Eliciting the influence of molehills on the stability of river dikes with structured expert judgment

Abstract

Earth embankments (e.g., dams, dikes and levees) are critical components of flood defense systems, providing protection against riverine and coastal inundation. The integrity of these structures can be compromised by biotic factors such as burrowing activity from rodents, foxes, and other animals, whereby excavation of the burrows can negatively impact the structure by altering seepage paths and creating defects in the surface that lead to higher turbulence during overflow events and exacerbating pore pressure. Therefore, burrowing activity can contribute to failure of these structures through mechanisms such as slope instability or localized erosion. Despite being a recognized risk, incorporating the influence of animal burrows in the design phase is difficult because of two main reasons: (1) no standardized methodologies exist, to the authors’ knowledge, to translate burrow activity into engineering parameters (variables), and (2) numerical methods that could be used to perform a conventional geotechnical design require quantifiable parameters such as hydraulic conductivity, or geometry of discontinuities, which are typically unknown or impractical to measure in the field. In cases where animal burrows are present, these parameters are not easy to measure or predict due to their high variability and limited literature on the topic. In this study, Structured Expert Judgement using Cooke’s method is explored as a tool to quantify geotechnical parameters to incorporate burrowing activity in embankment design, focusing on the case of mole activity. The questionnaire includes calibration questions covering both the activity of moles during the mating season (quantified as density measured on the ground surface), and geotechnical parameters of soil. The questions of interest cover parameters necessary for assessing the influence of molehills on instability and overtopping erosion of embankments. The experts are asked to quantify those parameters without the presence of molehills and with a given density of molehills to allow comparison.