The Simulation And Assessment Of Soil Moisture In China Area At Different Depth

Land surface processes is the continuous exchange processes of momentum, energy and substances between underlying surface and atmosphere. Soil moisture is one of the most important physical quantities of land surface process. This paper constructs an atmospheric forcing data, which was derived by combining retrieved products of the high spatial and temporal resolution precipitation and the surface incident solar radiation from geo-stationary satellite FY-2with the interpolated NCEP/NCAR reanalysis data. Then, this paper uses the atmospheric forcing data to drive CLM3.0, and consequently soil moisture products of China area with high solution and time space continuity are obtained. The evaluation of the CLM3.0model simulation results at different depth is conducted by the comparison to agricultural meteorology soil moisture observation data. The conclusions are summarized as follows:(1) The comparison to average state of the CLM3.0simulated soil moisture results and site observation data shows that model simulation results can reflect the spatial and temporal characteristics of soil moisture. The simulation results and observation data are consistent in time variation and space distribution. It’s found that the CLM3.0model simulation results at different depth is consistent with observations in space distribution, time change rate, and scale. The shallow results are better than deep ones.(2) From the error distribution of model simulation results with site observation data, except few sites, the statistics including bias, root mean square error and relative error are small, the correlation coefficient of shallow results are better than deep ones. Among them, comparatively speaking, the bias of north region is less than the south, the smallest root mean square error areas are Northeast and North China, the southern region of the relative error is less than the northern region, the highest correlation coefficient are in Northeast and Southwest regions, the best simulation seasons are summer and autumn. the statistics including bias, correlation coefficient, root mean square error and relative error in station scale show that the model has good simulation results in China area.(3) The whole China area is divided into eight regions according to climatic characteristics. This paper analyses the temporal variations of the regional average soil moisture. Among the results, the best simulation effect is the west of northwest region, which have good performance in three different depth; followed by the two southwest regins, which shallow simulation results is better than deep ones; low correlation coefficient; then followed by northeast regin and north regin. Overall, from the time series analysis of eigth regions, shallow simulation results and observational data are consistent changes over time, and have feedback to pre-precipitation; but deep simulation results have less time rate of change.(4) The land surface model is real-time operated and its application to drought monitoring. All the parameters needed by constructing atmospheric forcing data for CLM3.0model can be calculated just in real time including:precipitation, downward solar radiation, air temperature of near surface, relative humidity, wind speed. Then they are automatically downloaded and processed with projecting, interpolating, supplying missing data. CLM3.0can be run daily, and output various maps to monitor drought in China. The distribution of stimulated soil moisture corresponds well to drought area in winter in2009and in spring in2010in China.