Comparison of Ground Inventory Method and Remote Sensing Data Using Nonparametric Algorithm for Estimating Forest Quantitative Characteristics
Comparison of Ground Inventory Method and Remote Sensing Data Using Nonparametric Algorithm for Estimating Forest Quantitative Characteristics
Roya Abedi1
1) Department of Forestry, Ahar Faculty of Agriculture and Natural Resources,University of Tabriz, Iran,
IOI Code
:
XBAE-DZFGE
Publication :
Demo - Website of Saman Conference Management System(ngtu.ir)
Abstract :
Forest inventory provides an important source of comprehensive information. The purpose of this study was to compare the ground inventory method and non-parametric K-nearest neighbor algorithm in estimating the most important forest stands characteristics including mean diameter, height, basal area, volume and density of trees. In the present study, 0.1 ha circular sampling plots in systematic random grid was used to ground inventory. Then, forest characteristics were estimated by using the non-parametric k-nearest neighbor algorithm on IRS P6 LISS III satellite images in the R software environment and the yaImpute package. The results showed that the statistical averages of forest characteristics using the remote sensing data were very close to the ground inventory results and had a slight difference, which indicates the power of these images and the algorithm used to estimate forest characteristics. It is suggested that such studies be conducted to introduce the best alternative or complementary method of ground inventory in forest studies in order to provide more accurate information to forest management.
Keywords : forest inventory k nearest neighbor non parametric algorithm