Quantifying Parameter Sensitivity and Calibration in Simulating Soil Temperature and Moisture

The Shuffled Complex Evolution Algorithm is used to estimate the parameters in the Noah model, and soil hydraulic parameters are also estimated by soil texture components to evaluate the sensitivity of model hydraulic parameters in simulating soil moisture and soil temperature. The results show that soil hydraulic parameters are more sensitive than parameters related to vegetation, and more sensitive to soil moisture than soil temperature. When surface soil moisture is used to calculate the objective function, the estimated hydraulic parameters can not only improve the soil moisture, but also soil temperature. However, when deep soil moisture is used to calculate the objective function, the soil temperature simulation is even worse than with default parameters. Soil texture components are also used to calculate soil hydraulic parameters; and when they are used as estimated parameters, the results can improve both soil moisture and soil temperature simulation. Parameter estimation experiments in which soil temperature is used to calculate the objective function cannot improve the simulation of soil moisture. When soil texture components are used as estimated parameters and soil moisture and soil temperature are used to calculate the objective function, the estimated hydraulic parameters can improve both soil moisture and soil temperature simulation. Decreasing the number of parameters has little impact on the results. Accordingly, when soil texture components are used as estimating parameters in this land surface model, the estimated hydraulic parameters show good consistency-more so than when estimating hydraulic parameters directly.