摘要
以四川省遂宁市安居区为研究区域,研究了基于小班对象的森林蓄积量遥感估测模型的构建与合理性诊断。首先论述了建立森林蓄积量模型所需的各种数据,包括定性、定量数据以及遥感影像的派生数据。其次采用最大膨胀因子法及残差平方和法对选取的60个小班样地数据进行最优变量的选择,最终TM2、TM5、TM7、TM4×37、有林地、郁闭度等6个变量成为估测模型的主要因子。再次选取24个样地数据并采用残差分析方法对构建的模型进行合理性诊断,诊断结果显示模型基本合理。最后选用岭迹法对其余36个样地进行模型构建。构建的模型所测蓄积量的相对误差达8.0%,可应用到生产实践当中。
The construction of remote sensing estimation model for forest volume and rationality diagnosis based on small class object were studied using the Anju district of Suining in Sichuan province as research object.Firstly,the data to establish the model for forest volume was described,including qualitative and quantitative data and the data derived from remote sensing images.60 classes sample data were selected for optimal variable selection with the maximum expansion factor method and the residual sum of squares method.Finally,6 variables such as TM2,TM5,TM7,TM4×37,forested land and canopy density were selected as the major factors of estimation model.Then 24 plots were selected and residual analysis method was used to make a signisis for rationality of the constructed model,and the diagnostic results showed that the model was reasonable.Finally,ridge estimation method was selected to construct the model using the remaining 36 plots.The relative error of volume measured by the constructed model reached 8.0%,which could be applied to production practices produced.It was proved that the forest volume remote sensing estimation model based on small class objects could be applied to the production practice.
出处
《湖北农业科学》
北大核心
2011年第23期4945-4948,共4页
Hubei Agricultural Sciences
基金
国家自然科学基金项目(40771144)
四川师范大学校级项目(10MSL08)
关键词
小班
遥感估测
森林蓄积量
合理性诊断
small class
remote sensing estimation
forest volume
rationality diagnosis
