Application of generalized Hurst dimension rose plot in terrain altitude analysis

Abstract

This paper presents sound innovations in the processes for geographical data collection and data evaluation for multiscale analysis of terrain altitude. Data sets were collected along several transects, as well as different geographical directions, in three dissimilar zones of Spain, for the sake of Detrended Fluctuation Analysis (DFA) proper trade off. Used terrain data base consists of a 1003 × 1003-pixel matrix, with 1-m pixel resolution, provided by Official Spanish GIS.For multiscale terrain altitude analysis, generalized Hurst dimensions (H(q)) and multifractality index (ΔH(q)) were calculated on several transects crossing the areas under study. The results show, for all cases, a self-affine characteristic within most of the scale range, as well as a persistent character and a variable value of ΔH(q). Directionality analysis results, for both parameters, when represented by means of a rose plot, clearly demonstrate the differences in the scaling behavior along and across mountains, natural vegetation and crop lands and it could define a new metric for the landscape. Each of the three geographical areas covered by current statement of work clearly shows its particular anisotropy, this situation will be verified using a wider geographical data base in forthcoming works.