西北林学院学报2024，Vol.39Issue(1)：73-80,8.DOI:10.3969/j.issn.1001-7461.2024.01.10
不同植被类型森林生物量反演模型研究
Inversion Models of Forest Biomass for Different Vegetation Types——A Case Study of Public Welfare Forests in Hunan Province
摘要
Abstract
Ecological public welfare forest is an important foundation for building national ecological security and an important guarantee for implementing the"Two Mountains Theory".In this study,public welfare forests with different vegetation types(coniferous forests,broad-leaved forests,mixed forests of coniferous and broad-leaved trees,bamboo forests,and shrubs)occurring in Hunan Province were selected as research objects.By using the fixed plot monitoring data of public welfare forests in Hunan Province and Landsat 8 remote sensing data in 2021,three inversion models,including the biomass support vector machine model,decision tree model,and random forest model,were constructed for public welfare forests with different vegetation types.The results showed that among the three models,the random forest model had the high-est estimation accuracy,with the best fit for bamboo forests(R2:0.79 and RMSE:25.60 t·hm-2).The re-search results confirmed that vegetation classification inversion based on the random forest model could ef-fectively improve the estimation accuracy of forest biomass and provide a new method for improving the ac-curacy of forest biomass estimation.关键词
植被类型/生物量/Landsat 8 OLI/机器学习/湖南省公益林Key words
vegetation type/biomass/Landsat 8 OLI/machine learning/Hunan public welfare forest分类
农业科技引用本文复制引用
刘慧婷,潘俊,符玥,王光军,樊红波,胡孔飞..不同植被类型森林生物量反演模型研究[J].西北林学院学报,2024,39(1):73-80,8.基金项目
湖南省重点研发项目(2022NK2018) （2022NK2018）
广西壮族自治区科技攻关计划项目(桂科AB21220026). （桂科AB21220026）
