Rapid characterization of landslide hydrogeology through simple clustering of seismic refraction and electrical resistivity surveys

Abstract:

Landslide hazards pose risks to human life and economic security across the world. They are often one hazard in a chain of cascading multihazards, prone to triggering by earthquakes or erratic rainfall patterns, with potential to dam downslope waterways which can in turn cause outburst floods. The future timings of earthquakes and extreme rainfall events can be difficult to predict, but understanding the spatial distribution of key subsurface characteristics in an unstable slope can inform stability assessments, providing a basis for planning mitigation measures against future failure. These key subsurface characteristics include information on the lithological properties, hydrological setting and subsurface discontinuities found within a slope. Geophysical surveys provide a means of assessing these key landslide characteristics in a rapid and non-intrusive manner, and inverted geophysical models provide proxies of subsurface properties that can inform slope stability modelling and risk assessment.

Here, we present the results of rapid reconnaissance fieldwork, composed of P- and S-wave seismic refraction tomography (SRT) surveys and an electrical resistivity tomography (ERT) survey, at a slow-moving landslide in North Yorkshire, UK. The aim of these surveys was to rapidly acquire geophysical field data and apply integrated processing and data classification methods in order to elucidate hydrogeophysical patterns and trends that can aid in identifying the key subsurface characteristics of the landslide. These include, (i) co-registration of the disparate survey measurements to a common subsurface grid for improved analysis of subsurface data distributions, (ii) separate SRT and ERT inversions on this common grid using pyGIMLi, and (iii) the application of clustering methods to produce an objective, semi-automated ground model that requires minimal input from an operator. The results provide an overview of the distribution of hydrogeophysical properties in the landslide subsurface that can be linked to the predominant hydrogeological units known to form the basis of the unstable slope.