林地余火隐燃特性具有隐蔽性强、持续时间长、目测难度大、且具有死灰复燃等特点,从而一直困扰森林火灾的彻底扑灭。为了及时高效发现林地余火阴燃点，以无人机搭载高光谱成像系统为工具，首先结合设置样地、地面调查、GIS空间分析、模型构建、层次分析法等建立林地余火识别与定位系统；在此基础上，通过高光谱影像分析死灰复燃阴燃点的燃烧状态与消耗量的变化规律；探讨林地余火阴燃点分布与地形之间的关系；利用高光谱影像温度识别技术，标定不同程度的林地余火死灰温度场，量化死灰复燃的火环境及温度参数阈值，形成死灰复燃点诊断的科学方法；明确不同环境下，各温度林地余火阴燃点引燃概率与引燃时间，并建立初步的林地余火死灰复燃概率模型。本项目研究将推动森林防火技术的发展，研究成果将为安全扑火提供重要的方法和资料。

Wildland fire can not put out completely because of the smoldering fires, which are often hidden, last long, invisible and resurgent easily. To observe and locate the embers quickly, this study will construct an initial model for computing the probability of resurgences. In this study, a recognition and location system for embers will firstly be built by using plots, ground surveys, GIS spatial analysis, models and Analytic Hierarchy Process(AHP) with a Hyperspectral UAV. Under different conditions, the probabilities and igniting times of the smoldering fires vary in different temperatures. On hyperspectral analysis of the combustion phase and rules of resurgences, a scientific diagnostic technique will be proposed by determining the relationship between location and topography to quantize the thresholds of temperature and conditions. This research will promote the development in preventing forest-fire and provide more materials and technologies for safe firefighting.