Soil Moisture Content Estimation Based on Sentinel-1 SAR Imagery Using an Artificial Neural Network and Hydrological Components

Abstract

this study estimates soil moisture content (SMC) using Sentinel-1A/B C-band synthetic aperture radar (SAR) images and an artificial neural network (ANN) over a 40 × 50-km2 area located in the Geum River basin in South Korea. The hydrological components characterized by the antecedent precipitation index (API) and dry days were used as input data as well as SAR (cross-polarization (VH) and copolarization (VV) backscattering coefficients and local incidence angle), topographic (elevation and slope), and soil (percentage of clay and sand)-related data in the ANN simulations. A simple logarithmic transformation was useful in establishing the linear relationship between the observed SMC and the API. In the dry period without rainfall, API did not decrease below 0, thus the Dry days were applied to express the decreasing SMC. The optimal ANN architecture was constructed in terms of the number of hidden layers, hidden neurons, and activation function. The comparison of the estimated SMC with the observed SMC showed that the Pearson’s correlation coefficient (R) and the root mean square error (RMSE) were 0.85 and 4.59%, respectively.

Estimation of soil moisture using environmental covariates and machine learning algorithms in Cathedral Peak Catchment, South Africa

“…Through these results, the dependence of SM on the terrain has allowed topography to be a principal predictor in many mapping applications (McBratney et al, 2003). This finding agrees with the works of Martínez-Murillo et al ( 2017), who found a significant influence of terrain on the SM, and Chung et al (2022), who found that the addition of terrain attributes aided in improving model performance in an ML approach for SM estimation. From the correlation plot, it could be assumed that the LST is not a major predictor for SM since it had the lowest correlation to SM, as shown in Figure 12, contrary to the findings of Mohamed et al (2020) and Ahmadi et al (2022).…”

“…(2017), who found a significant influence of terrain on the SM, and Chung et al. (2022), who found that the addition of terrain attributes aided in improving model performance in an ML approach for SM estimation. From the correlation plot, it could be assumed that the LST is not a major predictor for SM since it had the lowest correlation to SM, as shown in Figure 12, contrary to the findings of Mohamed et al.…”

“…The advantages of employing ML algorithms include rapidly producing high‐accuracy results and deciphering complex relationships (Khanzode & Sarode, 2020). Numerous literature has cited their success in the prediction of SM from observed and meteorological data (Babaeian et al., 2021; Sanuade et al., 2020) as well as from assimilating RS data (Adab et al., 2020; Chaudhary et al., 2022; Chung et al., 2022).…”

Soil Moisture Retrieval in Bare Agricultural Areas Using Sentinel-1 Images

“…Compared to utilizing either the neural network or the change detection alone, the hybrid approach boosts the correlation by 54% and 33%, respectively, highlighting its superior efficacy. Chung et al [18] trained an ANN model on Sentinel-1 SAR imagery to estimate soil moisture content Their model was trained using a variety of hydrological components, including soil texture, topography, and precipitation data using a leave-one-out approach. The ANN model that uses all the previously cited hydrological components exhibited superior performance in terms of accuracy compared to ANNs trained only on some components.…”

A Polarimetric Decomposition and Copula Quantile Regression Approach for Soil Moisture Estimation From Radarsat-2 Data Over Vegetated Areas

“…For the estimation of SSM in vegetation-covered areas, some vegetation scattering models have developed successively [29]- [31]. With the development of SAR and the increase of remote sensing data, some new methods gradually appear in the study of SSM estimation, such as the multi-angle method [32], [33], multifrequency method [34], [35], machine learning and artificial neural network method [36], [37], change detection method [38], [39], probabilistic graphical estimation method [40], [41], multi-polarization method [42]- [44] and so on. However, acquiring multi-angle and multi-frequency data simultaneously at the same location is challenging.…”