Estimation of soil moisture via semiempirical and machine learning methods using SAR satellite images: A comparative field study

Remote sensing technology is widely used in many fields thanks to its ability to observe in a wide range and to collect data on time and with less constraints. Agriculture is one of the most important areas for the application and research of remote sensing. Active microwave SAR sensors are suitable for large-scale soil moisture estimation due to their high resolution, not affected by weather conditions and sensitivity to soil moisture. In recent years, in the field of remote sensing nonlinear machine learning techniques have been recommended for soil moisture estimation. In this study, surface soil moisture was retrieved from Radarsat-2 and polarimetric target decomposition data by using semiempirical models and machine learning methods. The semiempirical models and machine learning techniques employed were Oh (1992), Dubois (1995), Oh (2004) and Generalized Regression Neural Network (GRNN), Least Squares – Support Vector Machine (LS-SVM), Extreme Learning Machine (ELM), Kernel based Extreme Learning Machine (KELM), Adaptive Neuro-Fuzzy Inference System (ANFIS), respectively. In addition, Yamaguchi, van Zyl, Freeman-Durden, H/A/α and Cloude polarimetric target decomposition methods were used in this study. For soil moisture inversion, firstly, preprocessing was applied to the Radarsat-2 image of two different dates with bare and moderately vegetated soil. Then, sigma nought coefficients and the polarimetric decomposition components were extracted as feature vector from preprocessed SAR image pixels corresponding to ground measured points. Lastly, sigma nought coefficients were used in semiempirical inversion models, and sigma nought coefficients and polarimetric decomposition components were used as input to machine learning methods. The best accuracy results for semiempirical models were 13.01 and 17.91 Root Mean Square Error (RMSE) for bare and moderately vegetated soil, respectively. The best accuracy for machine learning techniques were 4.04 and 2.72 RMSE for two dates, respectively. The results indicated that the machine learning techniques performed much better than the semiempirical models.