Machine Learning for Modeling Wildfire Susceptibility at the State Level: An Example from Arkansas, USA

Fire susceptibility modeling is crucial for sustaining and managing forests among many other valuable land resources. With 56% of its area covered by forests, Arkansas known as the “natural state”. About 1000 wildfires occurred burned more than 10,000 acres each year during 1981–2018. In this paper, we use remote-sensing-based machine learning methods to address natural anthropogenic factors influencing model fire in Arkansas. Among 15 explored variables, potential evapotranspiration, soil moisture, Palmer drought severity index, dry season precipitation were recognized most significant contributing density. The obtained R-squared values are significant, with 0.99 training 0.92 validation. results show that Ouachita National Forest Ozark Forest, west-central west Arkansas, respectively, have highest wildfires. southern part has low-to-moderate susceptibility, while eastern state lowest susceptibility. These new demonstrate potency random forest predicting at level can be adapted study fires states help preparedness reduce loss save precious environment.