Earlier GCM studies have suggested that an enhancement of greenhouse warming might increase the occurrence of summer droughts in mid-latitudes, especially in southern Europe and central North America. This could represent a severe threat for the agriculture of the regions concerned, where summer is the main growing season. However, most studies featuring enhanced summer dryness in mid-latitudes generally used the ``bucket'' model, a very simple representation of the land surface processes known to exaggerate daytime evaporation, a behaviour which tends to emphasize summer dryness.

This study uses a regional climate model including a land-surface scheme of intermediate complexity to investigate the issue for the American Midwest. A surrogate climate-change scenario is used for the simulation of a warmer climate. The control runs are driven by observations, while the sensitivity experiments are forced by a modified set of initial and lateral boundary conditions. The modifications consist of a uniform 3K temperature increase and an attendant increase of specific humidity (unchanged relative humidity). The atmospheric CO2 concentration of the sensitivity experiments is set to four times its pre-industrial value. The simulations are conducted for the springs and summers of 4 years, corresponding to drought (1988), normal (1986, 1990) and flood (1993) conditions.

The numerical experiments do not present on average any large enhancement of summer drying under warmer environmental conditions. First, the overall changes in the hydrological cycle are of small magnitude despite the strong forcing applied. Second, precipitation increases in spring lead to a higher water storage during this season, compensating for the enhanced soil moisture depletion occurring later in the year. Though further processes would need to be accounted for in order to fully understand the potential mechanisms at play, our results suggest that the risk of enhanced summer dryness in the studied region might be less acute than previously assumed.