DETECTING CHANNEL INCISION IN THE NORTHERN GREAT BASIN, U.S.A. USING HIGH RESOLUTION DIGITAL TERRAIN MODELS

Channel incision is a pervasive and long-acknowledged environmental problem throughout the western United States. Incised channels – or, gullies - are a major source of erosion in semi-arid regions, and often disproportionately affect agricultural lands. There, the rapid lowering of channel bottoms relative to adjacent floodplains locally alters hydrology and diminishes the agricultural productivity of the land.

Though there is a considerable body of research concerning gullies, we do not yet know the extent of incision in the intermountain west, its contribution to sediment budgets, nor how rapidly it is progressing. High resolution photogrammetry provides a heretofore impossible opportunity to partially automate the mapping and quantification of gullies from digital terrain models (DTM). Previously, DTMs were unable to resolve smaller gullies (<5 m), whose rapid propagation cannot easily be captured by field survey methods. High resolution DTMs (<0.1 m) not only resolve gullies, but provide quantifiable data, the automatic extraction of which would permit relatively rapid mapping following imagery flights. Gullies are particularly well suited to automated detection; despite ranging considerably in size, they exhibit a fairly consistent set of morphometric characteristics. In this, they provide an opportunity to test a toolkit whose utility may extend to other geomorphic features.

Here, we present initial results of gully extent and sediment volumes in the Northern Great Basin. The rugged landscape provides an ideal setting to refine established detection techniques, as there are a number of features (e.g. canals, jeep roads, narrow valleys) that are easily misidentified as gullies. Gullies are mapped as linear features at the landscape scale (HUC-8) to estimate total length and as polygons at the catchment scale (HUC-12) to estimate volumes of eroded sediment. Estimates were validated on the ground to verify both gross classification and magnitude of incision.