基于全红外光谱的火点识别与林火灾后评估

Fire spot recognition and post-fire assessment based on full infrared spectroscopy

以2018年底美国加州史上死伤最惨重、也最具破坏性的“坎普”林火(Camp Fire)为研究对象,根据近红外、短波红外和热红外光谱段对林火灾害不同生命周期的敏感度,采用归一化燃烧指数NBR、热红外地表温度LST和归一化植被指数NDVI等模型进行灾时高温火点识别及溯源、灾后植被损失和植被恢复模式评估。结果表明,dNBR高于0.1的烧伤区域面积占比达66.53%,其中高强度烧伤区超19%;火灾造成平均植被覆盖度下降11.55%。经LST反演识别的7个典型高温火点受灾状况远高于其他区域,NDVI和FVC最高降幅为0.45和49.4%,分别是全区NDVI和FVC降幅的7.4倍和6.9倍,可见热红外LST反演技术在高温火点精确定位和受灾程度定量判定上的高效和准确程度。灾后植被恢复研究表明,林火对高植被覆盖区破坏较为严重。灾后1 a内植被恢复速度较慢。过火区总体植被恢复情况较差,部分区域出现土壤退化的现象。预计植被完全恢复还需要更长的时间。

The most deadly and destructive‘Camp Fire’in the history of California,USA at the end of year 2018 was taken as the research object to comprehensively analyze the spread of forest fires and their effects on regional vegetation.Several remote sensing retrieval methods such as Normalized Burn Radio(NBR),Land Surface Temperature(LST)and Normalized Difference Vegetation Index(NDVI)were introduced to identify and trace high temperature fire points,and to evaluate the post-disaster vegetation loss and restoration patterns.The results showed that,the burned area with dNBR higher than 0.1 accounted for 66.53%,among which the high-intensity burn rate was over 19%.The average vegetation coverage decreased by 11.55%.The damage degree of 7 typical high temperature fire spots identified by LST retrieval was much higher than that of other regions.The highest decrease rates of NDVI and FVC were 0.45 and 49.4%respectively,7.4 times and 6.9 times of that in the whole region,indicating high efficiency and accuracy of thermal infrared LST retrieval technology in accurate positioning of high temperature fire point and quantitative judgment of disaster degree.Research on post-disaster vegetation restoration showed that high-vegetation-covered areas had been more severely damaged by the fire,and the recovery rate was slow.The overall vegetation restoration was poor,and soil degradation had occurred in the burned areas.It is expected to take a long time for the vegetation to be fully restored.