Evaluation of Standalone In-situ Simulations of Frozen Soil

The performance of a climate model to reproduce frozen soil depends on the modeled atmospheric forcing and the parameterizations in the land surface. Due to the complex land-air interactions caused by snow and soil freezing and thawing, biased simulations of climate models may be compensated or amplified by errors in land surface models. This may lead to a misjudgment of the simulation capabilities of the land surface model itself, especially when we are trying to improve the overall performance of the climate model without being able to balance the results in frozen soil. In order to separately investigate the simulation performance of the land surface model in the frozen soil region, we conduct simulations using the stand-alone land surface models CLM5, TERRA, and JSBACH at representative sites in Siberia, Alaska, and the Tibetan Plateau and explore the performance of the models from daily to interannual scales using the same atmospheric forcing and initial conditions. The main evaluation objects will be the insulating effect of snow, soil energy balance, and soil moisture transportation to a depth of 3 meters below the ground surface. We look forward to the offline simulation experiments to evaluate the accuracy of different land surface model simulations, the optimal soil hydrothermal parameterization scheme, and important physical processes that may be neglected by the models’ prediction of frozen soil in daily and monthly time scales.