Floodplain Sedimentation in Vegetated Areas of the Elwha River Floodplain, 2012-2014

Abstract:

The removal of the Elwha and Glines Canyon Dams from the Elwha River, near Port Angeles, Washington, has released a large pulse of sediment into the middle and lower reaches of the Elwha River. This sediment has important geomorphic, hydraulic, and ecological implications. Our project focuses on the deposition of fine sediment on vegetated parts of the Elwha floodplain using field observations of sediment accumulation in combination with a simplified physics numerical model, CAESAR-Lisflood. The floodplain of the Elwha is densely vegetated and in places is characterized by large amounts of local topographic variation. This makes measuring centimeter-scale overbank sedimentation difficult using traditional approaches such as lidar and total-station based cross-section surveys. To address this problem and to provide ground truth for more traditional surveying methods, we set up over 50 short (10-20 m long) cross-sections between sets of flagged trees and surveyed, at 1-meter intervals, ground elevation with respect to a spike set in each section. Nails in the trees ensure that the horizontal position of our measurements do not shift by more than a few centimeters from year to year. This approach allows sediment accumulation to be measured repeatedly with a precision we estimate to be on the order of a few centimeters, allowing us to estimate annual rates of local sedimentation. At a given point on the floodplain, sedimentation should depend significantly on the frequency of inundation. We simulate this for the 2012-2014 period using a CAESAR-Lisflood 2-D numerical model calibrated using a set of continuously recording staff gages. CAESAR-Lisflood uses simplified-physics hydraulic routines to efficiently simulate flow depth and velocity and to drive size-specific sediment transport and morphodynamic change. This allows the model to simulate changes in flood inundation probability for the post- removal period. CAESAR-based hydraulic results are used to interpret our field-based sedimentation measurements and to develop a reach-wide estimate of overall fine sediment accumulation on the floodplain. Our CAESAR runs, which are performed for the entire post-dam removal period, produce rich 2-D representations of velocity and bed texture that should be useful to other researchers studying this system.