Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Comparative Assessment of Multimodal Earth Observation Data for Soil Moisture Estimation

Ioannis Kontogiorgakis, Athanasios Askitopoulos, Iason Tsardanidis, Dimitrios Bormpoudakis, Ilias Tsoumas, Fotios Balampanis, Charalampos Kontoes

TL;DR

A high-resolution SM estimation framework for vegetated areas across Europe, combining Sentinel-1 SAR, Sentinel-2 optical imagery and ERA-5 reanalysis data through machine learning, suggests that domain-specific spectral indices combined with tree-based ensemble methods offer a practical and computationally efficient solution for operational pan-European field-scale soil moisture monitoring.

Abstract

Accurate soil moisture (SM) estimation is critical for precision agriculture, water resources management and climate monitoring. Yet, existing satellite SM products are too coarse (>1km) for farm-level applications. We present a high-resolution (10m) SM estimation framework for vegetated areas across Europe, combining Sentinel-1 SAR, Sentinel-2 optical imagery and ERA-5 reanalysis data through machine learning. Using 113 International Soil Moisture Network (ISMN) stations spanning diverse vegetated areas, we compare modality combinations with temporal parameterizations, using spatial cross-validation, to ensure geographic generalization. We also evaluate whether foundation model embeddings from IBM-NASA's Prithvi model improve upon traditional hand-crafted spectral features. Results demonstrate that hybrid temporal matching - Sentinel-2 current-day acquisitions with Sentinel-1 descending orbit - achieves R^2=0.514, with 10-day ERA5 lookback window improving performance to R^2=0.518. Foundation model (Prithvi) embeddings provide negligible improvement over hand-crafted features (R^2=0.515 vs. 0.514), indicating traditional feature engineering remains highly competitive for sparse-data regression tasks. Our findings suggest that domain-specific spectral indices combined with tree-based ensemble methods offer a practical and computationally efficient solution for operational pan-European field-scale soil moisture monitoring.

Comparative Assessment of Multimodal Earth Observation Data for Soil Moisture Estimation

TL;DR

A high-resolution SM estimation framework for vegetated areas across Europe, combining Sentinel-1 SAR, Sentinel-2 optical imagery and ERA-5 reanalysis data through machine learning, suggests that domain-specific spectral indices combined with tree-based ensemble methods offer a practical and computationally efficient solution for operational pan-European field-scale soil moisture monitoring.

Abstract

Accurate soil moisture (SM) estimation is critical for precision agriculture, water resources management and climate monitoring. Yet, existing satellite SM products are too coarse (>1km) for farm-level applications. We present a high-resolution (10m) SM estimation framework for vegetated areas across Europe, combining Sentinel-1 SAR, Sentinel-2 optical imagery and ERA-5 reanalysis data through machine learning. Using 113 International Soil Moisture Network (ISMN) stations spanning diverse vegetated areas, we compare modality combinations with temporal parameterizations, using spatial cross-validation, to ensure geographic generalization. We also evaluate whether foundation model embeddings from IBM-NASA's Prithvi model improve upon traditional hand-crafted spectral features. Results demonstrate that hybrid temporal matching - Sentinel-2 current-day acquisitions with Sentinel-1 descending orbit - achieves R^2=0.514, with 10-day ERA5 lookback window improving performance to R^2=0.518. Foundation model (Prithvi) embeddings provide negligible improvement over hand-crafted features (R^2=0.515 vs. 0.514), indicating traditional feature engineering remains highly competitive for sparse-data regression tasks. Our findings suggest that domain-specific spectral indices combined with tree-based ensemble methods offer a practical and computationally efficient solution for operational pan-European field-scale soil moisture monitoring.
Paper Structure (14 sections, 1 equation, 3 figures, 2 tables)

This paper contains 14 sections, 1 equation, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Data Collection
  3. Ground Truth Data
  4. Satellite Data
  5. Reanalysis
  6. Methods
  7. Preprocessing & Feature Engineering
  8. Model Training and Experiments
  9. Results & Discussion
  10. E1: Modality and Temporal Matching Evaluation
  11. E2: ERA5 temporal window optimization
  12. E3: Prithvi embeddings vs. hand-crafted features
  13. Conclusion
  14. Acknowledgements

Figures (3)

  • Figure 1: Distribution of ISMN stations across Europe used in this study after filtering, colored by network.
  • Figure 2: Experiments methodology diagram.
  • Figure 3: Model performance as a function of ERA5 temporal lag.