Climate Feedback

The authors, Reed Maxwell at Lawrence Livermore National Laboratory, California, and Stefan Kollet of Bonn University, Germany, use actual climate data from 1999 as control conditions for the Great Plains watershed model. Then they ask how the results alter when a temperature rise and/or rainfall changes are added to that baseline.

This is where groundwater comes in. Write Maxwell and Kollet:

There is evidence that whereas drought timing may depend on sea surface temperature, the length and depth of major droughts in the region depend on soil moisture conditions and land-atmosphere interactions.

Their various climate simulations do show differences in soil moisture, as well as in the land-atmosphere interactions that contribute to soil-drying evaporation. And those differences are related to the flow of water through the soil, says Maxwell. Groundwater flows, he says, "dictate whether we’re going to be in a dust-bowl drought,” like the one the region experienced in the 1930s, “a mega-drought that’s beyond what we’ve seen on record, or a wet period.”

Where the water table is shallow - that is, where water saturates the soil at or near its surface, as in river valleys - the temperature rise of 2.5 C that Maxwell and Kollet test in the model is the main factor controlling whether the ground shrivels in the sun. On hilltops, which have deep water tables, precipitation is key because groundwater is too far from the surface to influence dry-out. And at an in-between groundwater depth of two to five metres - a sweet spot also identified in their previous paper (subscription, Water Resources Research) - groundwater level itself is the main influence.

This connection between water underground and moist, bountiful soils should be important in forseeing whether the Great Plains can remain the agricultural heartland of a warming US. But groundwater levels themselves depend on human responses to climate change: some solutions use more water than others. The new results suggest that groundwater shortages caused by intensive irrigation, for example, could feed back as worse climate change. And there’s every reason, Maxwell says, to expect that similar types of systems operate in other regions - though exact numbers such as the most influential water-table depths may differ from place to place.

"The water-management issue and the climate-feedback issue cannot be decoupled,” he adds. “These two are thought about in very different boxes, but they’re all part of the same thing.”

Anna Barnett

Photo: Soil blown by dust bowl winds piled up in large drifts in March 1936 near Liberal, Kansas; Library of Congress.

You can vote or comment on the importance of this paper in the Journal Club at Nature Reports Climate Change.