关键词: 地形因子, 岩性分类, 聚类分析, 因子分析, DEM, 宏观地形复杂度指数, 微观曲率指数, 桂林-阳朔地区

Abstract:

Topographic and geomorphological features are important for remote sensing lithologic interpretation, but there still lacks quantitative analysis of the geological correlation between topography and lithology. Digital Elevation Model (DEM) is a digital representation and simulation of topographic and geomorphological features in space. Topographic factors derived from DEM can describe the characteristics of concave and convex changes of different topographic slopes and undulations, thus quantitatively describing the characteristics of different topographic and geomorphological features. This paper focuses on studying the mathematical and geological significance of topographic factors, by establishing the quantitative correlation between lithology and topographic factors combination for classifying rock types. Based on ASTER GDEM data, this paper extracted 12 topographic factors, such as slope, profile curvature, maximum curvature, topographic relief, and elevation coefficient of variation. Based on analysis of the geological significance of each topographic factor, the characteristics and correlation of each lithologic topographic factor were studied by cluster analysis and variance analysis, and the quantitative difference between lithologies in the study area was established. In addition, the factor analysis method was used to study the dominant factors in the process of lithological classification, and to determine the appropriate lithological classification method and achieve quantitative lithological classification. Quantitative research was carried out on the two basic issues: whether topographic factors can be used for lithologic classification and how to use topographic factors to identify lithology. Experimental results show: (1) There was a significant correlation between lithology and topographic factors and the topographic factors of different lithologies could be distinguished significantly. (2) There were significant differences among the topographic factors (combinations) of different lithology and topography. The macro-topographic complexity index (MTI) and micro-curvature index (MCI) based on factor analysis had a relatively high classification accuracy of 77.36% for the rock types in the study area, highly consistent with the actual lithologic types. Our findings suggest that topographic factors such as slope and topographic complexity can be used for lithological classification, and that factor analysis can be used to obtain quantitative indicators reflecting macro-topographic complexity and micro-topographic curvature characteristics. This study can serve as a methodological reference for quantitative classification of lithology.

Key words: topographic factors, lithological classification, cluster analysis, factor analysis, DEM, macro-topographic complexity index, micro-curvature index, Guilin-Yangshuo