The soil hydrological processes of a force-restore type land surface modeling is re-examined by comparisons of several recent modifications from a variety sources in the current literature. It is found that during wet period of the warm season, after proper treatment of the average soil temperature lapse rate, correct specification of vegetation coverage, and a suitable initialization of the bulk soil moisture content, the land surface model in Advanced Regional Prediction System (ARPS), which is based on the Interaction between Soil, Biosphere, and Atmosphere model (ISBA), can satisfactorily simulate the soil moisture contents and surface fluxes, as compared with the direct observations from Oklahoma Atmospheric Surface layer Instrumentation System (OASIS) project. For another selected very dry period, it turns out the simulated soil moisture contents and surface fluxes are no longer sensitive to vegetation properties as stomatal conductance and leaf to canopy scaling, i.e., leaf area index (LAI). Rather, a parameter signifying the water extraction rate from the superficial layer is critical for the successful simulation of soil moisture contents and the surface energy fluxes. Several proposed parameterization schemes for this parameter are compared and the best results are obtained by combining two of the approaches.