A New Landscape Classification Approach for Quantifying Spatial Pattern of Bac Kan Province, Vietnam

Abstract

Landscape theory and its application have played an important role in natural resource exploitation and environmental protection. Various classification approaches had been employed worldwide in landscape ecology studies. This paper had developed a new hierarchical landscape classification framework for quantifying spatial pattern of Bac Kan province. A landscape formation equation was applied with three natural factors (geology, topography, and soil) and cultural factor (land use). A multi-level segmentation technique with multiresolution segmentation algorithm was chosen to segment landscape units (patches) and to categorize landscape types at different levels. The results revealed that the landscape classification of Bac Kan province has 4 hierarchical levels. Level 4, which provided full details of spatial pattern based on geologic period, elevation, soil depth, and land use, had 315 landscape types. At this level, there are 8,427 landscape units mapped with a minimum and maximum areas of 0.02 km 2 and 116.63 km 2 , respectively. A new Bac Kan landscape map at a scale of 1:100,000 along with 16 different attributes for each landscape unit was also produced. In conclusion, the framework of research methodology presented in this paper can be used as a guideline for landscape classification at provincial and national levels.