Research On Lunar Slope Spectrum Variations Based On Digital Elevation Models

Lunar geomorphology has long been controlled by inner and exterior geologic process in the moon’s top layers. A systematic study of lunar geomorphological characteristics and their combination forms, spatial distribution and evolution, is important to interpret the formation and evolution process of lunar geomorphology. Recently, many scholars have deep studied on lunar geomorphology, especially on the detection algorithms and geomorphic features of craters, and have already made great achievements. However, macro-analysis of lunar geomorphology still shows insufficient and needs to be further studied. As an effective way to express the macro-terrain on the basis of the study on micro-terrain, theories and methods of slope spectrum based on DEMs have been successfully applied into researches on loess geomorphology. Therefore, in this study, slope spectrum is considered to study the spatial variations of lunar geomorphology on the macro level with multi-resolution DEMs and geologic maps. The major work and conclusions in this paper include:1) The geographical conditions to make the slope spectrum feasible in geomorphology research have been analyzed and then, the applicability of slope spectrum on lunar geomorphology research has been proved.2) A method based on spatial sampling to detect the stable slope spectrum of the Moon has been studied, which is proved to work better for lunar geomorphologic features with great complexity and highly discrete distribution than traditional ways did.3) Scale dependence of lunar slope spectrum has been researched, and it is found that lunar slope spectrum could maintain its stability when it is derived based on DEMs with resolutions ranging from 60 m to 240 m, which suggests that lunar geomorphologic features show strong self-similarity in a cell with its cell size varying from 60 m to 240 m.4) Slope gradient and surface roughness are chosen as key parameters to explore the spatial distributions of lunar slope spectrum. It is shown that lunar slope spectrum has notable spatial variation with significantly different measurements of slope spectrum among the four types of lunar landform:information entropy and the integral values of area - slope gradient and of area-surface roughness in mare, terrae highland, basin and crater increase in turn, while skewness in mare, terrae highland, basin and crater decreases in order. These differences illustrate the importance of geomorphic origin to the spatial patterns of topographic relief.5) The types of lunar landform in the test area have been classified through unsupervised classification method with 19 factors of slope spectrum and the overall classification accuracy reaches 70.3%.The work discussed in this paper is not only beneficial to provide new ideas for lunar geomorphology research, but also is effective to extend the field study of digital terrain analysis.