Study On Estimate Model Of Biomass Of Chinese Pie Forest In Huanglong Based On RS And GIS Technology

Huanglong is one of the five major forest areas in Shaanxi which is located in the southeastern Loess Plateau in northern Shaanxi. Chinese pine is one of the dominant species and takes on laminated distribution. This research uses RS and GIS technology, takes the TM image as the main data, combines forest resource inventory data and field survey data of 70 plots, has carried on the biomass estimation to the artificial Chinese pine in Huanglong mountain. The research indicated:The pertinence between field survey biomass and remote sensing data in the study area is good, therefore, it is feasible to estimate Chinese pine biomass in the research area using the remote sensing data. Correlation analysis showed that: field survey biomass has relevance to TM1-TM5, TM7 , TM3 has the highest relevance and R2 is 0.6836; RVI, GVI and NDVI are related to each other and RVI has the highest relevance; Through analysis Correlation between the biomass and terrain factor, Slope within terrain has the highest correlation coefficient which is 0.689.The biomass was well linear correlated with vegetation index and terrain factor. The results that the well relevant factors are carried through a unitary linear and nonlinear regression analysis is,R~2 of single-band linear are that: TM1:0.1610,TM2:0.4953,TM3:0.827,TM4:0.787,TM5:0.585,TM7:0.652, 0.1029,0.0001,0.0002,0.0007,0.0002,0.0007;R2 of vegetation index and terrain factors are that: NDVI: 0.1358,RVI:0.2731,GVI:0.1936. R2 of terrain linear and nonlinear are that: SLOPE:0.4749,ASPCET:0.0847,and nonlinear are that: NDVI:0.0002,RVI:0.0001,GVI:0.0001,SLOPE:0.0072,ASPCET:0.8709.This research that builds a multiple regression model on RS for the biomass of Chinese pine . The optimal regression model is determined as follows:Y=0.376TM3+0.260TM7+0.379TM4+0.197TM5+0.107SLOPE+0.191TM2+0.122NDVI-3.550R~2 is 0.952. According to this model, Chinese pine biomass in Caijiachuan Forest Farm of Huanglong mountain is 37.892Tg.Using the established remote sensing forecast model to verify the accuracy, the estimate accuracy of field spot model is up to 90%. the number of field spot is 57, accounting for 87%, with an average accuracy of 93.88%. The estimate accuracy of the established model is relatively higher, so it is suitable for Chinese pine biomass estimation of the study area.