森林火灾应急救援情景演化Web三维模拟——以四川省西昌市泸山“3.30”森林火灾为例

Web-based 3D simulation of the emergency rescue scenario evolution for forest fires-A case study of the“3.30”forest fire in Lushan,Xichang City,Sichuan Province

目前大多数森林火灾信息系统局限于对林火要素的单一模拟,对森林火灾应急响应情景演化过程进行综合性表达的系统较少。本文构建基于知识元和元胞自动机的多要素、多事件情景演化模型,采用Cesium创建森林火灾应急救援情景演化模拟的三维地理信息系统;通过四川省泸山“3.30”森林火灾的实际应用案例验证了可行性和有效性。结果表明:系统模型能够较准确地模拟林火扩散范围;系统模拟1 h林火扩散并可视化所需时间约为10s,手动修正火灾扩散范围更新速度快于5s/次;系统可以清晰地模拟森林火灾中火焰扩散、地面扑火、隔离带开辟、飞机扑火等过程中各要素的动态变化。研究成果可为森林火灾灭火救援工作提供决策支持,有助于提高森林火灾应急救援能力,减少火灾对环境和经济造成的损失。

As climate change intensifies and extreme heat events become more frequent,forest fires are escalating globally.Yet,most existing systems are limited to singular simulations of forest fire propagation and few offer comprehensive representations of the scenario evolution during forest fire emergency responses.This study introduces a multi-element,multi-event scenario evolution timeline model.Utilizing this model,a three-dimensional Geographic Information System(3D GIS)for simulating forest fire scenario evolution was developed using the open-source Web 3D visualization engine,Cesium.Symbols representing four key elements of forest fires-disaster factors,disaster bearing bodies,disaster resistant bodies,and disaster bearing environments-were designed.For each element,large-scale and small-scale visual symbols were created to accommodate varying observation scales and perspectives in the three-dimensional scene.Large-scale symbols include highly realistic three-dimensional models such as humans,aircraft,and houses.Small-scale symbols are simple,intuitive two-dimensional dot,line,or icon markers.Cesium’s particle system was employed to simulate fire,smoke,and various weather conditions like rain and snow.An enhanced 3D cellular automata model,integrating digital elevation model(DEM)elevation data,meteorological factors,and vegetation attributes,was used to mimic forest fire spread.The system’s front end was built using the Vue3 framework,while the back end leveraged the Spring Boot 3 framework.The GDAL library was utilized for raster and vector spatial data computations and processing,with PostgreSQL managing the database infrastructure.The devised system adeptly replicates processes such as flame dispersion,firefighting team interventions,firebreak establishment,and aerial fire suppression.Its practicality and efficacy were substantiated through the real-world case of the“3.30”forest fire incident in Lushan,Sichuan.While the model generally accurately simulates forest fire spread,certain areas exhibited discrepancies,which could be manually rectified in near real-time.Testing indicated that the manual correction update rate for the fire spread range was under 5 seconds per iteration,satisfying the demands of real-time fire rescue command operations.This system will provide decision support for forest fire combating and mitigating,help to improve forest fire emergency response capabilities,and reduce environmental and economic losses due to fire.