Minor changes in precipitation may have major impact on groundwater supplies in arid regions

December 6, 2010

By Cathryn Delude
Correspondent, Civil & Environmental Engineering

People living in the great expanses of flat, dry regions of the world that lack natural freshwater lakes depend on episodic rainstorms to replenish the groundwater they use for domestic and agricultural purposes. New work by researchers in CEE shows that even minor variations in precipitation caused by climate change could cause major, often counterintuitive changes to fragile groundwater supplies in these areas.

For example, if the future brings drier winters but wetter summers to the Texas High Plains, annual rainfall could increase 10 percent but recharge — the water that soaks deep into the ground to replenish groundwater — would drop 10 percent, largely because the area’s rain-fed cotton crops would soak up the summer rain. Likewise, 20 percent less precipitation, falling mainly in summer, could reduce groundwater recharge by 75 percent, but because of the rain’s timing, cotton crops might still be sustainable. Alternatively, 10 percent more annual precipitation falling in intense spring storms before the growing season begins could yield 20 percent more recharge. These patterns also match historical rainfall variability and recharge response patterns, the researchers found.

“The most important factors affecting diffuse groundwater recharge are the intensity, timing and duration of discrete rainstorms,” said Professor Dennis McLaughlin, a co-author of two related articles in the September 2009 and July 2010 issues of Water Resources Research. “Climate change models, which can differ significantly, make regional forecasts over relatively long time scales, but the mechanisms that affect recharge are very localized and sporadic.” 

Uncertainties about local short-term variability make it difficult to estimate recharge in the current climate, much less predict conditions in a changing future. For that reason, McLaughlin, Professor Dara Entekhabi, recent graduate Gene-Hua Crystal Ng, (now at the U.S. Geological Survey in Menlo Park, Calif.), and Bridget Scanlon of the University of Texas, Austin used a probabilistic framework to understand the dynamics of groundwater recharge.

For the first study, they analyzed the mechanisms that control groundwater recharge in the southern High Plains. They combined past rainfall and meteorological records with measurements of soil moisture and chloride ion profiles deposited by rainfall, using procedures similar to those developed to analyze ice-core records. After analyzing 71 years of historical weekly recharge rates, they found that a few intense storms during the fallow season account for most of the annual recharge, while steady light rains during the growing season contribute relatively little

For the second study, the researchers replaced the historical data with future weather “samples” simulated by a weather generator program. This program essentially rolled the dice thousands of times to simulate random rainstorms that reproduce the monthly rainfall distribution predicted by each of five possible future climate scenarios: wetter, more intense storms, more seasonal variation, drier, and very dry. With this, the researchers could assess probable changes in the timing, frequency and