期刊论文

Remote sensing inversion of forest volume based on HLS data

中文

森林与环境学报

2096-0018

2021

41

6

620-628

10.13324/j.cnki.jfcf.2021.06.008

国家自然科学基金面上项目(31971578) “十三五”国家重点研发计划项目:人工林资源监测关键技术研究(2017YFD0600900) 湖南省教育厅科学研究重点项目(17A225)。

本校为第一机构

通过对比陆地卫星8号(Landsat 8)与协同陆地卫星和哨兵2号(HLS)数据在波段信息及森林蓄积量建模效果的差异,探索HLS数据源在森林蓄积量反演中的应用潜力。以内蒙古自治区旺业甸林场为研究区,以Landsat 8和HLS为数据源提取遥感变量,结合森林蓄积量样地实测数据,利用线性逐步回归(LSR)和逐步随机森林(SRF)进行特征变量筛选,分别构建线性和非线性模型开展森林蓄积量反演,并进行精度验证及对比分析。结果表明:HLS影像中的7个波段与Landsat 8影像各对应波段均表现为极显著相关,相关系数绝对值均大于0.7(P0.05)。利用HLS影像进行森林蓄积量估测,可得到与Landsat 8相似的效果,且HLS数据无需进行繁琐的预...

By comparing the differences between Landsat 8 and Harmonized Landsat and Sentinel-2 (HLS) data in terms of band information and forest volume modeling, the application potential of the HLS source in forest volume inversion was explored. Considering Wangyedian Forest Farm in the Inner Mongolia Autonomous Region as the study area, Landsat 8 and HLS images were used as data sources to extract remote sensing variables. In combination with the measured data of forest volume sample plots, linear stepwise regression and stepwise random forest were us...