Just Accepted
This article has been peer reviewed and accepted for publication. It is in production and has not been edited, so may differ from the final published form.
Simulating wildfire spread based on continuous time series remote sensing images and cellular automata
Abstract
Background. The acquisition of behavior parameters in wildfire propagation and the development of firefighting strategies necessitate a precise and efficient simulation method, where the accuracy of fireline coordinates and rate of spread (ROS) are two crucial parameters closely associated with the precision of the simulation. Aims. The present study proposes an adaptive simulation method of wildfire propagation by integrating the continuous time series remote sensing images with the Cellular Automata (CA) model. Methods. The ROS in each direction is calculated using continuous time fireline coordinates derived from multi-source remote sensing images. A time-adaptive propagation algorithm is developed based on the CA model (Time-adaptive Cellular Automata, TCA). The distribution information of vegetation is derived to establish the simulation system of wildfire propagation for verification experiments. Key results. The developed TCA model demonstrates satisfactory simulation performance, as evidenced by a prediction accuracy of 92.2% for the burned area and 87.3% for fireline length within local regions in the MuLi Forest Fire on 28 March 2020. The model effectively characterizes the gradual spread simulation based on low ROS. Conclusion. The adaptive simulation method enables the acquisition of relatively precise results. The TCA model demonstrates its capability to achieve localized and gradual propagation. The created simulation software serves as a powerful tool for wildland surface fire prevention and control. Implications. The adaptive simulation method based on continuous time remote sensing images and the CA model are essential for accurately predicting wildfire propagation.
WF24130 Accepted 13 December 2024
© CSIRO 2024
