Determination of grid size for digital terrain modelling in landscape investigations—exemplified by soil moisture distribution at a micro-scale

Abstract

The central problem of a combined analysis of digital terrain models (DTMs) and other landscape data is determination of a DTM grid size (w) providing a correct study of relationships between topographic variables and landscape properties. Generally, an adequate w is determined by an expert estimate, and solutions are largely subjective. We developed an experimental statistical method to determine an adequate w for DTMs applied to landscape studies. The method includes the following steps: (a) derivation of a DTM set using a series of wi , (b) performance of a correlation analysis of data on a landscape property and a topographic variable estimated with various wi , (c) plotting of correlation coefficients obtained versus w, and (d) determination of smoothed plot portions indicating intervals of an adequate w. We applied the method developed to study the ifluence of topography on the spatial distribution of soil moisture (M) at a micro-scale. We investigated the dependence of M on gradient (G), horizontal (kh ), vertical (kv ), and mean (H) landsurface curvatures. For DTM derivation, we used 13 values of wi from 1 to 7m. An interval of adequate wi for M falls between 2.25 and 3.25m in the given terrain conditions. In absolute magnitudes, correlation coefficients are largest within this interval; correlation coefficients of M with G, kh , kv and H are 0.28, 0.52, 0.50 and 0.60, respectively, for w = 3m. The results obtained demonstrate that the method actually works to identify an adequate w at a micro-scale. The method developed allows estimation of an adequate area of landform which realise a topographic control of landscape properties.