Modelling Initial Attack Containment Success: a Tool to Support Initial Attack Dispatching Decisions
This project focuses on dispatching, particularly building machine learning models to predict IA success. Previous studies have used historical fire report data for the province of Alberta to model IA containment in relation to a range of variables including IA response time, fire load, fuel conditions (time-since-fire, linear features), and fire behaviour conditions (Arienti et al. 2006, Beverly 2017). A key innovation of this project is developing a model to predict IA success that incorporates information about the type and the amount of dispatched resources, in addition to fuel and weather conditions.