiative cooling of the atmosphere and latent heating through condensation and precipitation, Allan says. "If latent heating is more than radiative cooling, the atmosphere will warm up and become more stable thus inhibiting convection and reducting latent heating to balance again."

As the atmosphere warms up due to climate change, radiative cooling will also increase because there is more thermal emission. To balance this rise in radiative cooling there must be an increase in latent heating, i.e. more precipitation. "The slow rise in radiative cooling constrains average precipitation to rise only at one to three per cent for each degree centigrade of warming," said Allan. "Since average precipitation is constrained to rise at a slower rate than convective region precipitation, rainfall in the drier regions away from convection must diminish."

So the rain debt must be repaid by regions away from convective areas. Unfortunately, the largest debt is paid by the already arid sub-tropical regions.

The next step is to try and determine the cause of the discrepancy between the climate models and satellite observations.

"It is first important to establish that the satellite measurements are robust and not subject to previously overlooked limitations," said Allan. "Further research will aim to quantify the effects of changes in aerosol particles that may be influencing present day changes in rainfall. It is also possible that higher space resolution models will improve the representation of convective processes."