Construction and Evaluation of Machine Learning Models for Near-Real-Time Fire Detection from MTG FCI Imagery

2026-10-04

AI Summary

This study addresses the limited accuracy of traditional threshold-based algorithms and their tendency to miss small fires in Meteosat Third Generation (MTG) satellite fire detection. To overcome these limitations, we propose a machine learning-based near-real-time fire detection model utilizing MTG Flexible Combined Imager (FCI) imagery. The model is trained using VIIRS reference data and validated across diverse ecoregions, including Europe and Africa. Results demonstrate that the 1 km resolution model significantly outperforms both its 2 km variant and the operational baseline, achieving an F1-score improvement of up to 0.36. Furthermore, it advances fire detection time by 260 minutes compared to the baseline, substantially enhancing both the detection rate of small fires and overall timeliness. The proposed model has been released as open source to facilitate broader adoption and further research.

Abstract

Geostationary satellite observations are important for wildfire detection and monitoring. The current study evaluates machine learning models for MTG FCI near-real-time fire detection in 1- and 2-km spatial configurations and compares them with threshold-based algorithms. The models were trained and evaluated using VIIRS fire reference data across diverse ecological regions in Europe, Africa, and the Middle East. The key results are that 1-km models significantly outperform both their 2-km variants and operational threshold products. The constructed 1-km models achieved F1 scores higher by up to 0.36 compared to baseline products. Importantly, the 1-km models detected small fires with higher probability compared to competing models. Finally, the models robustly detected fires up to 260 min earlier than baseline products. To support opensource applications, our trained models are publicly available.

Problem

• wildfire detection
• near-real-time
• geostationary satellite
• MTG FCI
• machine learning

Innovation

• Machine Learning
• Near-Real-Time Fire Detection
• MTG FCI
• Geostationary Satellite
• Small Fire Detection