Real-Time Data Driven Wildland Fire Modeling

TL;DR

This paper presents a real-time wildland fire modeling approach combining semi-empirical spread estimates, atmospheric modeling, and data assimilation techniques to improve fire front prediction using thermal imagery.

Contribution

It introduces a novel integration of semi-empirical fire spread models with WRF and advanced data assimilation for real-time fire behavior prediction.

Findings

Effective fire front identification using level set methods

Successful assimilation of thermal image data

Enhanced real-time fire spread estimation

Abstract

We are developing a wildland fire model based on semi-empirical relations that estimate the rate of spread of a surface fire and post-frontal heat release, coupled with WRF, the Weather Research and Forecasting atmospheric model. A level set method identifies the fire front. Data are assimilated using both amplitude and position corrections using a morphing ensemble Kalman filter. We will use thermal images of a fire for observations that will be compared to synthetic image based on the model state.