Rapid Wildfire Hotspot Detection Using Self-Supervised Learning on Temporal Remote Sensing Data

TL;DR

This paper introduces a new dataset and a self-supervised learning model for rapid wildfire hotspot detection using multi-temporal satellite data, achieving promising results for near real-time monitoring.

Contribution

It presents a novel dataset of European fire-related remote sensing time series and a self-supervised learning approach for identifying wildfire hotspots.

Findings

Achieved an F1 score of 63.58 on the Thraws dataset.

Demonstrated effectiveness of SSL in analyzing multi-temporal remote sensing data.

Enabled near real-time wildfire hotspot detection.

Abstract

Rapid detection and well-timed intervention are essential to mitigate the impacts of wildfires. Leveraging remote sensed data from satellite networks and advanced AI models to automatically detect hotspots (i.e., thermal anomalies caused by active fires) is an effective way to build wildfire monitoring systems. In this work, we propose a novel dataset containing time series of remotely sensed data related to European fire events and a Self-Supervised Learning (SSL)-based model able to analyse multi-temporal data and identify hotspots in potentially near real time. We train and evaluate the performance of our model using our dataset and Thraws, a dataset of thermal anomalies including several fire events, obtaining an F1 score of 63.58.