Soil water and the persistence of floods and droughts in the Mississippi River Basin

In six independent multiyear simulations of the National Center for Atmospheric Research (NCAR) Community Climate Model (CCM3) coupled to the NCAR Land Surface Model (LSM 1.0), the Mississippi River Basin stands out as a region in which surface air temperature is, during summer months, negatively correlated with soil water and precipitation. To determine if these correlations represent, in part, a land surface feedback in which wet soils caused by high rates of precipitation lead to higher latent heat fluxes, creating a cooler, moister atmosphere, and vice versa, several experiments in which soil water was initialized to either wet (saturated) or dry (wilting point) were performed. In an ensemble of five wet and five dry 60-day simulations for summer conditions, initially wet soils resulted in cold surface air temperature compared to initially dry soils. Latent heat flux increased and sensible heat flux decreased in the wet experiment compared to the dry experiment only for the first 30 days even though the surface was colder and wetter than the dry experiment throughout the full 60 days. Precipitation increased for the initially wet soils compared to the dry soils over much of the Mississippi River Basin, but only during the first 30 days. This increase was small, and most of the increased flux of water into the atmosphere was advected out of the region. The magnitude of the wet-dry signal depended on the manner in which soil water limited evapotranspiration. A step-function response, in which soil water did not limit evapotranspiration until near wilting point, produced a stronger signal than a linear ramping function scaled between field capacity and wilting point. These results suggest that soil moisture feedbacks amplify the severity and persistence of floods and droughts and that the manner in which soil water limits evapotranspiration, which is poorly understood, is a key control of this response.