Empirical Analysis of Landslide Runout in Late Pleistocene Glacial Strata in the Eastern Salish Lowland, Washington

ABSTRACT Empirical relationships were developed to estimate potential landslide runout length and deposit area for slopes in the eastern Salish (Puget) Lowland in northwestern Washington. Published empirical equations use landslide parameters of runout length (L), height (H), reach angle (α), deposit area (A), and volume (V) and linear or log-based regressions. Most data sets include varying landslide types and sizes, geology, and geographic locations, and many are for long-runout debris flows, lahars, and rock avalanches. The Salish Lowland is well suited for a regional study because of the available high-quality topographic data, slopes with relatively consistent topography and geology, and numerous landslides. The total data set includes 223 landslides mapped in the Cedar River (N = 141), North Fork Stillaguamish River (N = 17), and Skagit River–Finney Creek (N = 65) valleys. The data include H and L for debris slides (N = 162), debris flows (N = 39), and flow slides and debris avalanches (N = 22), and V and A data for the Cedar River landslides. Log-log regressions for the total data and landslide type subsets have good correlations for L as a function of H (R2 = 0.71 to 0.77) and A as a function of V (R2 = 0.76 to 0.95). However, relationships using α or tan α have very poor correlations. This methodology can be used to assess probabilities of L and A for landslide runout risk assessment of slopes in the Salish Lowland primarily underlain by late Pleistocene glacial and interglacial deposits.