RECONSTRUCTION OF COMPOSITE LANDSLIDES IN FLAT-LYING SEDIMENTARY STRATA USING BALANCED CROSS SECTIONS

The Colorado Plateau region of the United States contains numerous examples of composite megalandslides with basal slip surfaces developed in overconsolidated (OC) shales. The largest identified slides displace over 850 m of strata and slides exceeding 600 m are common in the Grand Canyon. Although some of these slides have been noted in the literature, none have been evaluated in sufficient detail to explain their triggering, or kinematic mechanisms. Simple saturation of these shales engenders two-thirdsreduction of their intrinsic cohesion. Clastic dikes permeating the zones of displacement are suggestive of undrained conditions, fostering excess pore pressures and rapid movement, which would also reduce effective stress and shear strength. Earthquake shaking and rapid drawdown conditions might also contribute to destabilize slopes floored in shale.

The megalandslides in the Colorado Plateau appear to evolve through a progressive failure mechanism, likely ascribable to strain softening. Strain incompatibility between adjoining units of contrasting stiffness also appears to play a significant role in favouring specific stratigraphic horizons, because strain accumulation promotes strength loss through concentration of shear strain, to residual strength conditions.

We have been developing kinematic models using the principle of balanced cross sections to re-create the likely genesis of these landslides, from their initiation to final form as we view them today. The modelling of these features takes a dozen or more individual cross sections, depending on their complexity. These 2D cross sections account for stratigraphy and illustrate multiple sequences of movement within the various blocks of the landslide complexes. They suggest that toe removal and reactivation are repeatedly triggered by river erosion. Like tectonic reconstruction, these models work backwards using existing topography, with the aim of re-creating a cliff profile similar to those of unfailed slopes near the slides being modelled. This information is crucial to the accuracy of numerical analysis methods, including finite element, finite difference, and discontinuous deformation analysis (DDA). This work has concentrated on landslides in the central Grand Canyon, but can be applied anywhere similar conditions exist.