Estimation of Spatial Distribution and Occurrence Probability of Oak species Using IKONOS Satellite Imagery and GIS.
-A case Study on Yangpyeong-gun in Gveonggi-do-
Bo-Young Ham, Woo-Kyun Lee, J ae-Seo Chong and Jun-Hak Lee

 

ABSTRACT : In this study. spatial distribution of oak species was prepared by classifying the IKONOS imagery and the relationship between spatial distribution of oak species and geomorphological factors was analysed through logistic regression. Comparing to pixel based classification, misclassification could be reduced by segment
based classification. Segment based classification appeared also free from salt and pepper effect and showed higher overall accuracy. The occurrence probability model derived with the data of segment based classification showed better statistical performance than that of pixel based classification. The occurrence probability model represent well the well-known silvicultural and ecological knowledge about spatial distribution of oak three species. The approach in this study could apply to other three species and activities related forest resources.

 

Key words : Oak three species, spatial distribution, logistic regression, occurrence provability, GIS, IKONOS