Extreme-value forest fire prediction A study of the Loss Function in an Ordinality Scheme

TL;DR

This paper introduces an ordinal classification framework for wildfire severity prediction, emphasizing the impact of loss function design on rare event forecasting, and demonstrates that ordinal-aware loss functions improve model performance in operational settings.

Contribution

The study presents the first ordinal classification approach for wildfire severity prediction, comparing various loss functions and showing the effectiveness of ordinal supervision in imbalanced, real-world data.

Findings

Ordinal supervision improves wildfire severity prediction.

Weighted Kappa Loss outperforms other loss functions.

Performance on rarest events remains limited due to data imbalance.

Abstract

Wildfires are highly imbalanced natural hazards in both space and severity, making the prediction of extreme events particularly challenging. In this work, we introduce the first ordinal classification framework for forecasting wildfire severity levels directly aligned with operational decision-making in France. Our study investigates the influence of loss-function design on the ability of neural models to predict rare yet critical high-severity fire occurrences. We compare standard cross-entropy with several ordinal-aware objectives, including the proposed probabilistic TDeGPD loss derived from a truncated discrete exponentiated Generalized Pareto Distribution. Through extensive benchmarking over multiple architectures and real operational data, we show that ordinal supervision substantially improves model performance over conventional approaches. In particular, the Weighted Kappa Loss (WKLoss) achieves the best overall results, with more than +0.1 IoU (Intersection Over Union) gain on the most extreme severity classes while maintaining competitive calibration quality. However, performance remains limited for the rarest events due to their extremely low representation in the dataset. These findings highlight the importance of integrating both severity ordering, data imbalance considerations, and seasonality risk into wildfire forecasting systems. Future work will focus on incorporating seasonal dynamics and uncertainty information into training to further improve the reliability of extreme-event prediction.