Enhanced landslide mobility promoted by liquefaction of underlying sediments: Evidence from detailed field, lab, and modelling investigations of the deadly Oso, USA landslide

Enhanced landslide mobility can project devastation across extensive areas, greatly affecting hazard and risk. Despite this importance, assessing potential mobility can be challenging as underlying causes of enhanced mobility vary. Liquefaction can dramatically decrease shear resistance and promote mobility, and pervasive liquefaction is well known to boost the mobility of debris flows and other flow slides. However, liquefaction’s potential effect on more coherent slide masses can be difficult to identify in the field. The 2014 Oso, Washington (USA) debris avalanche provides an exceptional opportunity to understand specific causes of liquefaction and enhanced mobility. The slide was more mobile than typical debris avalanches, sweeping over 1 km across a flat alluvial plain to the opposite side of the river valley and killing 43 people as it travelled. Following the 2014 event, we performed detailed investigations aimed at illuminating the event sequence and the mechanisms promoting mobility, with a strong focus on the role of liquefaction.

The landslide initiated in stratified glacial materials and created a variety of landslide deposit types, including a widespread debris-avalanche hummock field covering much of the formerly flat river valley. Our field investigations revealed clear and widespread evidence for sub-bottom (basal) liquefaction as the cause for the slide’s long reach. Soon after the slide event, we mapped more than 350 sand boils – classic indicators of liquefaction – as both isolated vents and groups of multiple vents within the hummock field. We found sand boils in the depressions between hummocks; the hummocks themselves were not liquefied and commonly contained rafted materials such as intact pieces of glacial stratigraphy and forest floor on their surfaces. The sand boils erupted through a variety of glacial sediments, including lacustrine clays. Sand boil grain-size characteristics most closely matched the underlying alluvial sands, rather than the overriding glacial sediments. Evidence of sand boils was transient; most features were eroded from the landscape within a year.

Liquefaction can be induced by several mechanisms, including rapid loading, shearing of loose contractive sediment, and cyclical loading during ground shaking. Given these plausible mechanisms, we used a fully coupled fluid-sediment elastic deformation analysis, as well as triaxial geotechnical testing of the alluvium, to assess potential liquefaction of the materials overrun by the Oso slide. Our results demonstrate that the large failure rapidly loading loose, already wet alluvial sediments likely resulted in their liquefaction. The greatly reduced shear strength of the liquefied alluvium enabled enhanced mobility of the overriding landslide mass on a liquefied base. This process differs from liquefaction of the slide material itself and is therefore not directly dependent on slide-mass properties. Liquefaction of underlying sediments, similar to that observed at Oso, may have enhanced the mobility of other large, coherent landslides in Europe and Asia.