Optimizing resilient disaster readiness: a network and probabilistic modeling approach to firefighting resource prepositioning

Abstract

Wildfire response effectiveness depends on strategically prepositioning resources to balance accessibility and hazard exposure, as wildfires can escalate rapidly and delayed access to equipment in the early stages and initial attack increases operational risk and firefighter casualties. This study presents a geographic information system (GIS) and multi-modal transportation network-based optimization framework that integrates probabilistic fire spreading potential with multi-modal travel time accessibility to support resilience-informed wildfire resource allocation. Fire risk is estimated using a spatial spreading model based on forest distribution and elevation, and accessibility is calculated over an integrated road–forest network including drivable, walkable, and off-trail routes. Candidate cache locations are selected using a $$p$$ -median facility location formulation under both travel time-only and risk-weighted strategies. Results show that travel time-only optimization can neglect high-risk regions, whereas risk-weighted optimization reduces the expected response time to a randomly occurring early-stage fire from 34 to 31 min, approximately a 10 percent improvement, while also lowering extreme response times from 117 to 97 min. The framework is modular, reproducible, and adaptable to improved data inputs, providing a transparent approach for resilience-oriented wildfire resource prepositioning.

Bibliographic Information