CAJ | 학술논문

利用密云县2006年TM遥感影像和国家森林资源清查资料，以乔木林为研究对象，通过探讨森林生物量与影响森林生物量因子之间的关系，建立森林生物量估测模型。通过相关性分析，发现所选自变量间存在多重相关性，影响模型稳定性与估测精度。针对自变量间多重共线性问题，采用偏最小二乘法建立密云县森林生物量遥感估测模型，并与主成分回归方法建立的模型进行对比，用预留的样本对模型进行检验。结果表明：预测值与实测值相关系数为0．718，精度高达90．1％。运用模型反演，密云县森林乔木生物量估测值为893．388万t。
이용밀운현2006년TM요감영상화국가삼림자원청사자료，이교목림위연구대상，통과탐토삼림생물량여영향삼림생물량인자지간적관계，건립삼림생물량고측모형。통과상관성분석，발현소선자변량간존재다중상관성，영향모형은정성여고측정도。침대자변량간다중공선성문제，채용편최소이승법건립밀운현삼림생물량요감고측모형，병여주성분회귀방법건립적모형진행대비，용예류적양본대모형진행검험。결과표명：예측치여실측치상관계수위0．718，정도고체90．1％。운용모형반연，밀운현삼림교목생물량고측치위893．388만t。
With TM remote sensing image and National Continuous Forest Inventory data of Miyun County in 2006, we studied the relationship between tree forest biomass and influencing factor , and established the forest biomass estimation model . Forest biomass remote sensing estimates were influenced by remote sensing factors and topographic factors .By the correla-tion analysis, the presence of multiple correlation between selected variables affected the stability and accuracy of the mod -el.In order to solve the problem of multicollinearity among variables , we used partial least squares method to establish re-mote sensing of forest biomass in Miyun County estimation model , and used principal component regression model to com-pare the models with control samples .The correlation coefficient between predicted and observed values was 07.18 with the accuracy of 90.1%.By examining the model with inversion forest trees in Miyun County , the estimated biomass was 8.934 million tons.