Petropoulos et al. (2025) RegreSSM: A novel software tool for downscaling the SMAP L3 soil moisture operational product utilizing the Ts/VI feature space and Sentinel-3 data

Identification

• Journal: Environmental Modelling & Software
• Year: 2025
• Date: 2025-12-18
• Authors: George P. Petropoulos, Spyridon E. Detsikas, Vasileios Anagnostopoulos, Christina Lekka
• DOI: 10.1016/j.envsoft.2025.106836

Research Groups

• Department of Geography, Harokopio University of Athens, Athens, Greece
• Agile Actors Hellas S.A., Halandri, Greece

Short Summary

This study introduces RegreSSM, a novel Python-based software tool for downscaling the SMAP L3 surface soil moisture product from 36 km to 1 km resolution using Sentinel-3 optical and thermal data and the Ts/VI feature space. The tool provides a user-friendly, automated, and reproducible workflow, demonstrating satisfactory soil moisture retrieval accuracy over the Iberian Peninsula.

Objective

• To develop and present RegreSSM, a user-friendly, standalone software tool that enables the automated and reproducible downscaling of the SMAP L3 Surface Soil Moisture (SSM) operational product from 36 km to 1 km spatial resolution by fusing Sentinel-3 optical and thermal data based on the Ts/VI feature space.

Study Configuration

• Spatial Scale: Downscaling from 36 km to 1 km. The demonstration was conducted over the Iberian Peninsula, specifically the Duero River Basin in north-western Spain (REMEDHUS network, approximately 1300 km²).
• Temporal Scale: Daily SMAP L3 products (06:00 a.m. and 18:00 p.m. local times). Sentinel-3 data with approximately 1-2 days temporal resolution. Validation performed for the full calendar year 2022.

Methodology and Data

• Models used: Semi-empirical polynomial regression model (third-degree polynomial) based on the Ts/VI feature space: SSM = ∑ 3 p=0 ∑ 3 q=0 apq(LSTt)p(FVCt)q
• Data sources:
  • Satellite:
    • SMAP L3 Surface Soil Moisture (SSM) operational product (v09), 36 km spatial resolution, daily (06:00 and 18:00 local time).
    • Sentinel-3 Level-2 Fractional Vegetation Cover (FVC) (OLCI instrument), 300 m spatial resolution, ~1 day temporal resolution.
    • Sentinel-3 Level-2 Land Surface Temperature (LST) (SLSTR instrument), 1 km spatial resolution, ~1 day temporal resolution.
  • Observation (in-situ):
    • REMEDHUS operational network stations (Spain), providing volumetric soil surface moisture (0-5 cm) point observations, hourly measurements.

Main Results

• The RegreSSM tool successfully downscaled SMAP L3 SSM from 36 km to 1 km.
• Validation over the Iberian Peninsula (REMEDHUS network) for 2022 showed:
  • Average bias: 0.01 m³/m³
  • Mean Absolute Difference (MAD): 0.06 m³/m³
  • Root Mean Square Difference (RMSD): 0.07 m³/m³
  • Unbiased Root Mean Square Difference (ubRMSD): 0.04 m³/m³
  • Coefficient of Determination (R²): 0.63 (overall average), with AM overpass R² = 0.68 and PM overpass R² = 0.62.
• The downscaled maps provided improved representation of local soil moisture heterogeneity and effectively captured seasonal variability.
• The results are comparable to other downscaling techniques and fall within the SMAP mission's target accuracy error levels (0.04 m³/m³).

Contributions

• RegreSSM is the first open-source, Python-based software tool that operationalizes a well-established Ts/VI downscaling methodology for SMAP SSM products.
• It provides an integrated, automated workflow with a user-friendly Graphical User Interface (GUI), making high-resolution SSM generation accessible to non-expert users and practitioners.
• The software enables standardized and reproducible downscaling runs across diverse study areas, distributed as citable software with a documented workflow.
• It leverages state-of-the-art Sentinel-3 optical and thermal data for downscaling, addressing a gap where most studies used MODIS/MeteoSat.
• RegreSSM facilitates scalable and transferable downscaling of SMAP products globally with low computational requirements, advancing the operational use of NASA and ESA datasets.

Funding

• LISTEN-EO project, implemented in the framework of H.F.R.I called “Basic research Financing (Horizontal support of all Sciences)” under the National Recovery and Resilience Plan “Greece 2.0” funded by the European Union-Next Generation EU (H.F.R.I. Project Number: 15898).

Citation

@article{Petropoulos2025RegreSSM,
  author = {Petropoulos, George P. and Detsikas, Spyridon E. and Anagnostopoulos, Vasileios and Lekka, Christina},
  title = {RegreSSM: A novel software tool for downscaling the SMAP L3 soil moisture operational product utilizing the Ts/VI feature space and Sentinel-3 data},
  journal = {Environmental Modelling & Software},
  year = {2025},
  doi = {10.1016/j.envsoft.2025.106836},
  url = {https://doi.org/10.1016/j.envsoft.2025.106836}
}