Quantifying terrain fabric in digital elevation models
Eigenvector analysis of a topographic landform reveals a directional fabric consisting of surface roughness or slope, organization or fabric strength, and preferred orientation. This analysis uses a digital elevation model (DEM) to compute slope and aspect at all points in a region and uses those values to define the normal surface. Standard techniques contour the distributions, extract eigenvectors and eigenvalues from the matrix of the sum of cross products of the directional cosines, and compute eigenvalue ratios. The terrain fabric at a point depends on the size of the region used for the computation and reveals different scales over which directional fabrics operate. With large-scale DEMs, the directional fabric varies in a systematic manner and proves relatively insensitive to the horizontal resolution of the DEM or its quality and creation method. Quantitative measurement of terrain fabric belongs in all studies of terrain analysis and geomorphometry.