# Monitoring digestate application on agricultural crops using Sentinel-2 Satellite imagery

**Authors:** Andreas Kalogeras, Dimitrios Bormpoudakis, Iason Tsardanidis, Dimitra A. Loka, Charalampos Kontoes  
**arXiv:** [2504.19996](https://arxiv.org/abs/2504.19996)  
**Date:** 2025-04-28  
**Summary page:** https://www.summarizepaper.com/en/paper/2504.19996v2/

## Key points

- Study focuses on monitoring digestate application on agricultural crops using Sentinel-2 satellite imagery
- Evaluates effectiveness of optical Sentinel-2 satellite imagery in detecting digestate application and its impact on soil and crop health
- Utilizes Sentinel-2 satellite image time series (SITS) analysis to examine spectral indices (EOMI, NDVI, EVI) after digestate application on soils growing four different crop types in Thessaly, Greece
- Machine Learning models (Random Forest, k-NN, Gradient Boosting, Feed-Forward Neural Network) used to detect presence of digestate with F1-scores up to 0.85
- Emphasizes potential of combining remote sensing technology with ML techniques for scalable and cost-effective monitoring of EOM applications in agriculture
- Supports precision agriculture practices and contributes to sustainability efforts by providing insights into effects of organic matter application on soil and crop health
- Results accepted for presentation at the 2025 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2025), highlighting significance of research in advancing agricultural monitoring methods for improved environmental outcomes

## Summary

The study conducted by Andreas Kalogeras, Dimitrios Bormpoudakis, Iason Tsardanidis, Dimitra A. Loka, and Charalampos Kontoes focuses on monitoring the application of digestate on agricultural crops using Sentinel-2 satellite imagery. The research evaluates the effectiveness of optical Sentinel-2 satellite imagery in detecting digestate application and its impact on soil and crop health. The study utilizes Sentinel-2 satellite image time series (SITS) analysis to examine specific spectral indices (EOMI, NDVI, EVI) after digestate application on soils growing four different crop types in Thessaly, Greece. Additionally, Machine Learning (ML) models including Random Forest, k-NN, Gradient Boosting, and a Feed-Forward Neural Network are employed to detect the presence of digestate with F1-scores reaching up to 0.85. The findings emphasize the potential of combining remote sensing technology with ML techniques for scalable and cost-effective monitoring of EOM applications in agriculture. This approach supports precision agriculture practices and contributes to sustainability efforts by providing valuable insights into the effects of organic matter application on soil and crop health. The study's results have been accepted for presentation at the 2025 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2025), showcasing the significance of this research in advancing agricultural monitoring methods for improved environmental outcomes.