The Research Of Estimating Pinus Massioniana Lamb Volume In The Middle Of Guizhou Based On Landsat8 Data

Forest volume is an important indicator which reflects the number of forest resources,it plays an important role on the prediction of the number of forest resources as well as forest management. Given this,I take Guiyang as the research area, landsat8 Land Imager data(concluding the gray value of B1- B7 bands, DVI, NDVI and RVI) as the information sources of remote sensing,together with topographic factors, stand factors and soil factors in my paper to do the research of Pinus massioniana Lamb stand volume estimating model in the middle of Guizhou. Meanwhile, the multi-spectral data processing method of landsat8 Land Imager, the processing method of the independent variable factor and a combination of ecological significance, since the method for selecting the optimal variable factor variables, the independent variable factor of the model as well as the contribution of the model Inspection and precision analysis were studied, the conclusions are as follows:(1) Metadata conducted radiometric calibration and atmospheric correction processing to avoid the detrimental effect of water vapor and carbon dioxide. Imaging postprocess is closer to the grain phenomenon triplet vegetation, explaining the necessity of radiometric calibration and atmospheric correction.(2)In order to solve the sample across the pixel element even when an exception as the impact on accuracy and other issues, focusing analysis enhanced is executed,which eliminates the abnormal pixel luminance value and improves the extraction of pixel gray precision values effectively.(3) In this study,the independent variable facators was chosen with taking the theory about the growth of Pinus massioniana Lamb as a guidance,combined with the mathematical statistics,to demonstrate the choice of every variable factor and the division category of every independent variable factor was reasonable and applicable significance.(4)To solve the problem of multicollinearity between independent variable factors,the study makes use of stepwise regression method to select the optimal combination of independent variable factors :slop,aspect,soil depth,elevation,canopy density,age group,B2,B5,B6,B7,DVI and RVI;(5) The independent variables to model the contribution factor in the order: stand factor(37.90%), remote sensing factor(32.07%), site factor(30.03%),;(6) With a random sample of 30 test plots,the prediction accuracy of the model was tested, the average level of 83.22% accuracy, it can provide a reference for scale pine in central Guizhou region Volume Estimation.