From Black Box to Insight: Explainable AI for Extreme Event Preparedness

TL;DR

This paper explores explainable AI for extreme event preparedness with wildfire forecasting as a case study. It evaluates several predictive models and uses SHAP to reveal feature contributions, temporal dynamics, and potential biases, supported by custom time-aware visualizations. The results show deep learning models, particularly Transformer-based architectures, achieve higher accuracy than tree ensembles, while SHAP explanations align with physical understanding and offer actionable guidance for emergency management. By applying the NASA FAIRUST principles, the work emphasizes trustworthy, accessible, and reusable climate risk forecasting to support mitigation and policy decisions.

Abstract

As climate change accelerates the frequency and severity of extreme events such as wildfires, the need for accurate, explainable, and actionable forecasting becomes increasingly urgent. While artificial intelligence (AI) models have shown promise in predicting such events, their adoption in real-world decision-making remains limited due to their black-box nature, which limits trust, explainability, and operational readiness. This paper investigates the role of explainable AI (XAI) in bridging the gap between predictive accuracy and actionable insight for extreme event forecasting. Using wildfire prediction as a case study, we evaluate various AI models and employ SHapley Additive exPlanations (SHAP) to uncover key features, decision pathways, and potential biases in model behavior. Our analysis demonstrates how XAI not only clarifies model reasoning but also supports critical decision-making by domain experts and response teams. In addition, we provide supporting visualizations that enhance the interpretability of XAI outputs by contextualizing feature importance and temporal patterns in seasonality and geospatial characteristics. This approach enhances the usability of AI explanations for practitioners and policymakers. Our findings highlight the need for AI systems that are not only accurate but also interpretable, accessible, and trustworthy, essential for effective use in disaster preparedness, risk mitigation, and climate resilience planning.

Figures (10)

• Figure 1: Costs of U.S. extreme disaster events over time disasters.
• Figure 2: Explainable AI (XAI) Techniques.
• Figure 3: Representative feature distributions from Mesogeos dataset.
• Figure 4: Representative feature distributions from California Wildfires dataset.
• Figure 5: Visualization for average SHAP values for the Mesogeos dataset using the LSTM model.