Soil Moisture Estimation Based on Polarimetric Decomposition and Quantile Regression Forests

The measurement of surface soil moisture (SSM) assists in making agricultural decisions, such as precision irrigation and flooding or drought predictions. critical challenge for SSM estimation vegetation-covered areas is the coupling between vegetation scattering. This study proposed an method based on polarimetric decomposition quantile regression forests (QRF) to overcome this problem. Model-based separates volume scattering, double-bounce while eigenvalue-based provides additional parameters describe scattering mechanism. combined use these explains SAR information from multiple perspectives, vegetation, roughness, SSM. As different crops differ morphology structure, it essential investigate potential varying estimate covered by crops. QRF, a applicable high-dimensional predictor variables, used parameters. In addition estimates, QRF can also provide predicted uncertainty intervals quantify importance estimates. performance was tested using data active passive validation experiment 2012 (SMAPVEX12) compared with copula (CQR). estimated consistent situ SSM, root-mean-square-error ranging 0.037 cm3/cm3 0.079 correlation coefficients 0.745 0.905. Meanwhile, both