基于多源数据的草原可燃物特征提取及易燃性评价研究

Supported by Changling grassland experimental station of northeast normal university, study sites were selected in grassland region in the western Jilin Province. Aim at the weaknesses of grassland fire disaster risk assessment, how fuel physicochemical properties under different conditions affect on grassland fire occurring and development was studied by grassland fire disaster risk formation mechanism. Started with indoor experiments on different indexes of grassland fuel physicochemical properties, fuel characteristic retrieval models were established based on multi-source remote sensing data and rapid extraction of the key features parameters of grassland fuel was realized. And thus, by screening indexes on that basis, the grassland fuel flammability assessment models were established, and region grading division of fuel flammability were done to achieve fuel flammability assessment from single species composition to community, from community to a type of grassland and also from a type of grassland to total study area. Because of grassland fuel flammability assessment taken as the precondition of grassland fire disaster risk assessment, this study can make the results of risk assessment more scientific, reliable and maneuverable. It is significant to grassland fire disaster monitoring under large scale and instructional to grassland fire suppression.

本项目以吉林省西部草原为研究区，以东北师范大学长岭草原实验站为依托，针对草原火灾风险评价的薄弱环节，从草原火灾风险形成机理角度出发，研究不同条件和环境下可燃物理化性质影响草原火的发生发展的机理，从草原可燃物的理化性质的多种指标的室内实验着手，利用多源遥感数据建立可燃物特征反演模型，实现草原可燃物关键特征参数的快速提取。在此基础上筛选指标，构建草原可燃物易燃性评价模型，并对研究区的可燃物易燃性等级区划，实现从单种类到一个群落、一个群落到某个草地类型，从草地类型到整个研究区的可燃物易燃性评价。草原可燃物易燃性评价是草原火灾风险评价的前提和基础，通过本项目的研究，可使草原火灾风 险评价的结果更科学、可靠、更具有可操作性，对于大范围的草原火灾风险监测具有重要意义，同时对草原火灾的扑救工作具有指导意义。

本项目以吉林省西部草原为研究区，以东北师范大学长岭草地实验站为依托，从草原火灾风险形成机理角度出发，通过草原可燃物的理化性质的多种指标室内实验提取着手，并结合室内可控实验和野外点火实验与遥感技术，建立基于多源数据的草原可燃物特征反演模型。实现包括草原可燃物的承载量、覆盖率、含水率、连续度、热值、燃点、灰分含量等指标的遥感反演模型。实现草原可燃物关键特征参数的快速提取。同时，研究将综合运用室内外实验、数理方法、3S技术，实现多源信息融合，进行客观的草原火灾风险评价，将风险评价结果应用在草原火灾风险预警及防火资源布局优化与救助等应急管理的关键问题上，构建草原火灾应急管理辅助决策系统，实现草原火灾风险评价、预警和应急对应管理的自动化和智能化。本研究将弥补我国草原火灾火行为指标遥感反演技术研究的不足。并引入新方法解决传统遥感技术无法很好解决的草原火问题。可以为日后的相关研究奠定一定的理论基础和相关技术路线。