Soil Profile Moisture Simulation Study Based On Distributed Dynamical Model Driven By The ET Of Satellite Remote Sensing

Soil water is an important junction between groundwater and surface water,and an important link of water resources transformation and hydrological cycle.The movement law and accurate simulation of soil water is a hot academic issue in current international research.It is very important for agricultural water saving and regional ecological environment.The study of soil moisture provides a solid theoretical basis for agricultural water use and natural water resources cycle.It has certain guiding significance to grasp the law and optimize the agricultural irrigation system.In this study,based on MODIS remote sensing data,the distributed simulation of evapotranspiration(ET)was realized by SEBS(Surface Energy Balance System)model,which drived the soil hydrodynamic model.At the same time,the measured soil water content data of Daxing soil moisture monitoring point were used to verify the simulation of the distributed soil profile moisture during the growing period of Summer Maize.The main conclusions are as follows:(1)Remote Sensing interpretation of EvapotranspirationThe ET obtained by remote sensing method is more efficient and accurate than the traditional method.At the same time,the ET obtained by remote sensing method can better reflect the spatial variation law,so it is suitable for estimating ET in larger areas.In this study,the ET was simulated on 500 m spatial scale with MODIS data,and the validity of SEBS model was verified by the measured values of Eddy covariance system.According to the actual situation of the study area,the underlying surface parameters of the model were improved to improve the simulation effect of the model on ET in the study area.The results showed that the energy closure of Eddy covariance system of the experimental station reached 0.89 and 0.79 in the growth period of Winter Wheat in 2014 and 2015,respectively.Therefore,it can be considered that the energy closure of the station was relatively good.For the unclosed part,energy closure was achieved by assuming that sensible heat flux was accurate,direct subtraction of latent heat flux was low,and Bowen's ratio method which considered sensible heat flux and latent heat flux were lower than the actual value.The results showed that the simulated values of latent heat flux in the growing period of Winter Wheat in 2014 and 2015 were 313w/m~2 and 277w/m~2,respectively.The relative root mean square error were 26.8 w/m~2 and 28.5 w/m~2,and the certainty coefficient were 0.737 and 0.840.Therefore,it could be concluded that SEBS model can better reflect the spatial and temporal distribution pattern of ET in the study area,and it was feasible to apply this data to the soil water movement model.Therefore,it was feasible to use SEBS remote sensing evapotranspiration data to drive the soil water dynamics model.(2)Simulation of Soil relative humidity based on statistical methodBased on the statistical method of remote sensing data,the soil relative humidity was simulated in the study area.The statistical model was divided into two methods:Temperature Vegetation Drought Index Model(TVDI)and apparent thermal inertia model(ATI).The simulated mean values of soil relative humidity by TVDI model and ATI model were 71.06%and 92.55%,the root mean square error were 3.76%and3.52%,and the deterministic coefficients were 0.56 and-0.11,respectively.The TVDI model was used in the area with higher vegetation index when the normalized vegetation index(NDVI)was at the threshold of 0.2.On the contrary,the simulation effect of ATI model was better than that of two models when they were simulated separately.It can be seen that the statistical method based on remote sensing data can obtain more accurate spatial and temporal distribution of soil relative humidity.(3)Distributed Soil Moisture Model driven by SEBS ModelBased on the principle of soil hydrodynamics of Hydrus model,this study simulated the soil moisture content in the region.At the same time,the ET data of SEBS model was used to drive the soil hydrodynamics model,and the soil profile moisture content was accurately simulated on the GIS platform,so as to realize the construction of the soil moisture distributed model.The results showed that the simulated soil water content of 40-60 cm in summer maize growth period was better,the simulated mean value was 13.40 m~3/m~3,the root mean square error was 3.9 m~3/m~3,and the certainty coefficient was 0.734.Because of the difference of principle between the statistical model and the soil hydraulics model,the former needs to calibrate the specific research area before using,so the applicability is poor;the latter uses the mechanism model,which can be used to simulate soil moisture conveniently and accurately only depending on meteorological data and remote sensing data after modeling.Therefore,it can be concluded that the distributed soil moisture model driven by satellite remote sensing evapotranspiration data can more efficiently and conveniently obtain the temporal and spatial variation of soil profile moisture.