Modelling wildfire spread and spotfire merger using conformal mapping and AAA-least squares methods

TL;DR

This paper introduces two numerical methods based on conformal mapping and AAA-least squares to model wildfire spread and merger, accounting for wind effects and firebreaks, with results matching existing data.

Contribution

It presents novel conformal mapping and AAA-least squares methods for efficient wildfire modeling, incorporating complex factors like pyrogenic winds and firebreaks.

Findings

Methods are fast and accurate for operational wildfire prediction.

Results match well with existing numerical and experimental data.

Effects of wind and firebreaks on wildfire spread are demonstrated.

Abstract

A two-dimensional model of wildfire spread and merger is presented. Three features affect the fire propagation: (i) a constant basic rate of spread term accounting for radiative and convective heat transfer, (ii) the unidirectional, constant ambient wind, and (iii) a fire-induced pyrogenic wind. Two numerical methods are proposed to solve for the harmonic pyrogenic potential. The first utilizes the conformal invariance of the Laplace equation, reducing the wildfire system to a single Polubarinova-Galin type equation. The second method uses a AAA-least squares method to find a rational approximation of the potential. Various wildfire scenarios are presented and the effects of the pyrogenic wind and the radiative/convective basic rate of spread terms investigated. Firebreaks such as roads and lakes are also included and solutions are found to match well with existing numerical and experimental results. The methods proposed in this work are suitably fast and accurate to be considered for operational use.