Answers are grounded in this paper and cite their sources.
Evaluates soil moisture estimation improvements in hydrology through NISAR-derived soil roughness, suggesting more accurate agricultural insights.
Surface soil moisture (SSM) is an important soil parameter used in hydrology, agronomy and environmental sciences. SSM estimation using Sentinel-1 (S1) data is generally limited by inherently poor estimation of soil roughness, represented as root-mean‑square height (RMSH), leading to errors in SSM estimation. This letter uses simulated radar backscattering coefficients to assess the potential of integrating RMSH derived from the NISAR satellite into S1-SSM inversion neural networks (NNs). NISAR-RMSH was approximated by adding noise to reference RMSH, matching the expected root‑mean‑square error (RMSE) on RMSH estimation from three different NISAR configurations (single-polarization ‘SP’, dual-polarization ‘DP’ and quad-polarization ‘QP’). Using noisy synthetic data, four NNs were constructed, three with RMSH from NISAR and one without. Validation was performed on noisy synthetic data. Results show that using NISAR-RMSH information in the DP and QP configurations improves SSM estimation, with QP resulting in the best SSM estimation with an RMSE of 4.3 vol%, compared to 6.0 vol% without RMSH information. NISAR-RMSH also improves the stability of the NNs, with consistent RMSE and bias values across the range of RMSH values and significantly reduces the effect of radar incidence angle on SSM estimation. These findings demonstrate the potential improvement of S1-SSM estimation through NISAR-derived RMSH.
No takes yet. Share an insight, caveat, or question.
Najem et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: