基于ALOS数据的云南松林蓄积量遥感估测方法  

The Remote Sensing Estimation Methods of Pinus yunnanensis Forest Stock Volume Based on ALOS Data

作　　者：王丽媛[1] 胡振华[1] 徐天蜀[1] WANG Liyuan HU Zhenhua XU Tianshu(College of Forestry ,Southwest Forestry University ,Kunming 650224, Yunnan, Chin)

机构地区：[1]西南林业大学林学院,云南昆明650224

出　　处：《福建林业科技》2017年第1期10-15,共6页Journal of Fujian Forestry Science and Technology

基　　金：云南省林学一流学科建设经费资助(51600625);国家自然科学基金项目(31260156)

摘　　要：基于ALOS数据和样地实测数据为数据源,云南省宜良县为研究区,以提高森林蓄积量遥感估测模型精度为目的,从遥感、GIS、郁闭度等方面选取与森林蓄积量相关性较高的因子作为自变量。采用逐步回归法、主成分分析法、偏最小二乘法3种模型估测宜良县的云南松林蓄积量。从3种模型的拟合效果和估测精度比较结果表明:偏最小二乘法精度最高,调整决定系数为0.754,预测精度为82.77%,与主成分分析等传统线性估测模型相比精度有较大改善。Based on ALOS data and forestry resource inventory data,in order to improve the accuracy of predicting forest volume by remote sensing. Analyzed the relative relation between remote sensing variables,forest canopy and forest volume,setting the highly relevant factor. The test site was located in Yiliang of Yunnan Province. Multi-stepwise regression model,principal component model and partial least square regression models were built to estimate forest stock volume of Pinus yunnanensis. And the accuracy of these three models were analyzed and compared to study the application of regression models in forest stock volume. The results showed that the accuracy and fitting effects of partial least square regression model are better than the other two models. Determination coefficient was 0. 754 and prediction accuracy reached 82. 77%. Comparing this method with traditional linear estimation model such as principal component analysis,accuracy shown a greatly improvement.

关 键 词：ALOS数据 偏最小二乘法 云南松 森林蓄积量 

分 类 号：S791.257[农业科学—林木遗传育种] S771.8[农业科学—林学]