Improved Simulation of Frozen‐Thawing Process in Land Surface Model (CLM4.5)

Abstract

AbstractSoil freezing‐thawing cycle is a hallmark feature of the land surface over cold regions. Variations in soil moisture and temperature during frozen‐thawing (FT) process are important for water and energy exchange between land and atmosphere. Results of regional simulations by Community Land Model version 4.5 (CLM4.5) over the Tibetan Plateau show that error of daily soil temperature is within 3 °C and error of daily soil moisture is with 0.10 mm3/mm3 in FT process, averaged at 10 sites of whole Tibetan Plateau. Single‐point simulations show that model biases partly attribute to uncertainties of initial condition and soil category data; excluding impacts of model setting, large biases of soil moisture still exist during soil thawing, and CLM4.5 fails to simulate the diurnal cycle of soil moisture in this period. Modifications of the FT parameterizations in CLM4.5 are proposed, which include (1) use of virtual temperature (Tv) instead of constant freezing point to determine occurring of phase change, (2) introduction of phase change efficiency (ε) to optimize variation rate of soil temperature and moisture in FT process, and (3) consideration of inferred impacts of phase change on soil heat conduction. Single‐point and global simulations show that compared to original parameterizations in CLM4.5, these modifications can reproduce the features of daily and diurnal variations in soil moisture and make simulation in FT process closer to the observations.