A New Simulation-Optimization Model for Wildland Fire Resource Pre-Positioning
Abstract
Every day, using detailed weather forecasts, paired with reports on the moisture content of soil and vegetation, the Los Angeles County Fire Department (LACoFD) must decide where to pre-position firefighting equipment and personnel for the following day. For years, LACoFD has relied on their own expert judgment to make these costly decisions. In 2019, NPS student Zachary Scholz developed the Augmentation Optimization Model (AOM), a mathematically based decision tool to guide resource pre-positioning. Unfortunately, AOM relies on weak estimations of expected burned acreage, complicating result interpretation. We address this problem by developing a simulation to estimate initial attack area containment as a function of pre-positioned resources. These estimates inform the new AOMs objective, producing improved, realistic, and interpretable results. In addition, we have followed LACoFD feedback to incorporate accessibility and steepness of terrain, hand-crew resources, and solution evaluation. We also standardize assembled resources as mixes of engines and exchangeable personnel and reformulate the model so it generates and solves faster. Through an upgraded user interface, LACoFD is using the new AOM daily and analyzing alternatives of protection and cost. The results improve those of legacy AOM and LACoFDs manual solutions on the critical days tested. Moreover, we demonstrate that protection can benefit from augmentation policies not solely based on burning index.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 2020
- Accession Number
- AD1127075
Entities
People
- Rachel A Seeberger
Organizations
- Naval Postgraduate School