基于小班对象的森林蓄积量遥感估测模型合理性诊断.docVIP

基于小班对象的森林蓄积量遥感估测模型合理性诊断 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6 0个小班样地数据进行 最优变量的选择，最终TM2、TM5、TM7、TM、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2 4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3 6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。 关键词：小班；遥感估测；森林蓄积量；合理性诊断 中图分类号：P 2 3 7文献标识码：A文章编号： 0439- 8 11423-4945-04 Rationality Diagnosis of Remote Estimation Model of Forest Volume Based on Small Class Objects CHENE Wu-xuel, NIYong2, YANG Cun-jianl，ZHOU Jie-mingl，ZHOU Wan-cun3 Abstract： The construct i on of remote sensing e s t i m a t i on model for forest volum e and rationality diagnosis ba sed on small class object we re studied using the A n j u d is t r i c t of Suining in Sichuan p r ovince as research o b j e c t . Firstly, the data to e s t a b 1 ish the model for forest v o 1 u m e was described, includ in g qualitative and quantit at i v e data and the data deri v e d from remote sensing i m a g e s ? 60 classes sample data we re selected for optimal v a r iable selection with the ma ximum expans ion factor me thod and the residual sum of squares me thod. Finally， 6 v ariables such as T M 2 , T M 5 , T M 7， TM， forested land and can o p y density were selected a s the major factors of estima t i o n mode 1 . Then 24 plots we re selected and residual an alysis method was used to ma k e a signisis for rat i o n a 1 i t y of the constructed model， and the diagnostic results s h o wed that the model was reasona ble. Finally， ridge e s t i m a t ion method was selected to co nstruct the model using the remaining 36 plots. The r e 1 a t i v e error of volume measur e d by the constructed model reached 8. 0%, which could be a pplied to production practi c e s produced. It was proved t h at the forest volume remote sensing estimation model ba s e d on small class objects c ould be applied to the product ion practice. Key words： small class； re mote sensing estimation； for est volume; rational