Set Prediction for Next-Day Active Fire Forecasting
2026-05-11 • Machine Learning
Machine Learning
AI summaryⓘ
The authors developed a new model called WISP that predicts exactly where wildfires will start the next day, instead of just giving broad fire danger scores over large areas. WISP uses data from the past two days, like weather and vegetation info, to predict precise fire locations on a detailed grid worldwide. It ranks potential fire spots and matches predictions to real fires during training. Their tests showed that WISP can detect over half of the important fire clusters fairly accurately, making it a promising method for detailed wildfire forecasting.
wildfire forecastingmachine learningpoint-set predictionHungarian matchingfire radiative powersatellite vegetation productsmeteorologyfire cluster detectionclassification-localization loss
Authors
Yuchen Bai, Georgios Athanasiou, Xin Yu, Diogenis Antonopoulos, Ioannis Papoutsis, Stijn Hantson, Nuno Carvalhais
Abstract
Accurate next-day active fire forecasts can support early warning, disaster response, forest risk assessment, and downstream estimation of fire-related carbon emissions. Existing machine learning approaches to wildfire forecasting typically predict wildfire danger or fire probability on kilometre-scale daily grids, which is useful for regional warning but does not directly represent localized fire events. We propose Wildfire Ignition Set Predictor (WISP), a query-based model that reformulates next-day active fire forecasting as point-set prediction. From 48 hours of covariates including meteorology, satellite vegetation products, static land, and fire history, WISP predicts a fixed-size ranked set of future active fire cluster centres on a 375 m grid across globally distributed regions. The model is trained end-to-end with Hungarian matching; to address the conflicting roles of the classification score in assignment, ranking, and query activation, we use asymmetric classification-localization weighting in matching and loss. We further construct a globally distributed, hourly, multi-source benchmark for this task. On a held-out test set spanning fire regions worldwide, the best WISP variant achieves 38.2% average precision (AP) for ranked fire-centre detections, covers 53.4% of fire cluster mass weighted by fire radiative power (FRP), and localizes 54.1% of observed clusters within 5 km. These results establish sparse set prediction as a viable formulation for high-resolution wildfire forecasting and provide a benchmark for future work in this regime.