Using the Weather Research and Forecasting (WRF) model, we simulated the diurnal planetary boundary layer (PBL) evolution over the IHOP2002 experiment field (including the Oklahoma mesonet stations) on June 14, 2002. Since the simulation area was controlled by an anticyclone system, the PBL development was mainly affected by local surface forcing, while spatial variations in PBL heights were strongly influenced by variations in surface energy fluxes, which resulted from variations in soil moisture.

The simulation period is from 00:00 UTC June 14 to 12:00 UTC June 15, 2002. Three cases are run with the different initial soil moisture conditions to investigate the influence of soil moisture on PBL structure. In the case CTL, the initial and lateral conditions and the land surface conditions, including soil moisture and soil temperature, are forced by the data from NARR database. In the case WET, the initial soil moisture is replaced by the wettest soil condition in this summer (June, July and August), and in the case DRY, the initial soil is set as the driest condition.

The results show that soil moisture changes in the case WET and the case DRY cause changes in Bowen ratio, which lead the difference in surface energy partitioning. The averaged sensible heat flux increase in the case DRY and decrease in the case WET. These changes in sensible heat flux have great impacts on the PBL structure. A considerable increase in midday PBL heights occurs in the case DRY. A warmer and drier boundary layer develops in the case DRY, while a cooler and wetter one in the case WET. A more convective and higher PBL is present over the drier soil layer and a strong south-north gradient in PBL heights develops in the case DRY. This indicates that soil moisture heterogeneity greatly affects the spatial variations in PBL.