On the role of vegetation in earthquake induced landslides

Breakage of vegetation (grasses, shrubs and trees) has been observed in past earthquake induced landslide events. But the role of vegetation during such occurrences has generally been overlooked due to the difficulty in quantifying the magnitude of the vegetation effect alone. In this paper, dynamic centrifuge testing has been conducted to investigate the seismic performance of a slope planted with vegetation. In the centrifuge modelling, 3-D printing of layered ABS plastic was used to produce repeatable root analogues which are highly representative of the geometry and mechanical behaviour of real woody root systems. The earthquake sequence was composed of three distinct motions to investigate the behaviour of the vegetated slopes under a sequence of strong motions (without intermediate reinstatement, e.g. aftershocks) and to provide a richer dataset concerning the dynamic response within the slope (e.g. more information on topographic amplification). It is shown that vegetated slopes performed better than the benchmark fallow slopes during earthquakes, especially in terms of the crest deformation response, which is a key parameter in performance-based slope assessment and design. However, vegetation had a very limited influence on the general propagation and amplification of the dynamic earthquake motion from the toe to the crest of the slope. Test results were then compared with other existing studies which considered slopes with similar geometry but reinforced by more traditional techniques (e.g. piles).