Fluvial geomorphic factors affecting liquefaction-induced lateral spreading

Abstract

Liquefaction-induced lateral displacements represent a major geohazard in earthquake-prone regions, yet the uncertainty associated with their prediction remains notoriously high. Documented observations after recent earthquakes provide evidence that depositional environment-specific geologic conditions play a crucial role in liquefaction susceptibility, and in the severity and spatial extent of liquefaction-induced ground deformations. However, this evidence is largely qualitative in nature, which limits the potential to incorporate the effects of depositional processes and environments in the next generation of lateral spreading predictive models. This study provides a framework to quantitatively assess the relationship between depositional environment-specific geologic factors and lateral spreading by means of simple fluvial geomorphic facies models, geotechnical engineering data (e.g. Cone Penetration Test data), and geospatial analytics. Three hypotheses are introduced and tested using lateral spreading ground deformations observed following the 2011 Christchurch earthquake along the Avon and Heathcote rivers in New Zealand. The results from this study indicate that the presence of an active (i.e. with active sediment deposition) compared to inactive (e.g. abandoned) channels is the most important fluvial geomorphologic variable out of the three tested. The other two are associated with the location relative to the meander bend position, including location within the point bar (inside) or the cut bank (outside), and upstream versus downstream within a given point bar. Findings from this study show that more lateral spreading occurs within point bars, and upstream (within a given point bar) in simple meander bends. However, the presence of geomorphic complexities (e.g. cut banks connected to an incised channel or tributary and/or channel confinement) can challenge the unbiased quantification of the contribution of a single geomorphic variable to the observed lateral displacements. These findings can be applied to other fluvial environments outside of New Zealand, and the proposed framework can be implemented for other non-fluvial depositional settings.