Suppression firing is a crucial approach to control the spread of forest fires. However, existing suppression firing mainly relies on rare quantitative analysis by experts, making efficient forest fire control efforts difficult to perform.
In this paper, a fire spread prediction model was implemented to quantitatively simulate and analyze suppression firing. This model adopted the cellular automata algorithm to define the fire spread as a grid dynamics problem. The forest landscape was divided into contiguous regular cells with different cell burning states (S0: unburned; S1: ignited; S2: flashover; S3: extinguishing; S4: extinguished). Then, multimodal environmental factors such as fuel type, slope, wind, and temperature were considered to construct the rate of the spread function and predict the fire spread speed in various complex scenarios. Next, state update rules were proposed to define how the burning state of forest cells was transformed for different fire conditions. The minimum travel time method was then adopted to iteratively calculate the ignition time of each cell in the forest landscape. Therefore, the spatiotemporal evolution of forest fires in complex environmental scenarios was quantitatively modeled. Additionally, a trigger mechanism was proposed to define reverse ignition behavior as a grid cell with specific time-trigger constraints. This mechanism realized a quantitative simulation analysis of human ignition factors with different spatiotemporal conditions.
To verify the reliability and feasibility of our model, a real forest fire that occurred in the Beibei District of Chongqing in August, 2022 was chosen as the study case. Fire data (fuel type, slope, historical weather, fire perimeter, etc.) and firefighting records (the location and time of fire ignition, suppression firing description, etc.) were collected to reconstruct the firing process. Our model was applied to the suppression firing in this forest fire to analyze the fire control effect for different environmental conditions. The experimental results showed that our model was superior in predicting the spatiotemporal spread of forest fire with competitive model performance (Jaccard: 0.732; Sorensen: 0.845). The spatial location and ignition time of the reverse ignition in suppression firing were quantitatively analyzed and visualized, demonstrating how the reverse fire burned the fuel in advance and impeded the spread of free fires.
Quantitatively modeling the suppression firing can provide effective decision-making for wildfire firefighters to formulate accurate fire control strategies and improve the modernization capability of forest fire management. As a highly complex, dangerous firefighting strategy, more research on the combustion mechanism and simulation method of suppression firing is needed, such as the formation mechanism and modeling method of local microclimate in a forest fire landscape, the barrier effect of the isolation zone, and spatial optimization.
