Using weather and climate information for landslide prevention and mitigation

Abstract

Landslides are significant natural hazards that degrade the productivity of soils, harm humans, and damage property. Extended and intense rainfall is the most common triggering mechanism of landslides worldwide. Sites are most susceptible to landsliding during wet antecedent conditions. Typically deep-seated, slow moving landslides (e.g., earthflows, slumps) are triggered or reactivated by an accumulation of precipitation over several days or weeks. In contrast, shallow, rapid landslides (debris avalanches, debris flows) usually initiate during individual intense or large storm events. Successfully predicting landslide hazards in large regions greatly depends on our ability to link meteorological conditions with various types and extents of slope failures. Four available methods for linking available weather and climate information to landslide initiation are discussed: (1) simple rainfall — landslide relationships; (2) multi-factor empirical assessment methods; (3) distributed, physically-based models; and (4) real-time warning systems. Each of these methods has certain strengths and weaknesses related to landslide hazard assessment. Of the land use practices that exacerbate landsliding, roads/trails and forest conversion to agriculture (typically associated with burning) exert the greatest impacts. Climate change scenarios that promote higher intensity storms, more rainfall, and vegetation with weaker root structure or less root biomass will likely increase landslide susceptibility; however, such impacts are currently speculative and will be difficult to unravel from anthropogenic effects.