Evaluation of Optical Remote Sensing Efficiency in Estimating Soil Surface Moisture and Comparing It with Thermal Data for Irrigation Management of Sugarcane

Soil moisture is one of the most important parameters in water, soil and plant resources management. Therefore, the present study was conducted to evaluate the efficiency of thermal and optical remote sensing data in order to estimate soil moisture and irrigation planning in sugarcane fields of Khuzestan Province, Iran. For this purpose, soil moisture content for 9 passes of Landsat 8 and Sentinel 2 satellites was calculated using thermal and optical trapezoidal methods from April to October 2020 in Amirkabir Sugarcane Agro-industry fields. To validate the results, the measured soil moisture content data of 337 ground control points located in 18 sugarcane-growing fields measured by TDR350 dehumidifier were used simultaneously with the passage of the satellites. The results showed that TOTRAM model with a determination coefficient of 0.82 and error rate of RMSE and NRMSE as 4.45% and 12.9%, and OPTRAM model with an explanation coefficient of 0.93 and RMSE and NRMSE error of 3.14% and 12.1% were able to properly estimate soil surface moisture in sugarcane fields. Also, the results of evaluation of soil moisture maps for irrigation planning of sugarcane fields showed that these data could be used for irrigation planning with average NRMSE error of 16% and 9% in relation to ground irrigation time data for TOTRAM and OPTRAM models, respectively. In this regard, OPTRAM model data were more efficient compared to thermal data, due to better spatial resolution of optical data and less effect by environmental factors such as temperature and relative humidity of air and also the effect of adjacent pixels.