Effects Of Atmospheric Forcing Data And Soil Texture On Soil Moisture Simulation By Land Surface Model

Land surface process is a general term for all kinds of physical,chemical and biological processes,and it is also one of the most important components of studying the complex processes of the climate system.Soil moisture affects the process of water circulation between the land and the atmosphere by affecting the surface evapotranspiration,and is one of the most important factors affecting the land surface process.However,because of the large spatial heterogeneity of soil moisture,the construction of measured sites is limited,and it is particularly difficult to obtain high quality and large scale soil moisture data.But using land surface model can simulate the continuous soil moisture data in time and space.The soil moisture of China in 2015 was simulated using different atmospheric forcing data and soil texture,and the following four groups of control experiments were designed.CLM_CTL uses the atmospheric forcing data from the China Land Data Assimilation System/Community Land Model(CLDAS/CLM)and use the soil texture data with pattern(FAO),CLM_HR uses the atmospheric forcing data from the High Resolution China Land Data Assimilation System/Community Land Model(HRCLDAS/CLM)and uses the Soil texture from FAO,CLM_SNSS uses the atmospheric forcing data from CLDAS/CLM and uses the soil texture data for land surface model(SNSS),CLM_NEW uses the atmospheric forcing data from CLDAS/CLM and uses the soil texture data for SNSS.The effects of atmospheric forcing data and soil texture on Soil Moisture Simulation of land surface model were studied by combining the four sets of experimental results with the measured data.The conclusions are as follow:(1)Compared with CLDAS/CLM,the atmospheric forcing data in HRCLDAS/CLM improved the data making method,merged more measured data and improved the spatial resolution from 0.0625° to 0.01°.Compared with CLDAS/CLM,the air temperature,ground pressure and relative humidity of HRCLDAS/CLM atmospheric forced data have a small decrease,and the numerical change is less than 5%.But the relative humidity changes greatly in winter,and the average change is about 15%.The overall reduction of wind speed in 10m is obvious,and the variation of data is between 10%-35%.However,the radiation data of HRCLDAS/CLM is larger than CLDAS/CLM,with an average variation of about 8%,while the precipitation data HRCLDAS/CLM has a significant decrease compared with CLDAS/CLM,especially in autumn,which is about 20%.According to CLDAS/CLM atmospheric forcing data simulation,soil moisture of CLM_CTL is higher than the measured data 0.lm3/m3,and the soil humidity of CLM HR based on HRCLDAS/CLM simulation is higher than the measured data 0.06m3/m3.The soil humidity is dry and closer to the measured data.This is due to the decrease of precipitation data and the increase of radiation data in HRCLDAS/CLM.(2)Comparing four soil layers(0-10cm)on two soil texture databases of SNSS and FAO,it is found that in the same soil database,the contents of sand and clay in vertical soil layer change little.Compared with the same soil layer,the percentage of sand in SNSS increased than FAO by 10%,while the percentage of clay content decreased by 20%.The difference was different in different regions,especially in Northeast of China,where the percentage of sand content increased about 20%.As for the comparison of simulation results of soil moisture,the simulation result of CLM_SNSS is drier than CLM_CTL.The RMSE and Bias from the measured data are reduced from 0.8TIm3/11m3 and 0.19rrm3/rrm3 to 0.5m3/m3 and 0·14m3/m3 respectively,it is proved that the simulated value of soil moisture decreases with the increase of sand content and the decrease of clay content,and the change of SNSS of the soil texture is closer to the real situation.(3)According to the simulation results of the four groups,the spatial distribution of soil moisture is consistent,and the gradient trend from the northwest to the southeast is from the drought to the moist.The decrease of precipitation and the increase of radiation in the data of atmospheric forcing will cause the simulated value of soil moisture to be smaller.The increase of sand content in soil texture and the decrease of clay content will also affect the hydrological characteristics of the soil itself,and reduce the simulated value of soil moisture.The degree of influence of the two depends on the degree of change in the factor itself.The simulation effect of land surface model on soil moisture has seasonal difference,and the simulation effect is best in summer,followed by autumn,and the simulation effect is poor in spring from the root mean square deviation and the correlation coefficient.