RxFire Optimization Engine - Eglin AFB case study

Description

The data and code accompanies the following article: R. Majumder, A. J. Terando, J. K. Hiers, J. A. Collazo, B. J. Reich (2025). A spatiotemporal optimization engine for prescribed burning in the Southeast US. The article develops a decision tool (the RxFire Engine) to augment fire manager expertise in prescribed burn decision making in the southeast US. A case study is presented based on the Eglin Air Force Base in Florida, USA. This dataset and code aims to replicate the results presented in the article. However, it can be used to obtain future prescribed burn allocation portfolios if the engine is provided with updated weather forecast information.

Steps to reproduce

1. Install R and the packages as required in each R script. All R scripts are provided at the top level of the directory structure. 2. JAGS and rjags used for Bayesian inference. Installation guide can be found here - https://mcmc-jags.sourceforge.io/ 3. R scripts with filenames starting with 00 are for preparing the fire data. The output is already made available in the EglinBurns.RData file. 4. R script with filenames starting with 01-09 do Bayesian hierarchical modeling on meteorological data (the fail state estimation in the article). The output is already made available in the JAGS_4var and JAGS_POP2 folders. 5. R script with filename starting with 10 corresponds to the RxFire engine described in the article, and contains the allocation algorithm which uses historical fire information and Bayesian models 6. R scripts whose filenames which do not start with a number contain utility functions used by other R scripts, and do not need to be run on their own.

Institutions

Categories

Funding

Licence