Identifying and Evaluating Fire Severity: A Case Study of Wulin Fire

The main purpose of the present study is to develop a methodology to classify fire severity in a very short period following a fire. The methodology developed here is to correlate the relationship between ground data and NDVI from the SPOT satellite image. We surveyed the burned site of Wulin fire in 2002 and set up 20 plots along the different altitudes. Fire severity was identified by stem charred, crown consumed, litter remained, scorch height, and diameter of trees. The results showed that fire severity consists of 3 classes. All of them have high correlation with NDVI. Among the 3 classes, the severity of heavy class relates to NDVI < 0. The moderate class is identified by NDVI from 0~0.103, and the light class is those NDVI >0.103. Using the above criteria, only 7.84 ha was estimated as heavy class. One third of total burned area (106.73 ha) was classified as moderate. Most of the burned areas (236.35 ha) were just light influenced.本研究的核心目标为开发一套可在火灾后极短周期内实现火灾烈度分类的方法，该方法旨在关联地面实测数据与SPOT卫星影像的归一化植被指数（Normalized Difference Vegetation Index, NDVI）。研究团队于2002年针对武林火灾的过火区域开展野外调查，沿不同海拔梯度布设了20个样地。火灾烈度通过树干炭化程度、树冠烧毁情况、枯落物残留量、灼烧高度以及树木径级进行判定。结果显示，火灾烈度可划分为3个等级，且各等级均与归一化植被指数存在较高相关性。其中，重度烈度对应的归一化植被指数值小于0；中度烈度的归一化植被指数取值范围为0~0.103；轻度烈度则对应归一化植被指数值大于0.103。基于上述判定标准，研究估算得到重度烈度过火区域面积仅为7.84公顷，总过火面积的三分之一（106.73公顷）被划分为中度烈度区域，而绝大多数过火区域（236.35公顷）仅受轻度火灾影响。