Research On Quantifying Terrain Complexity

Terrain complexity is one of important terrain parameters in digital terrain analysis (DTA). However unlike slope, aspect parameter which has clear mathematic and topographic meaning, terrain complexity is ambiguous. Terrain complexity, which describe roughness and complexity of the terrain surface, is not only an important terrain parameters in digital terrain analysis, but also widely applied in the fields of reduction of topographic data, terrain classification and visualization, mapping and surveying, soil erosion, surface turbulence and biological richness assessment. Terrain complexity has a widely semantic implication and often used to describe the variability of the terrain surface. Up to now, the existed terrain descriptor can be classified into three types:statistical sense, geometrical sense and semantic sense. These indices partly describe roughness and complexity of the terrain surface, but cannot appropriately characterize it. Therefore, the important problems should be solved are what the scientific meaning of terrain complexity is and how to quantify terrain complexity. This paper wants to seek an approach to quantitatively evaluate topographic complexity based on grid DEM.Firstly, this paper discusses the scientific connotation of the terrain complexity and three basic characteristics that is objectivity, local, scale invariance and rotation invariant. And this paper presents an approach of quantifying terrain complexity that integrates the Constitution theory and fuzzy mathematics. According to express the spatial distribution characteristics of the terrain data, the general set should be extended to three tuples. The three tuples can make up the shortage of general set that is hard to describe the spatial data. Therefore, the article constructs the conceptual model of the terrain complexity based on the new general set.Secondly, according to the structure characteristics of the grid DEM, this paper chooses a single grid as the unit of the conceptual model of terrain complexity. Also slope is used to describe the attribute of the terrain complexity conceptual model. With the method of landscape ecology, this study investigates the quantitative depiction and spatial distribution of the terrain complexity in detail. So a single grid, the distribution function and the four parameters of landscape pattern are used to describe. Lastly, terrain complexity of an area is given by fuzzy logic inference which combines the terrain parameters of the terrain complexity quantification model and the semantic sense. That is fuzzy membership values is used to represent the terrain complexity. With the difference of the terrain complexity, it divides into three degrees that is complex, medium complex and simple. The results showed that the computed terrain complexity presented in this paper is much higher over areas in dramatic changes area than flat. And it shows that this model is a sound terrain parameter to evaluate terrain complexity.