GSA 2020 Connects Online

Paper No. 17-11
Presentation Time: 4:35 PM

IMPROVING LANDSLIDE SUSCEPTIBILITY MODELS IN THE OREGON COAST RANGE BY INCORPORATING DETAILED LITHOLOGY AND STRUCTURE


LAHUSEN, Sean Richard, Geology, Minerals, Energy, and Geophysics Science Center, United States Geological Survey, Moffett Field, CA 94035 and GRANT, Alex, United States Geological Survey, Moffett Field, CA 94035

The central Oregon Coast Range (OCR) contains thousands of deep-seated landslides that fundamentally shape the morphology of the landscape and pose a significant risk to the residents of coastal Oregon. Despite the proximity of the OCR to the Cascadia Subduction Zone megathrust, recent work suggests that rainfall, not large magnitude earthquakes, is the predominant trigger of bedrock landslides here. While increasing mean annual rainfall is positively correlated with landslide frequency, the spatial pattern of slide occurrence remains enigmatic. Much of this region is underlain by the rhythmically-bedded sandstones and siltstones of the Eocene Tyee Formation, and previous research has shown landslide density increases to the north, where the ratio of siltstone to sandstone is highest. However, our newly mapped landslide inventory has revealed a more complex pattern of landslide density that ranges from ~0-3 landslides/km^2 from west to east. Here we analyze sedimentology, bedding attitudes, and estimated rock mass strength at dozens of sites across a 25x70 km east-west swath of the Tyee Mountain Member of the Tyee Formation to test whether subtle variations in subsurface properties within the same mapped geology exert strong control on landslide occurrence in the OCR. We find regions of high landslide density (1-3 slides/km^2) tend to be underlain by bedrock with a greater overall ratio of siltstone to sandstone, corroborating previous work in the region. We also show the importance of bedding thickness: regions with few to no deep-seated landslides often host the thickest sandstone beds. Our study reveals higher variability in landslide density across the Tyee Mountain member than previously recognized that is likely due to changes in lithological properties within the same mapped geology. We argue that landslide susceptibility models could be improved by accounting for variations in lithology and structure within generalized geologic units.