A Study Of Core Algorithm On Forest Stock Volume Quantitative Estimation Based On GIS And High Resolution Remote Sensing

Forest stock volume is a key factor in forest quantity evaluation, and it occupies animportant position in the National Forest Investigation. Due to the limitation of technologyand capital, the traditional forest stock volume estimation method cannot providehigh-precision and high real-time result for policy makers. With the rapid development andapplication of Geographic Information System (GIS) and Remote Sensing (RS), estimatingforest stock volume with high-resolution remote sensing image and a small amount of sampleplots has become a hot research direction. In recent years, most of the GIS and RS basedstudies about stock volume estimation interiorly are focused on classical Multiple LinearRegression (MLR) and have made great progress. However there is one problem that theclassical Multiple Linear Regression such as Least Square Regression cannot get good resultwhen multicollinearity exists among factors.This paper introduces the Partial Least Squares Regression (PLS-Regression) method tobuild models to estimate forest stock volume, and use a non-parametric method calledK-Nearest Neighbor as a contrast. A PLS based variable selection method—Bootstraparithmetic and the use of Remote Sensing image classification in forest stock volumeestimation are also discussed in this paper. The main algorithms were realized by C#programwith the ArcGIS Engine development component.In order to compare the efficiency and the precision of the PLS-Regression and theK-Nearest Neighbor method, this paper use the seventh National Forest Inventory sampleplots,Landsat TM image in same period and30m cell size DEM of Miyun County of Beijingto estimate its forest stock volume. With a small amount of sample plots, the RMSE ofPLS-Regression estimation is6.853m3/hm2,the Relative Standard Deviation is18.289%,whilethe RMSE of K-Nearest Neighbor estimation is7.369m3/hm2, the Relative StandardDeviation is19.009%.The results show that both the PLS-Regression and K-NearestNeighbor method can be used in forest stock volume quantitative estimation.