摘要:随着遥感技术的不断发展，中高分辨率影像逐渐在植被监测中发挥着重要的作用。为了明确高分辨率传感器在不同生态系统植被提取中的必要性，以内蒙古鄂伦春自治旗为研究对象，设置城市区域和森林区域两个应用靶区，以GF1-WFV和Landsat8-OLI 两种传感器同期影像为对比数据集，探究不同分辨率下遥感植被信息提取差异。结果表明：① Landsat 8对比GF-1在城市区域和森林区域的植被指数高估、低估状态相反，城市区域GF-1的NDVI（Normalized Difference Vegetation Index，NDVI）和SAVI（Soil-Adjusted Vegetation Index，SAVI）均值比Landsat 8低5.69%和1.41%，在森林区域则高出0.77%和5.86%；② 高分辨率影像避免了城市中绿化植被（GF-1植被占比71.30%和71.31%，Landsat 8为58.28%和58.30%）和森林中裸地、道路（GF-1植被占比94.97%和94.92%，Landsat 8为95.00% 和94.99%）被漏提。③在分级面积上，Landsat 8相比GF-1数据在城市区域存在低覆盖度等级的6.67%和 6.77%低估，在森林区域出现高覆盖度等级的12.11%和12.47%高估。研究结果反映了低绿化建成区和高密度林区更加需要使用高分影像作为植被监测工具。



Abstract:With the development of remote sensing technology, medium and high resolution images are playing an important role in vegetation monitoring. In order to define the advantages of high resolution sensors in extracting vegetation information of different ecosystems, this research select urban and forest as target area in Olunchun autonomous banner, Inner Mongolia. Two kinds of sensor images, GF1-WFV and Landsat8-OLI, were used as comparison dataset to explore the difference of vegetation extracting using different spatial resolution in these two ecosystems. The results showed that: (1) compared with GF-1, Landsat 8 showed opposite overestimation and underestimation of its vegetation index in urban areas and forest areas, while the Normalized Difference Vegetation Index (NDVI) and Soil-adjusted Vegetation (SAVI) of GF-1 in urban areas were 5.69% and 1.41% higher than that of Landsat 8, and 0.77% and 5.86% lower in forest areas. (2) High-resolution images avoided the green vegetation of urban (71.30% and 71.31% of GF-1 vegetation, 58.28% and 58.30% of Landsat 8) and bare land and roads of forest (94.97% and 94.92% of GF-1 vegetation, 95.00% and 94.99% of Landsat 8) were omitted. (3) In terms of graded area, compared with GF-1, Landsat 8 underestimates 6.67% and 6.77% of low coverage levels in urban areas, and overestimates 12.11% and 12.47% of high coverage levels in forest areas. This research reflects it is more necessary to use high-resolution images as vegetation monitoring tools in low green built-up areas and high density forest areas.





PDF全文下载地址:

https://www.ecologica.cn/stxb/article/pdf/stxb201903040405