Characteristics and spatial differentiation of slope spectrum in different areas of Pisha Sandstone Areas

Abstract: Pisha Sandstone Area generally includes the exposed, soil- and sand-covered regions, which suffer from the most intensified water loss and soil erosion in China. It is therefore highly demanded for the slope spectrum and spatial differentiation for different geomorphological types of Pisha Sandstone. In this study, an advanced spaceborne thermal emission and reflection radiometer global digital elevation model (ASTER GDEM) of Pisha Sandstone with 30m resolution was used to acquire the slope spectrum. The regional topography was then described using the slope spectrum to combine slope characteristics. The slope spectrum was extracted with a 3° equal difference level for better expression of geomorphological features in the study areas. A quantitative analysis was performed on the curved slope and slope-landscape spectrum. The spatial differentiation was then used to effectively express the surface coverage of Pisha Sandstone. Skewness (S) and entropy of slope spectrum (H) were selected to represent the slope spectrum. In addition, some parameters of the spatial structure were used for the slope-landscape spectrum, including the patch density (PD), mean area (AREA_MN), mean perimeter-area ratio (PARA), patch cohesion index (COHESION), aggregation index (AI), and interspersion and juxtaposition index (IJI). The results showed that the slope spectrum curve of the soil-covered area was nearly normal distribution. The quantitative analysis found that the maximum entropy of slope spectrum was 1.57-1.99 nat in the soil-covered areas with a low level of landscape aggregation, indicating the large slope changed in soil-covered areas. There were obvious transitional characteristics in the sand-covered areas. The information entropy of the slope spectrum varied greatly from 0.76 to 1.71 nat, indicating that the slope of the sand-covered area was increasingly concentrated from east to west. The slope was concentrated particularly in some watersheds of sand-covered areas, where the slope spectrum was in "L" shape. There was a high frequency within 3°-6° in the slope spectrum for exposed areas. The minimum patch density was 18-35/100 hm2 in the exposed area of intensive erosion. The slope landscape index referred to the concentration of patches with different graded slopes. The curve of slope spectrum with larger information was gentle in the west of severe erosion areas. There was also a low-level aggregation of different slope graded patches in the slope landscape spectrum. The indexes indicated a great change of large slope, where the surface was relatively broken. Additionally, the slope spectrum also effectively described the spatial coupling of slope spectrum and geomorphic features, as well as the surface slope changes of different types. Consequently, different measures needed to be taken for special slopes, because the soil-, sand-covered and exposed areas varied greatly in the Pisha Sandstone Area with serious soil erosion. The finding can provide a sound reference to explore soil erosion in different types of Pisha Sandstone Area. Significantly, it still remains open in the influence of the digital elevation model (DEM) with different resolutions and slope classifications on slope spectrum in the Pisha Sandstone Areas.