Improving Permafrost Physics in a Distributed Cryosphere-Hydrology Model and Its Evaluations at the Upper Yellow River Basin

Frozen soil undergoing freeze-thaw cycles has effects on local hydrology, ecosystems, and engineering infrastructure by global warming. It is important to clarify the hydrological processes of frozen soil, especially permafrost. In this study, the performance of a distributed cryosphere-hydrology model (WEB-DHM, Water and Energy Budget-based Distributed Hydrological Model) was significantly improved by the addition of enthalpy-based permafrost physics. First, we formulated the water phase change in the unconfined aquifer and its exchanges of water and heat with the upper soil layers, with enthalpy adopted as a prognostic variable instead of soil temperature in the energy balance equation to avoid instability when calculating water phase changes. Second, more reasonable initial conditions for the bottom soil layer (overlying the unconfined aquifer) were considered. The improved model (hereinafter WEB-DHM-pf) was carefully evaluated at three sites with seasonally frozen ground and one permafrost site over the Qinghai-Tibetan Plateau ("the Third Pole"), to demonstrate the capability of predicting the internal processes of frozen soil at the point scale, particularly the "zero-curtain" phenomenon in permafrost. Four different experiments were conducted to assess the impacts of augmentation of single model improvement on simulating soil water/ice and temperature dynamics in frozen soil. Finally, the WEB-DHM-pf was demonstrated to be capable of accurately reproducing the zero curtain, detecting long-term changes in frozen soil at the point scale, and discriminating basin-wide permafrost from seasonally frozen ground in a basin at the headwaters of the Yellow River.