چكيده به لاتين
In situ measurements to generate a time series of soil moisture is often not justified. Also, satellites that measure soil moisture directly have low spatial resolution for field-scale. In addition, remote sensing information provides only an observation of the surface layer of the soil and is limited to measuring soil moisture at depth. The use of remote sensing observational data and calibration of soil water movement simulation model by data assimilation method is proposed as an alternative to in situ measurements and reduces possible simulation errors. This method significantly increases the accuracy of the simulation by combining the results of the one-dimensional model and observational data. In this study, to estimate soil moisture and hydraulic properties, observations obtained from in situ measurements and satellite data in the HYDRUS-1D model are assimilated using the ensemble kalman filter (EnKF). Data assimilation was performed using three different observational sources including in situ measurements of soil moisture, land surface temperature (LST) from the LANDSAT-8 satellite and MODIS sensor. Comparison of the obtained results by considering the statistical index of NSE, RMSE,PBIAS and R^2 shows that in assimilation all three types of observations with increasing soil depth, the simulation accuracy decreases. Also, the assimilation data obtained from in situ measurement of soil moisture compared to satellite data provides the best estimate of soil moisture profiles. But in situ measurements of soil moisture are limited. In contrast, the spatial extent and ease of access to satellite data, the assimilation of soil surface temperature from the LANDSAT-8 satellite in the HYDRUS-1D model are a good alternative to using in situ measurements to estimate soil moisture profiles with acceptable accuracy.
