Balouei et al. (2025) A novel high-resolution soil-moisture mapping using Sentinel-1-imagery and optimization-based for a new precise remote sensing drought index

Identification

• Journal: International Journal of Environmental Science and Technology
• Year: 2025
• Date: 2025-12-06
• Authors: Fatemeh Balouei, Mostafa Kabolizadeh, Hamidreza Rabiei‐Dastjerdi
• DOI: 10.1007/s13762-025-06953-w

Research Groups

• Faculty of Earth Sciences, Shahid Chamran University of Ahvaz, Ahvaz, Iran
• Department of Remote Sensing and GIS, Faculty of Earth Sciences, Shahid Chamran University of Ahvaz, Ahvaz, Iran
• School of History and Geography, Faculty of Humanities and Social Sciences, Dublin City University (DCU), Dublin, Ireland

Short Summary

This study developed a novel high-resolution (10 m) Optimized Soil Moisture Condition Index (OSMCI) for agricultural drought monitoring in Khuzestan Province, Iran, using Sentinel-1 imagery and optimization algorithms, demonstrating superior accuracy compared to existing coarse-resolution indices.

Objective

• To bridge gaps in agricultural drought monitoring by leveraging Sentinel-1 imagery for high-precision soil moisture simulation as an alternative to global soil moisture datasets.
• To evaluate the performance of metaheuristic (Particle Swarm Optimization) and traditional (Interior Point method) optimization methods in remote sensing drought indices.
• To assess the utility of Sentinel-1 imagery for drought monitoring at a spatial resolution of 10 meters.
• To introduce a new agricultural drought index, OSMCI, exclusively based on Sentinel-1 data.

Study Configuration

• Spatial Scale: Plains of Khuzestan Province, Iran (48° to 49° E longitude, 30° to 32° N latitude), specifically Amirkabir Sugarcane Farms (48°10' to 48°25' E longitude, 30°50' to 31°50' N latitude). Data resolution for Sentinel-1 is 10 meters.
• Temporal Scale: Sentinel-1 and OSMCI analysis from 2017 to 2023. Ground meteorological data (SPI) from 1997 to 2023. Soil moisture measurements were conducted on six specific dates.

Methodology and Data

• Models used:
  • Optimization Algorithms: Particle Swarm Optimization (PSO) and Interior Point method (implemented via MATLAB’s fmincon function) for determining optimal weights.
  • Drought Indices: Optimized Soil Moisture Condition Index (OSMCI) and Soil Moisture Condition Index (SMCI) for comparison.
  • Preprocessing: Range-Doppler Terrain Correction (using SRTM DEM) for geometric correction, Lee filter (5x5 window) for speckle noise reduction.
• Data sources:
  • Satellite Imagery: Sentinel-1A and Sentinel-1B (C-band Synthetic Aperture Radar, Ground Range Detected data, Interferometric Wide mode, VH and VV polarizations, 10 m spatial resolution) accessed via Google Earth Engine (GEE).
  • Ground Observations:
    • Standardized Precipitation Index (SPI) from 35 ground meteorological stations (1997-2023, 3-, 6-, and 9-month intervals).
    • Gravimetric soil moisture measurements from 80 Amirkabir Sugarcane Cultivation and Industry experiment farms (at 0-30 cm, 30-60 cm, and 60-90 cm depths, with an average at 45 cm).
  • Global Soil Moisture Products (for comparison):
    • Soil Moisture Active Passive (SMAP) satellite mission (9 km spatial resolution, root zone soil moisture).
    • Famine Early Warning Systems Network Land Data Assimilation System (FLDAS) (11 km spatial resolution, SoilMoi40_100cm).
    • TerraClimate (4638 m spatial resolution, monthly soil moisture).
  • Digital Elevation Models (DEMs): Shuttle Radar Topography Mission (SRTM) DEM (30 m spatial resolution) for geometric correction.

Main Results

• The developed OSMCI, utilizing PSO and Interior Point methods, achieved high accuracy in soil moisture simulation, with Root Mean Square Error (RMSE) values below 0.022, Median Absolute Deviation (MAD) below 0.019, and Mean Squared Error (MSE) below 0.01 during both training and validation phases.
• PSO demonstrated a slightly higher correlation (R ≈ 0.50) in soil moisture simulation compared to the Interior Point method (R ≈ 0.32-0.43).
• OSMCI exhibited superior performance in drought assessment compared to the conventional SMCI index derived from global products (SMAP, FLDAS, TerraClimate) at most ground stations.
• The Normalized Root Mean Square Error (NRMSE) for OSMCI ranged from 0.2 to 0.3 across most stations, indicating a high confidence level (>70%) and outperforming the SMCI indices from global products, which often exceeded this threshold.
• The correlation coefficient (R) between OSMCI and the ground-based SPI (3-, 6-, and 9-month intervals) was generally between 0.42 and 0.62 at most stations, indicating an acceptable correlation.
• OSMCI provided significantly higher spatial resolution (10 m) for drought mapping, enabling detailed observation of local environmental changes, a notable advantage over the coarser resolutions of SMAP (9 km), FLDAS (11 km), and TerraClimate (4.6 km).
• The temporal trend of the OSMCI index closely mirrored that of the SPI, showing parallel transitions between drought and wet conditions from 2017 to 2023.

Contributions

• Development of a novel, high-resolution (10 m) agricultural drought index (OSMCI) based on Sentinel-1 Synthetic Aperture Radar (SAR) imagery, offering a significant improvement over existing coarse-resolution methods.
• Successful integration and evaluation of both metaheuristic (PSO) and traditional (Interior Point method) optimization algorithms for determining optimal weights in a remote sensing drought index, enhancing model flexibility and accuracy.
• Demonstration of the superior performance of the Sentinel-1-based OSMCI for local-scale agricultural drought monitoring compared to the conventional SMCI index derived from global soil moisture products (SMAP, FLDAS, TerraClimate).
• Overcoming the limitations of optical sensors (e.g., cloud cover) by relying solely on radar data, ensuring continuous and reliable data acquisition for drought monitoring.
• Provision of a flexible and adaptable framework for drought index development that can be tailored to specific regional conditions, improving water resource management and agricultural planning.

Funding

• Shahid Chamran University of Ahvaz, project code SCU.EG403.827, awarded to Mostafa Kabolizadeh, effective 01/06/2024.

Citation

@article{Balouei2025novel,
  author = {Balouei, Fatemeh and Kabolizadeh, Mostafa and Rabiei‐Dastjerdi, Hamidreza},
  title = {A novel high-resolution soil-moisture mapping using Sentinel-1-imagery and optimization-based for a new precise remote sensing drought index},
  journal = {International Journal of Environmental Science and Technology},
  year = {2025},
  doi = {10.1007/s13762-025-06953-w},
  url = {https://doi.org/10.1007/s13762-025-06953-w}
}