GSA Annual Meeting in Seattle, Washington, USA - 2017
Paper No. 31-2
Presentation Time: 1:45 PM
LANDSLIDE HAZARDS ASSOCIATED WITH SUBDUCTION-ZONE EARTHQUAKES (Invited Presentation)
JIBSON, Randall W., U.S. Geological Survey, Box 25046, MS 966, Denver Federal Center, Denver, CO 80225, firstname.lastname@example.org
Although most moderate and large earthquakes trigger landslides, the types and distributions of those landslides differ widely depending, in part, on the characteristics of the strong shaking generated by the earthquake. Most landslides triggered by strong shaking that is in the higher frequency ranges (~1-10 Hz) and that has short to moderate duration (<30 s) tend to be shallow, disrupted slides and falls from steep slopes. Longer period shaking of longer duration, more commonly associated with subduction-zone earthquakes, can trigger much deeper, larger landslides that can be highly destructive if they occur in developed areas. For example, the 1964 Alaska earthquake (M 9.2) generated shaking of only moderate intensity (~0.15-0.20 g), but the shaking persisted for 4-7 minutes. This long-period, long-duration shaking triggered the massive block slides that devastated downtown Anchorage and nearby residential areas including Turnagain Heights. The landslides occurred on a basal layer of sensitive clay that lost strength because of cyclic shearing during the long duration of shaking. Eyewitnesses reported that the deep landslides did not begin moving until about 1-2 minutes into the shaking sequence.
Current methods for modeling regional seismic landslide hazard focus primarily on shallow landslide hazard because that is the most common mode of failure that is present in the data sets and modeling procedures used to calibrate the hazard models. Addressing hazard from deeper landslides that are more likely to be triggered by subduction-zone earthquakes will require identification of the geologic and topographic conditions conducive to deep landsliding in areas where subduction-zone earthquakes could occur. Separate hazard models for shallow and deep landslides might need to be developed to address this issue.