PROPAGATOR: An Operational Cellular-Automata Based Wildfire Simulator

Andrea Trucchia,Andrea Gollini,Francesco Baghino,Massimiliano Severino,Mirko D’Andrea,Luca Ferraris,Dario Negro,Paolo Fiorucci

doi:10.3390/fire3030026

Abstract

PROPAGATOR is a stochastic cellular automaton model for forest fire spread simulation, conceived as a rapid method for fire risk assessment. The model uses high-resolution information such as topography and vegetation cover considering different types of vegetation. Input parameters are wind speed and direction and the ignition point. Dead fine fuel moisture content and firebreaks—fire fighting strategies can also be considered. The fire spread probability depends on vegetation type, slope, wind direction and speed, and fuel moisture content. The fire-propagation speed is determined through the adoption of a Rate of Spread model. PROPAGATOR simulates independent realizations of one stochastic fire propagation process, and at each time-step gives as output a map representing the probability of each cell of the domain to be affected by the fire. These probabilities are obtained computing the relative frequency of ignition of each cell. The model capabilities are assessed by reproducing a set of past Mediterranean fires occurred in different countries (Italy and Spain), using when available the real fire fighting patterns. PROPAGATOR simulated such scenarios with affordable computational resources and with short CPU-times. The outputs show a good agreement with the real burned areas, demonstrating that the PROPAGATOR can be useful for supporting decisions in Civil Protection and fire management activities.

Highlights

An operational wildfire simulator designed to respond to a broad scenario of wildfires, via an ad-hoc implementation of Cellular Automatas (CA) and Rate of Spread, is described
The model is applied to a set of Spanish and Italian case studies, which range from fires that were controlled by a high level of human intervention to wild-land fires that lacked such level of fire fighting
The proposed test cases represented ideal scenarios to test a CA model designed for wind-driven wildfires

Summary

Introduction

Mediterranean countries are prone to wildfires, which represent a significant menace to environment, properties, and human lives. Protection Authorities (CPAs) the main information to cope with direct impacts on exposed people. The tragic wildfires that occurred in Greece and in Portugal in the last few years, which caused many fatalities [1], and the more recent event that occurred last summer in Gran Canaria, where 9000 people were evacuated [2], constitute examples of the consequences of such shortcomings. Most of the wild-land fires in the Mediterranean are human caused; natural ignitions caused by lightning are not negligible and could be increased by climate change [3].

Objectives

Results

Discussion

Conclusion