Resumen
Accurate simulation of soil water and heat transfer is critical to understand surface hydrology under cold conditions. Using an extended freezing code in HYDRUS-1D (freezing module), this study was conducted: (1) to evaluate the freezing module using field data collected in a grazed steppe of Inner Mongolia; and (2) to further simulate grazing effects on frozen soil hydrological processes. The experimental data consisted of soil water and temperature profiles measured during freeze-thaw cycles from 2005 to 2006 in two plots (ungrazed since 1979 (UG79) and winter grazing (WG)). To check the sensitivity of the freezing module, a model without a freezing scheme (normal module) was used for comparison. We found that while the normal module can only simulate soil water and heat transfer under unfrozen conditions, the freezing module can simulate well under both frozen and unfrozen conditions. The freezing module can reasonably compute water phase change and, therefore, substantially improved the simulation of the evolution of liquid water and temperature in frozen soil. It overestimated liquid water content during spring snowmelt and, thus, underestimated surface runoff from underlying frozen soil layers. Furthermore, the weak prediction of soil moisture at the WG site, compared with the UG79 site, might relate to the less than ideal parameterization of soil hydraulic properties. Our results confirmed that the freezing module was able to accurately predict behaviors of soil freezing and thawing, as well as the effects of land management. We suggest that detailed knowledge of the soil-atmosphere processes is needed to improve the surface runoff algorithm in the frozen soil module.