Academic Article
Landscape Classification using Principal Component Analysis and Fuzzy Classification: Archaeological Sites and their Natural Surroundings in Central Mongolia
- Title
- Landscape Classification using Principal Component Analysis and Fuzzy Classification: Archaeological Sites and their Natural Surroundings in Central Mongolia
- Date
- 2011
- Is Part Of
- DIE ERDE – Journal of the Geographical Society of Berlin
- Volume
- 142
- Issue
- 3
- Pages
- 213-233
- Language
- eng
- Rights
- Copyright (c)
- ISSN
- 0013-9998
- Short Title
- Landscape Classification using Principal Component Analysis and Fuzzy Classification
- Abstract
- The middle and upper Orkhon Valley in Central Mongolia (47.5°N, 102.5°E) hosts a multitude of diverse archaeological features. Most of them – including the well-known ancient cities of Karakorum and Karabalgasun – have only rarely been described in their geographical setups. The aim of this study is to describe, classify and analyse their surrounding landscapes and consequently characterise these sites geographically. This analysis is based on freely available raster datasets that offer information about topography, surface reflectance and derivatives. Principal component analysis is applied as a dimensional reduction technique. Subsequently, a fuzzy-logic approach leads to a classification scheme in which archaeological features are embedded and therefore distinguishable. A distinct difference in preferences regarding to choose a site location can be made and confirmed by semiautomatic analysis, comparing burial and ritual places and settlements. Walled enclosures and settlements are connected to planar steppe regions, whereas burial and ritual places are embedded in mountainous and hilly environments.
Linked resources
Export
Position: 2856 (18 views)