Sediment Mobilization and Storage Dynamics of a Debris Flow Impacted Stream Channel using Multi-Temporal Structure from Motion Photogrammetry

Abstract:

High resolution topographic analysis methods have become important tools in geomorphology. Structure from Motion photogrammetry offers a compelling vehicle for geomorphic change detection in fluvial environments. This process can produce arbitrarily high resolution, geographically registered spectral and topographic coverages from a collection of overlapping digital imagery from consumer cameras. Cuneo Creek has had three historically observed episodes of rapid aggradation (1955, 1964, and 1997). The debris flow deposits continue to be major sources of sediment sixty years after the initial slope failure. Previous studies have monitored the sediment storage volume and particle size since 1976 (in 1976, 1982, 1983, 1985, 1986, 1987, 1998, 2003). We reoccupied 3 previously surveyed stream cross sections on Sept 30, 2014 and March 30, 2015, and produced photogrammetric point clouds using a pole mounted camera with a remote view finder to take nadir view images from 4.3 meters above the channel bed. Ground control points were registered using survey grade GPS and typical cross sections used over 100 images to build the structure model. This process simultaneously collects channel geometry and we used it to also generate surface texture metrics, and produced DEMs with point cloud densities above 5000 points / m2. In the period between the surveys, a five year recurrence interval discharge of 20 m3/s scoured the channel. Surface particle size distribution has been determined for each observation period using image segmentation algorithms based on spectral distance and compactness. Topographic differencing between the point clouds shows substantial channel bed mobilization and reorganization. The net decline in sediment storage is in excess of 4 x 10^5 cubic meters since the 1964 aggradation peak, with associated coarsening of surface particle sizes. These new methods provide a promising rapid assessment tool for measurement of channel responses to sediment inputs.