William H. Schulz

Landslides Triggered by the 2002 Denali Fault, Alaska, Earthquake and the Inferred Nature of the Strong Shaking

The 2002 M7.9 Denali fault, Alaska, earthquake triggered thousands of landslides, primarily rock falls and rock slides, that ranged in volume from rock falls of a few cubic meters to rock avalanches having volumes as great as 15×106 m3. The pattern of landsliding was unusual; the number of slides was less than expected for an earthquake of this magnitude, and the landslides were concentrated in a narrow zone 30-km wide that straddled the fault rupture over its entire 300-km length. The large rock avalanches all clustered along the western third of the rupture zone where acceleration levels and ground-shaking frequencies are thought to have been the highest. Inferences about near-field strong shaking characteristics drawn from the interpretation of the landslide distribution are consistent with results of recent inversion modeling that indicate high-frequency energy generation was greatest in the western part of the fault rupture zone and decreased markedly to the east.

Residual shear strength variability as a primary control on movement of landslides reactivated by earthquake‐induced ground motion: Implications for coastal Oregon, U.S.

Most large seismogenic landslides are reactivations of preexisting landslides with basal shear zones in the residual strength condition. Residual shear strength often varies during rapid displacement, but the response of residual shear zones to seismic loading is largely unknown. We used a ring shear apparatus to perform simulated seismic loading tests, constant displacement rate tests, and tests during which shear stress was gradually varied on specimens from two landslides to improve understanding of coseismic landslide reactivation and to identify shear strength models valid for slow gravitational failure through rapid coseismic failure. The landslides we studied represent many along the Oregon, U.S., coast. Seismic loading tests resulted in (1) catastrophic failure involving unbounded displacement when stresses represented those for the existing landslides and (2) limited to unbounded displacement when stresses represented those for hypothetical dormant landslides, suggesting that coseismic landslide reactivation may be significant during future great earthquakes occurring near the Oregon Coast. Constant displacement rate tests indicated that shear strength decreased exponentially during the first few decimeters of displacement but increased logarithmically with increasing displacement rate when sheared at 0.001 cm s−1 or greater. Dynamic shear resistance estimated from shear strength models correlated well with stresses observed during seismic loading tests, indicating that displacement rate and amount primarily controlled failure characteristics. We developed a stress-based approach to estimate coseismic landslide displacement that utilizes the variable shear strength model. The approach produced results that compared favorably to observations made during seismic loading tests, indicating its utility for application to landslides.

Three-dimensional surface deformation derived from airborne interferometric UAVSAR: Application to the Slumgullion Landslide

Author(s): Delbridge, BG; Burgmann, R; Fielding, E; Hensley, S; Schulz, WH | Abstract: ©2016. American Geophysical Union. All Rights Reserved. In order to provide surface geodetic measurements with “landslide-wide” spatial coverage, we develop and validate a method for the characterization of 3-D surface deformation using the unique capabilities of the Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) airborne repeat-pass radar interferometry system. We apply our method at the well-studied Slumgullion Landslide, which is 3.9 km long and moves persistently at rates up to ∼2 cm/day. A comparison with concurrent GPS measurements validates this method and shows that it provides reliable and accurate 3-D surface deformation measurements. The UAVSAR-derived vector velocity field measurements accurately capture the sharp boundaries defining previously identified kinematic units and geomorphic domains within the landslide. We acquired data across the landslide during spring and summer and identify that the landslide moves more slowly during summer except at its head, presumably in response to spatiotemporal variations in snowmelt infiltration. In order to constrain the mechanics controlling landslide motion from surface velocity measurements, we present an inversion framework for the extraction of slide thickness and basal geometry from dense 3-D surface velocity fields. We find that the average depth of the Slumgullion Landslide is 7.5 m, several meters less than previous depth estimates. We show that by considering a viscoplastic rheology, we can derive tighter theoretical bounds on the rheological parameter relating mean horizontal flow rate to surface velocity. Using inclinometer data for slow-moving, clay-rich landslides across the globe, we find a consistent value for the rheological parameter of 0.85 ± 0.08.

Slidequake Generation versus Viscous Creep at Softrock-landslides: Synopsis of Three Different Scenarios at Slumgullion Landslide, Heumoes Slope, and Super-Sauze Mudslide

In this study, we describe conditions for slidequake generation at three different creeping softrock landslides: the Slumgullion landslide in the San Juan Mountains, Colorado, U.S., the Heumoes slope in the Austrian Alps, and the mudslide in Super-Sauze, French Alps. From a geomorphologic point of view, all three landslides are classified as creeping landslides with average velocities between centimeters to meters per year. Associating creep with viscous flow, and considering the largely saturated, clayey consistency of the slope body, one would not expect any brittle behavior. Thus, it came as a surprise that impulsive seismic signals indicative of shear fracture could be discovered by sensitive passive monitoring methods at all three slopes. These fracture signals occur in episodes, have similar signatures as small earthquakes, and could be located within the slide bodies, i.e., are evidence of slidequakes. Our investigations identified seismic and aseismic slip in each slide, with slidequakes focusing at significant bedrock structures or at lateral boundaries. Synoptic comparison of three scenarios underlines the importance of landslide-bedrock and landslide-lateral boundary interactions under gravitational loading and Mohr-Coulomb-type failure. Comparison to frictional and asperity models of crustal- and plate-scale boundaries may pave the way to a comprehensive understanding of slidequake generation, and future slope failure prediction.

Clayey Landslide Initiation and Acceleration Strongly Modulated by Soil Swelling

Largely unknown mechanisms restrain motion of clay-rich, slow-moving landslides that are widespread worldwide and rarely accelerate catastrophically. We studied a clayey, slow-moving landslide typical of thousands in Northern California, USA, to decipher hydrologic-mechanical interactions that modulate landslide dynamics. Similar to some other studies, observed pore-water pressures correlated poorly with landslide reactivation and speed. In situ and laboratory measurements strongly suggested that variable pressure along the landslide’s lateral shear boundaries resulting from seasonal soil expansion and contraction modulated its reactivation and speed. Slope-stability modeling suggested that the landslide’s observed behavior could be predicted by including transient swell pressure as a resistance term, whereas modeling considering only transient hydrologic conditions predicted movement five to six months prior to when it was observed. All clayey soils swell to some degree; hence, our findings suggest that swell pressure likely modulates motion of many landslides and should be considered to improve forecasts of clayey landslide initiation and mobility.

Plenary: Progress in Regional Landslide Hazard Assessment—Examples from the USA

Landslide hazard assessment at local and regional scales contributes to mitigation of landslides in developing and densely populated areas by providing information for (1) land development and redevelopment plans and regulations, (2) emergency preparedness plans, and (3) economic analysis to (a) set priorities for engineered mitigation projects and (b) define areas of similar levels of hazard for insurance purposes. US Geological Survey (USGS) research on landslide hazard assessment has explored a range of methods that can be used to estimate temporal and spatial landslide potential and probability for various scales and purposes. Cases taken primarily from our work in the U.S. Pacific Northwest illustrate and compare a sampling of methods, approaches, and progress. For example, landform mapping using high-resolution topographic data resulted in identification of about four times more landslides in Seattle, Washington, than previous efforts using aerial photography. Susceptibility classes based on the landforms captured 93 % of all historical landslides (all types) throughout the city. A deterministic model for rainfall infiltration and shallow landslide initiation, TRIGRS, was able to identify locations of 92 % of historical shallow landslides in southwest Seattle. The potentially unstable areas identified by TRIGRS occupied only 26 % of the slope areas steeper than 20°. Addition of an unsaturated infiltration model to TRIGRS expands the applicability of the model to areas of highly permeable soils. Replacement of the single cell, 1D factor of safety with a simple 3D method of columns improves accuracy of factor of safety predictions for both saturated and unsaturated infiltration models. A 3D deterministic model for large, deep landslides, SCOOPS, combined with a three-dimensional model for groundwater flow, successfully predicted instability in steep areas of permeable outwash sand and topographic reentrants. These locations are consistent with locations of large, deep, historically active landslides. For an area in Seattle, a composite of the three maps illustrates how maps produced by different approaches might be combined to assess overall landslide potential. Examples from Oregon, USA, illustrate how landform mapping and deterministic analysis for shallow landslide potential have been adapted into standardized methods for efficiently producing detailed landslide inventory and shallow landslide susceptibility maps that have consistent content and format statewide.