基于MNF变换的森林乔木地上碳储量遥感估测模型  

Estimation model of tree aboveground carbon storage in forest based on minimum noise fraction transformation

作　　者：徐定[1] 刘书剑 彭道黎[1] 

机构地区：[1]北京林业大学林学院,北京100083 [2]贵州省黔西南州林业局,贵州兴义562400

出　　处：《中南林业科技大学学报》2012年第7期54-57,共4页Journal of Central South University of Forestry & Technology

基　　金：国家"十一五"林业科技支撑项目(2006BAD23B05);国家级林业推广项目(201145)

摘　　要：森林碳储量遥感监测是目前林业定量遥感的研究热点之一。以沽源县为研究对象,采用TM影像,对原始自变量进行MNF变换后得到新变量,建立森林乔木地上碳储量遥感估测模型。模型相关系数R为0.708,通过实测样地对估测结果进行检验,精度达到89.89%。结果表明,当自变量间存在显著复相关性情况下,通过MNF变换能有效去除复相关性,用新变量建立的模型估测效果较好。Remote sensing monitoring of forest carbon storage is one of the research hotspot in the forestry quantitative remote sensing at present. By taking Guyuan county as the research object, using the TM image, getting new variables after MNF transformation, the forest tree aboveground carbon storage remote sensing estimation model was established. The correlation was 0.708, and the estimating results were inspected through the measured sample, the accuracy reached 89.89%. The results show that when there were significantly complex correlation among the independent variables, the MNF transformation can effectively remove the multiple correlation, the predicting effects of the model established with new variables was very good.

关 键 词：MNF 复相关性:碳储量 遥感 

分 类 号：S718.556[农业科学—林学]