Search papers, labs, and topics across Lattice.
This study introduces WILDFIREIA, a national-scale benchmark for predicting initial attack (IA) failure in wildfire suppression using publicly available environmental data. By aligning over 38,000 wildfire events with various environmental and contextual datasets, the research evaluates 16 models to assess their predictive capabilities while preventing data leakage. The findings reveal that while public data can inform IA failure predictions, it remains incomplete, with XGBoost achieving the highest AUPRC of 53.3% and highlighting fuel as a critical predictor in the absence of dynamic observations.
Publicly available data can significantly inform wildfire initial attack failure predictions, but it only captures part of the picture.
Initial attack (IA) is the first wildfire suppression phase, when agencies must quickly decide which fires may escape early control. Existing IA failure prediction studies often use non-public response records or regional settings, so it remains unclear how well public data available at fire discovery time can support IA failure prediction at national scale. We present WILDFIREIA, the first U.S. national-scale benchmark for IA failure prediction from environmental and contextual data available at fire discovery time. WILDFIREIA aligns 38,128 naturally caused FPA-FOD wildfire events with FIRMS/VIIRS thermal detections, gridMET weather and fire-danger variables, LANDFIRE vegetation, fuel, and topography, OpenStreetMap access features, and WorldPop population density. To prevent data leakage, the benchmark fixes the event unit, size-based label rule, chronological split, metrics, and forbidden-feature list, and excludes final fire size, containment timestamps, and post-discovery satellite detections from model inputs. We evaluate 16 representative models across tabular, temporal, spatial, and spatiotemporal families under the same protocol. Results show that public discovery-time data provides useful but incomplete signal for IA failure prediction: XGBoost achieves the best AUPRC of 53.3%; FIRMS/VIIRS is the least redundant source; and fuel is the strongest static predictor when dynamic observations are unavailable. We release preprocessing outputs and model-ready caches to support reproducible research on early wildfire risk assessment: https://github.com/LabRAI/WildfireIA#.
