Towards improving wildland firefighter situational awareness through daily fire behaviour risk assessments in the US Northern Rockies and Northern Great Basin

Wildland firefighters must assess potential fire behaviour in order to develop appropriate strategies and tactics that will safely meet objectives. Fire danger indices integrate surface weather conditions to quantify potential variations in fire spread rates and intensities and therefore should closely relate to observed fire behaviour. These indices could better inform fire management decisions if they were linked directly to observed fire behaviour. Here, we present a simple framework for relating fire danger indices to observed categorical wildland fire behaviour. Ordinal logistic regressions are used to model the probabilities of five distinct fire behaviour categories that are then combined with a safety-based weight function to calculate a Fire Behaviour Risk rating that can plotted over time and spatially mapped. We demonstrate its development and use across three adjacent US National Forests. Finally, we compare predicted fire behaviour risk ratings with observed variations in satellite-measured fire radiative power and we link these models with spatial fire danger maps to demonstrate the utility of this approach for landscape-scale fire behaviour risk assessment. This approach transforms fire weather conditions into simple and actionable fire behaviour risk metrics that wildland firefighters can use to support decisions that meet required objectives and keep people safe.