The Influence of Information Timeliness on the Decision of Forest Fire Fighting Aircraft Scheduling

Firefighting aircraft are typically deployed during the initial phases of forest fires to reduce the risk of fire propagation. This study explores the scheduling and decision-making issues of aerial firefighting through a comprehensive analysis of multiple wildfire datasets from Northeast and Southwest. Accounting for temporal variations in information availability, we employ an obstacle model methodology to predict both the probability of deployment and the number of aircraft allocated. Two distinct models were developed: the "Fire Reporting Model",which utilizes initial fire detection data, and the "Initial Response Model",which incorporates information from the start of initial suppression efforts. Both models reveal critical factors influencing dispatch decisions, including fire severity index, fuel volatility, observed spread rates, initial fire area size, and ignition causes. Meanwhile, evaluation of predictive capabilities on validation datasets indicated that the "Initial Response Model" outperformed the "Fire Reporting Model". While both models perform well in predicting deployments of two or fewer aircraft during initial suppression phases, predictions beyond this threshold tend to underestimate firefighting effectiveness. These findings suggests a pattern of risk-averse decision-making in fire resource allocation strategies.