Characterizing the style, composition, and timing of Cedar River Landslide Complex

Abstract

The Cedar River Landslide Complex (CRLC) is a series of debris flowslides about 3 km down river from the Masonry Dam at Chester Morse Reservoir in the Cedar River Municipal Watershed, the source of drinking water for 1.4 million people in Seattle, WA. These hillslope failures all originated from the terminal moraine deposits of the Puget lobe of the Vashon glacier. The CRLC was discovered in 2009 by new geologic mapping and LiDAR imagery. This study created a landslide inventory, including dimensions, composition and dating all units of the CRLC. Fieldwork involved digging test pits at sites on three different units of the CRLC. Soil samples of all units, including both landslide and undisturbed material were collected. Sieve analysis tests, water content, and mineralogy data were collected. Volumes of each unit of the CRLC were calculated to determine mobility indices for each landslide to compare the CRLC to other large landslides in the region. To estimate the ages of the CRLC slides, radiocarbon samples were collected and surface roughness of the landscape was calculated using GIS. The underlying material in the Cedar River valley is well-graded gravel and the CRLC is composed mostly of well-graded gravel and sand. The landslide material has lower water content than the underlying material. Five absolute dates on three landslides were calculated using radiocarbon dating. The dated CRLC slides range in age from 170 Yr B.P. to younger than 6,850 Yr B.P. Standard deviation of slope (SDS) was used as a measure of surface roughness, the roughness of the CRLC has a small range in values (4.15 to 4.51). No correlation between surface roughness and landslide age was found. Several reasons for this are discussed. No correlation suggests the surface roughness measure used in this study (standard deviation of slope) is not a good measure of landslide age in this watershed. Absolute ages are calculated using organic samples found in soil horizons below the landslide material. As such, these dates represent only the maximum age and may not accurately date the events. A series of landslide triggers are considered, including coseismic, glacial retreat, and precipitation events, however our data set does not allow definitive conclusions or elimination of any particular trigger. Future work, including gathering more samples for landslide ages, is needed to consider the full landslide history of this complex.