Modeling the hydroclimatology of the midwestern United States: predicting soil moisture under a warmer climate

Climate change is likely to accelerate the hydrologic cycle, leading to enhanced global precipitation and evapotranspiration. In areas where the increase in evapotranspiration significantly exceeds that of precipitation, drought conditions will become more common. This could have extensive impacts on the entire world community if the newly created droughts occur in the midwestern United States or southern Europe, regions of substantial agricultural productivity. However, if the increase in precipitation exceeds that of evapotranspiration, little or no drying will occur.

A suite of 22-year numerical experiments were conducted to evaluate the ability of Regional Climate Model version 3 (RegCM3) to simulate the hydroclimatology of the midwestern United States and assess its response to climate change. RegCM3 was run using two surface physics schemes: Integrated Biosphere Simulator (IBIS) and Biosphere-Atmosphere Transfer Scheme 1e (BATS1e), and two convective closure assumptions: Fritsch & Chappell (FC80) and Arakawa & Schubert (AS74). Boundary conditions for control experiments were provided by the National Centers for Environmental Prediction-Department of Energy Reanalysis 2 and general circulation models. Future climate experiments were forced using a surrogate climate change scenario, in which the boundaries were warmed uniformly by 3°C, and general circulation models run under A1B emissions.

Overall, RegCM3 using IBIS and the AS74 convective closure assumption reproduces the observed seasonal cycles of the midwestern United States climate system best. Each model generally simulates incident surface shortwave radiation, absorbed surface shortwave radiation, downward longwave radiation, and net longwave radiation well. All models contain a significant wet bias and overestimate evapotranspiration during the spring and summer. Total runoff, surface runoff, and groundwater runoff are best simulated by RegCM3 using IBIS and the AS74 convective closure assumption. While BATS1e does capture the seasonal cycle of total runoff, gross errors in the partitioning of total runoff between surface runoff and groundwater runoff exist. The seasonal cycle of root zone soil moisture produced by RegCM3 using IBIS and the AS74 convective closure assumption is slightly dry, but otherwise agrees with observations. The rest of the models significantly underestimate root zone soil moisture.

RegCM3-IBIS and RegCM3-BATS1e simulate increased 2-m temperature, precipitation, evapotranspiration, and runoff during the spring and summer under all climate change scenarios. Two-meter temperature and precipitation are strongly influenced by boundary conditions and convective closure. Soil moisture is unchanged throughout the growing season as enhanced rainfall offsets increased evaporative demand. Negligible changes in soil moisture are robust across surface physics schemes, large-scale forcings, and convective closure assumptions.