A machine learning–based approach to regional‐scale mapping of sensitive glaciomarine clay combining airborne electromagnetics and geotechnical data

Abstract

Sensitive glaciomarine clays, often referred to as ‘quick clay’, commonly occur in many countries at high, northerly latitudes, causing frequent and occasionally devastating landslides. The salt content of quick clay is strongly correlated to both its shear strength and electrical resistivity. Hence, it can be mapped using electromagnetic methods more efficiently than traditional intrusive methods, the latter of which can often be slow and costly. However, the resistivity signature of quick clay is non‐unique, leading to ambiguous, imprecise interpretations of geophysical models. In this study, we present an improved method for predicting the probability of quick clay using airborne electromagnetics. Using machine learning algorithms, we combine geophysical models with geotechnical data to address the issue of their non‐unique resistivity signature. Beyond resistivity values, the machine learning algorithms use spatial derivatives of resistivity and spatial attributes. We evaluate the performance of this method using data collected from a road construction project in central Norway. Results show that this method is able to make plausible and accurate predictions of quick clay occurrence using as few as 10 boreholes across an area of 14.8 km2, and that it outperforms a simple interpretation based on resistivity intervals alone. In addition to a ‘best guess’ categorical classification, these algorithms output probability estimates, and we demonstrate that they are a reliable indication of uncertainty. The accuracy of these predictions also tends to increase as more geotechnical data are included as training data, helping compensate for the limited resolution of the airborne electromagnetics data. Given that the petrophysics of the clays at this test site are consistent with observations in other regions, we expect this method has the potential to make quick clay hazard mapping more efficient by offering valuable early‐phase insights, leading to large time and cost savings for both infrastructure planning and regional hazard mapping.

Three‐Dimensional Probabilistic Hydrofacies Modeling Using Machine Learning

“…Jia et al (2021), also incorporated location coordinates in machine learning for 3D geological modeling with geological-geophysical data sets. Including location data has been shown to enhance machine learning predictions (C. W. Christensen et al, 2021;Gottschalk & Knight, 2022).…”

“…Machine learning has been used to remove noise (X. Wu et al, 2020) and process AEM raw data (Asif et al, 2022), conduct geophysical inversions (S. Wu et al, 2022Wu et al, , 2023a, interpret AEM inversions (Haber et al, 2019), simulate AEM response (S. Wu et al, 2023b), model glacial till using electrical conductivity derived from AEM (Gunnink et al, 2012), map quick clay using AEM (C. W. Christensen et al, 2021), construct field-scale rock-physics transform and simulate AEM (Gottschalk & Knight, 2022), and to cluster AEM (Dumont et al, 2018). Friedel et al (2016) and Friedel (2016) used machine learning to estimate aquifer distributions and hydrostratigraphic units using AEM, borehole and hydrogeological data.…”

AEM in Norway: A Review of the Coverage, Applications and the State of Technology

“…We use a two-step modelling approach introduced by [47] and refined by [48] to predict the probability of quick clay. In the first step, laboratory tests, wherein remoulded shear strength has been measured directly, is used to train a random forest classifier interpret the presence of quick clay at geotechnical sounding locations.…”

“…These advantages can explain why the use of AEM in Norway has expanded in recent decades from mineral exploration to more engineering and environmental applications. While AEM still continues to be used for mineral exploration (e.g., [6,7,64]), both TEM systems [9,48] and FEM systems [10,11] have been successful in mapping quick clay occurrence. In fact, both ground-based and airborne resistivity imaging methods are now recommended as tools for large-scale quick clay mapping [65].…”

“…Using high-resolution intrusive investigations helps one recover detailed features in the subsurface. Other authors have noted that machine learning algorithms achieve this largely by way of the spatial coordinates [43,48], emphasizing the need for local ground-truth data. Beyond the correlative analyses and machine learning algorithms explored in this paper's two case studies, other integrated modelling approaches used outside of Norway include using multiple-point statistics for stochastic stratigraphic modelling [67][68][69], and local neighbourhood statistical analysis [69].…”