APPLICATION OF TERRESTRIAL-BASED STRUCTURE-FROM-MOTION (SFM) PHOTOGRAMMETRY TO THE MEASUREMENT AND MONITORING OF RIVER BLUFF EROSION

Excess sediment is problematic in many river systems. The cause of excess sediment may be natural (related to geologic history of the watershed) or human caused and/or amplified. Sediment budgets are one tool for accounting for shifts in sediment sources, sinks, and net efflux in a watershed. A fine sediment budget completed for the Le Sueur River, a river impaired for sediment in south central Minnesota, suggests that modern sediment loads are greater today than they were over the Holocene, and that river bluffs (comprised of Holocene glacial till and fluvial terrace deposits) are the greatest contributor of fine sediment to the Le Sueur River. The cause of accelerated bluff erosion, whether human, natural, or some combination, is currently not known. Although we have identified several mechanisms that drive bluff erosion (fluvial toe erosion, freeze-thaw, groundwater sapping, tree throw, positive pore water pressures), we do not know the relative contribution of sediment from each of these processes. The main purpose of this paper is to explore the use of Structure from Motion (SfM) photogrammetry techniques for measuring and monitoring the rates and processes by which river bluffs retreat. In 2013, we deployed a remote sensing platform (that is still active) consisting of three consumer-grade cameras to capture bluff activity on a daily time step at two actively eroding bluffs along the Le Sueur River. Since their deployment, we have captured several small bluff failures that occurred on the receding limb of the hydrograph. This suggests that fluvial undercutting of the bluff toe and oversteepening may be an important diver of bluff erosion. The largest failure captured so far occurred in June 2014 during an anomalously high rainfall event when positive pore water pressures were likely the driver bluff failure. To constrain the volumes and rates of bluff erosion we have performed multiple detailed photo surveys and surveyed using a terrestrial laser scanner (TLS) before and after bluff failures. We georeferenced the point clouds from common control points and built digital elevation models (DEMs). We differenced DEMs from before and after bluff failures to compute volumes of erosion and deposition. Here we report the magnitude of erosion and deposition resulting from different processes of bluff retreat.