Shen et al. (2025) Soil moisture retrieval under different land cover conditions based on Sentinel-1 SAR

Identification

• Journal: Remote Sensing of Environment
• Year: 2025
• Date: 2025-11-20
• Authors: Zhehui Shen, Qisheng He, Chun Yang, Zihao Cheng
• DOI: 10.1016/j.rse.2025.115147

Research Groups

• College of Geography and Remote Sensing, Hohai University, Nanjing 211100, China
• Jiangsu Province Engineering Research Center of Watershed Geospatial Intelligence (ERCWaGI), Nanjing 211100, China

Short Summary

This study systematically evaluates the performance of optical and radar-derived Vegetation Indices (VIs) in the WCM-Oh coupling model for soil moisture retrieval across five land cover types using Sentinel-1 SAR, proposing a new radar-based VI (SVIdp) and demonstrating improved accuracy in vegetated areas.

Objective

• To systematically evaluate the performance of optical and radar-derived Vegetation Indices (VIs) in the WCM-Oh coupling model for soil moisture retrieval under different land cover conditions using Sentinel-1C-band dual-polarization SAR data.
• To propose and test a new radar-based vegetation index, SVIdp, for soil moisture retrieval.

Study Configuration

• Spatial Scale: Regional, covering five distinct land cover types (grassland, crops, shrubland, forest, and desert).
• Temporal Scale: Not explicitly stated for the study period, but utilizes Sentinel-1 SAR data which provides long-term observations.

Methodology and Data

• Models used: WCM-Oh coupling model, WCM model.
• Data sources: Sentinel-1 C-band dual-polarization Synthetic Aperture Radar (SAR) data; optical and radar-derived Vegetation Indices (VIs).

Main Results

• The WCM-Oh model significantly improved soil moisture retrieval accuracy in typical vegetation areas (grasslands and crops), with the correlation coefficient (R) increasing by 0.10 and the root mean square error (RMSE) decreasing by 2 %–14 %, compared to the independent WCM model.
• The WCM-Oh2004 coupling model based on GNDVI demonstrated high prediction accuracy (R = 0.66–0.85, ubRMSE = 0.0392–0.0698 m³/m³) in areas with moderate vegetation cover (e.g., grassland A, herbaceous vegetation A/B, wheat/corn rotation areas). However, asymmetric errors (underestimation in low-value regions, overestimation in high-value regions) were observed in alpine grasslands, artificial lawns, sparse vegetation, and tea plantations.
• Biochemical optical indices (e.g., GNDVI, CVI) performed best in herbaceous vegetation areas, while structural radar indices (e.g., DpSVI, RVI) demonstrated superior stability in dense or structurally complex vegetation areas.
• The newly proposed SVIdp index outperformed DpRVI and NDVI in shrublands and grasslands.

Contributions

• Provides new insights into SAR-based soil moisture modeling under diverse land cover conditions.
• Systematically evaluates the performance of various optical and radar-derived Vegetation Indices within the WCM-Oh coupling model.
• Proposes and tests a novel radar-based vegetation index (SVIdp) based on dual-polarization scattering components.
• Supports large-scale SAR soil moisture spatial mapping.

Funding

• Not specified in the provided text.

Citation

@article{Shen2025Soil,
  author = {Shen, Zhehui and He, Qisheng and Yang, Chun and Cheng, Zihao},
  title = {Soil moisture retrieval under different land cover conditions based on Sentinel-1 SAR},
  journal = {Remote Sensing of Environment},
  year = {2025},
  doi = {10.1016/j.rse.2025.115147},
  url = {https://doi.org/10.1016/j.rse.2025.115147}
}