UC Santa Barbara

Oversteepened, convex segments of stream channels called knickpoints have been utilized as markers that commonly migrate upstream and delineate abrupt changes in erosional efficiency influenced by base-level fall, changes in rock strength, or strong spatial variations in discharge. Currently, few analyses based on DEMs enable discrimination among migrating knickpoints related to changes in external forcing (baselevel fall, stream capture, and climate) versus fixed knickpoints related to internal forcing within a catchment (resistant bedrock units, coarse debris flow run-outs at tributary junctions, or landslide dams). Furthermore, few analyses have characterized the extent that migratory knickpoints steepen downstream-adjacent hillslopes. To study these interactions, we exploit a 1-m resolution LiDAR DEM of Santa Cruz Island (SCI), CA and a new algorithm that automatically extracts and measures the dimensions of any knickpoint in a regional DEM. The algorithm reduces knickpoint selection time by >99% and removes selection bias from the traditional means of regional knickpoint identification that relies on visual inspections of individual longitudinal profiles.

The spatial pattern of knickpoints located by the algorithm highlight three dominant knickpoint-forming processes on SCI: knickpoints fixed to contacts between rocks of different strength, knickpoints migrating upstream from an incision pulse caused by a significant stream capture event, and mobile knickpoints stemming from relative sea-level fall and wave erosion of sea-cliffs. Nearly 36% of the hillslope area downstream from migratory knickpoints has slopes above threshold values of 35˚, and on average, these regions display a 4-6˚ steeper median hillslope-gradient than hillslopes upstream from migratory knickpoints. Hillslope-gradient histograms are nearly identical upstream and downstream from knickpoints fixed to spatial changes in rock strength or when re-analyzing hillslopes surrounding migratory knickpoints with a 10-m resolution DEM. From regionally extensive map of stream knickpoints, geological context, and hillslope attributes extracted from a 1-m resolution DEM, new insights emerge on the controls of landscape evolution; insights that would be much harder to obtain through individual knickpoint selection or lower resolution imagery.