首页 | 本学科首页   官方微博 | 高级检索  
     

多源遥感影像林火信息检测和植被恢复动态监测
引用本文:张兆鹏, 周晓敏, 闫敏, 等. 多源遥感影像林火信息检测和植被恢复动态监测[J]. 西南林业大学学报(自然科学), 2022, 42(5): 145–157.doi:10.11929/j.swfu.202105046
作者姓名:张兆鹏  周晓敏  闫敏  田昕
作者单位:1. 自然资源部第一大地测量队,陕西 西安 710054;2. 中国科学院空天信息创新研究院,北京 100091;3. 中国林业科学研究院资源信息研究所,北京 100091
摘    要:以Landsat 5 TM和GF–1 WFV影像为主要数据源,利用归一化植被指数和像元二分模型,结合0.2 m分辨率的航空CCD数据和全球30 m精细地表覆盖产品,建立植被覆盖度的估算模型,利用火烧前后NDVI差值提取了2001—2020年间大兴安岭地区的火灾信息;通过FVC定量分析了林区整体覆盖度变化和影响因素,并利用热点区域植被覆盖度变化对不同林火后植被恢复力进行了分析和研究。结果表明:以NDVI值为参数,运用基于像元二分模型的植被覆盖度遥感估算方法技术路线简单、可操作性强,适用于不同分辨率的遥感数据;dNDVI可以对过火区信息进行有效分离;研究区整体FVC较高,但频发的火灾是影响研究区FVC分布状况及其动态变化的主要因素;不同程度过火区的FVC恢复效果不同,小范围过火区的FVC可以恢复到原始状态,中等范围过火区的FVC经过长时间的自然恢复与原始FVC仍有一定差距,而对于大范围,尤其是特大范围的过火区经过长时间恢复虽然FVC会增加,但和原始FVC相比相差甚远。因此,火灾干扰对森林生态系统破坏性较强,受火灾干扰后森林恢复时间较长且其林分组成与物种难以恢复至原型。

关 键 词:林火   归 一化植被指数   像元二分模型   植被覆盖度   火烧迹地   植被恢复
收稿时间:2021-05-20

Forest Fire Information Detection and Vegetation Restoration Dynamic Monitoring Based on Multi-source Remote Sensing Images
Zhang Zhaopeng, Zhou Xiaomin, Yan Min and Tian Xin. Forest Fire Information Detection and Vegetation Restoration Dynamic Monitoring Based on Multi-source Remote Sensing Images[J]. Journal of Southwest Forestry University, 2022, 42(5): 145-157.doi:10.11929/j.swfu.202105046
Authors:Zhang Zhaopeng  Zhou Xiaomin  Yan Min  Tian Xin
Affiliation:1. The First Geodetic Surveying Brigade of MNR, Xi'an Shaanxi 710054, China;2. Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100091, China;3. Institute of Forest Resource Information Techniques, Chinese Academy of Forestry, Beijing 100091, China
Abstract:The novel fractional vegetation coverage model was developed from normalized difference vegetation index and the dimidiate pixel model based on Landsat 5 TM and GF-1 WFV images, the aerial CCD data of 0.2 m resolution and the global land cover map at 30 m. The fire information of Daxing'anling area from 2001 to 2020 was calculated using the difference of NDVI before and after fire. The change of forest coverage and its influencing factors were quantitatively analyzed by FVC, and the vegetation coverage of hot spot area was analyzed through the change of FVC. Vegetation resilience after different forest fires was analyzed in this study, and the results demonstrated: the vegetation coverage estimation method based on dimidiate pixel model is simple yet accurate, which is suitable for remote sensing data of different resolutions; dNDVI could differentiate the information of fire area effectively. FVC of the whole study area is high, frequent fire is the main factor affecting the distribution and dynamic change of FVC in the study area. FVC restoration effect varies different degrees of fire area. The FVC in small fire area can be restored to the original state, the FVC in medium fire area still has a certain gap with the original FVC after a long time of natural restoration; while FVC will increase in large fire area, especially in large fire area after a long time of restoration, but it is far from the original FVC. The research shows that it is difficult to restore the forest after the fire disturbance, which has destructive effects on forest ecosystem.
Keywords:forest fire  NDVI  dimidiate pixel model  FVC  burned area  vegetation restoration
点击此处可从《西南林业大学学报》浏览原始摘要信息
点击此处可从《西南林业大学学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号