Answers are grounded in this paper and cite their sources.
Effectively calibrating microwave scattering models and accurately simulating the soil–vegetation interaction of backscatter coefficients (σo) are prerequisites for improving soil moisture (SM) retrieval. Uncertainties in observations or parameterization methods produced unreasonable calibrated coefficients of the semiempirical water–cloud model (WCM). This leads to unrealistic simulations of vegetation scattering components and transmissivity (τ2). Two physically based constraint strategies were proposed to limit the range of the WCM coefficients, A and B. Constraint 1 limited the coefficient A by using the ranges of observed σo veg and assuming that τ2 was zero at the peak biomass stage of corn. Constraint 2 proposed two limitation curves based on the upper and lower bounds of τ2 under the corresponding vegetation conditions to limit the coefficient B. By introducing two constraints and applying the least squares method, the coefficients (A and B) were derived from the WCM-Oh04 simulated signals (σo sim) and the uninhabited aerial vehicle synthetic aperture radar observed signals (σo obs). The WCM-Oh04 model was parameterized based on five vegetation descriptors (VDs): vegetation water content (VWC), VH, cross-polarization ratio (Rvhvv), radar vegetation index, and dual-polarization radar vegetation index, and the performance of σo simulations was evaluated. For VV and HH polarizations, VWC and VH performed well as the VDs on both the training and validation sets, with high correlation (r > 0.68) and low error (ubRMSE < 1.91 dB). For VH polarization, Rvhvv showed the highest correlation (r ≥ 0.85) and lowest error (ubRMSE ≤ 2.31 dB) on both the training and validation sets. In addition, VH as the VD had the highest SM retrieval accuracy with an r of 0.35 and a ubRMSE of 0.114 m3/m3. When considering both corn phenology and the spatial resolution of σo (upscaled from 30 to 800 m), the SM retrieval accuracy using VH as the VD further improved, with the r increasing to 0.849 and the ubRMSE decreasing to 0.047 m3/m3. These findings indicate that constraining the WCM-Oh04 using physically based knowledge produces reliable results. This method retains the physical mechanism while reducing computational costs, which is beneficial for global SM monitoring based on active microwave remote sensing.
No takes yet. Share an insight, caveat, or question.
Feng et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: