Comparison of different soil temperature algorithms in permafrost regions of Qinghai-Xizang (Tibet) Plateau of China

Soil thermal diffusivity is a crucial physical parameter that affects soil temperature. By applying sinusoidal boundary conditions, an analytical solution using separated variables for the heat conduction-convection equation was developed. The thermal diffusivity and liquid water flux density were calculated with data collected at field observation sites in permafrost regions of Qinghai-Xizang (Tibet) Plateau (QXP). By taking the soil layer at the depth of 5 cm as the upper boundary, the soil temperature at a depth of 10 cm was modeled by the thermal conduction-convection method, amplitude method and phase method. The statistical analysis of the standard error of the estimate (SEE), the normalized standard error (NSEE) and the root mean square error (RMSE) demonstrated that the thermal conduction-convection method provided the most accurate prediction of soil temperature, with average SEE, NSEE, and RMSE of 0.72 degrees C, 9.26% and 0.72 degrees C, respectively. The thermal conduction-convection method provides a useful tool for calculating soil thermal parameters, simulating soil temperature and land surface processes parameterization for permafrost changes modelling under global warming. (C) 2016 Elsevier B.V. All rights reserved.