Causal Graph Neural Networks for Wildfire Danger Prediction

TL;DR

This paper introduces a causal graph neural network approach for wildfire prediction, improving accuracy and interpretability by modeling causal relationships among environmental variables, especially in imbalanced datasets.

Contribution

It integrates causality with GNNs to explicitly model causal mechanisms, enhancing wildfire forecasting accuracy and interpretability over traditional deep learning models.

Findings

Superior wildfire prediction performance in European biomes

Enhanced robustness in imbalanced datasets

Improved understanding through SHAP interpretability

Abstract

Wildfire forecasting is notoriously hard due to the complex interplay of different factors such as weather conditions, vegetation types and human activities. Deep learning models show promise in dealing with this complexity by learning directly from data. However, to inform critical decision making, we argue that we need models that are right for the right reasons; that is, the implicit rules learned should be grounded by the underlying processes driving wildfires. In that direction, we propose integrating causality with Graph Neural Networks (GNNs) that explicitly model the causal mechanism among complex variables via graph learning. The causal adjacency matrix considers the synergistic effect among variables and removes the spurious links from highly correlated impacts. Our methodology's effectiveness is demonstrated through superior performance forecasting wildfire patterns in the European boreal and mediterranean biome. The gain is especially prominent in a highly imbalanced dataset, showcasing an enhanced robustness of the model to adapt to regime shifts in functional relationships. Furthermore, SHAP values from our trained model further enhance our understanding of the model's inner workings.