Digital elevation models (DEMs) and terrain attributes derived from them have broad application in geomorphology, geology, hydrology, ecology and climatology. Surface landscape form and shape captured by DEMs has a major control on processes associated with weathering, erosion, and deposition. Landscapes have a fundamental influence on the distribution of in-situ and transported material and have been used as the primary surrogate to map soil/regolith Furthermore, landscapes strongly influence how water moves across and through the critical zone (i.e. the area between the tree tops and groundwater aquifer) and as such affect reactions associated with – hydrolysis, oxidation and reduction, movement of solutes, biological activity and sediment transport.

Two terrain attributes ruggedness and relative topographic position, are key parameters in understanding landscape processes and have been used widely in landform classification, digital soil/regolith mapping and ecological modelling. Ruggedness measures the degree of undulation or roughness, whereas, topographic position measures the relative situation in terms of upper, middle and lower parts of the landscape. Both parameters are therefore scale-dependent and will vary according to the size of the analysis window.

Topographic Position can be calculated across diverse spatial scales to delineate different landform associations and structures (Lindsay et al., 2015). Topographic position is measured as the deviation of a central cell from the mean elevation within a specific window size or kernel. To account for the degree of variability within the analysis window the ruggedness is calculated as the standard deviation of elevations. The multi-scale topographic position model is then calculated as the deviation from mean elevation (DEV) and ruggedness parameters in the following equation:

DEV(D)=(( Z0-Zp))/Sv

where D = size of the window measured in either map units or grid cells

Z0 is the elevation of the windows centre cell

Zp window mean elevation.

Sv ruggedness

DEV is therefore a measure of relative topographic position scaled by the corresponding ruggedness. This formula is applied to every cell or pixel in the elevation grid across many window sizes to capture variations in relative topographic position from local through to regional landscape scales.

The above formula developed by Lindsay et al. (2015) has been applied to the national 3 second resolution (~90m) Digital Elevation Model derived from the Shuttle Radar Topography Mission (SRTM). The SRTM DEM was processed to remove vegetation and noise using an adaptive filtering technique (Gallant 2011). The filtered version DEM-S used in the multi-scale analysis can be sourced from https://ecat.ga.gov.au/geonetwork/srv/eng/search#!a05f7893-0050-7506-e044-00144fdd4fa6). The DEM-S was re-projected into 90m grid cells using Lambert Conformal Conic (central meridian 134; standard parallel 1 -18.000; standard parallel 2 -36.000; EPSG:3112). The continent-wide DEM-S is 49200 rows by 40800 columns and stored as 32bit floating point precision in a ~8GB data file.

Relative topographic position scaled by the corresponding ruggedness was preformed across three spatial scales (window sizes) of 0.2-8.1Km; 8.2-65.2km and 65.6-147.6Km. Incremental window sizes of 1 cell (0.09 km), 5 cells (0.43 km), and 10 cells (0.85 km) were generated for each of these window sizes. The maximum deviation corresponding to these three scales is then calculated and combined into a RGB ternary image. Each band in the ternary composite was histogram clipped to .5-99.5% and scaled to 8-bit 0-255 range values. Red, green and blues hues show variations in topographic position for local, intermediate and regional scales, respectively.

Gallant, J.C., 2011. 3 second SRTM derived Digital Elevation Model (DEM) version 1.0 Geoscience Australiametadata: https://d28rz98at9flks.cloudfront.net/72759/1secSRTM_Derived_DEMs_UserGuide_v1.0.4.pdf

Lindsay, J, B., Cockburn, J.M.H. and Russell, H.A.J. 2015. An integral image approach to performing multi-scale topographic position analysis, Geomorphology 245, 51–61.