Terrain Ruggedness Index Calculator

The Terrain Ruggedness Index takes the eight elevation differences between a center cell and its neighbors, squares them, sums them, and takes the square root. The squaring means large departures dominate, so a single cliff face within the window drives the index far more than eight modest steps, and the result stays in the same units as elevation. Because the index is built from a fixed 3x3 neighborhood, it describes ruggedness at exactly one scale, set by the DEM cell size: the same hillside yields quite different values at 1 m, 30 m, and 90 m resolution, and the class breaks used here, level below 80 m through extremely rugged above 700 m, were established for 30 m grids and should not be transferred to other resolutions unchanged. The index has one structural limitation worth understanding before interpreting a value. It does not separate steepness from roughness. A perfectly smooth planar hillside falling at 30 percent, computed on 30 m cells, still returns a TRI near 20 m purely because the plane is tilted, even though there is no irregularity at all. Broken, hummocky ground with the same mean slope produces a much larger value, but the index alone cannot tell the two apart, which is why TRI is usually reported next to slope and to a topographic position or vector ruggedness measure. TRI is also sensitive to DEM error, since noise of a few meters in nominally flat terrain inflates the index above zero.

Attempts to reduce terrain roughness to a number predate digital elevation data. Through the 1970s workers proposed measures such as the ratio of true surface area to planimetric area, and during the 1980s wildlife biologists developed their own ruggedness indices because habitat selection by mountain ungulates clearly depended on terrain complexity that slope alone failed to capture. What these approaches lacked was a definition simple enough to compute uniformly across a large raster. That definition was supplied in 1999 by Shawn Riley, Stephen DeGloria, and Robert Elliot, who proposed the root-sum-of-squared-differences formulation together with the seven-class scheme built for 30 m elevation data. Its adoption was rapid because the arithmetic is trivial and the input is a single free DEM layer, and it now ships as a standard raster operation in GDAL, QGIS, GRASS, and WhiteboxTools. Refinements followed as the limitations became clear: in 2007 Sappington, Longshore, and Thompson introduced the vector ruggedness measure, which decomposes slope and aspect into unit vectors so that steepness and irregularity are not conflated, developed in the course of studying desert bighorn sheep in the Mojave. TRI has since spread well beyond earth science, notably into economics, where Nunn and Puga used it in 2012 to test how rugged terrain shaped historical outcomes in Africa.