Predicting movement using internal deformation dynamics of a landslide in permafrost

Abstract

Within the Brooks Range of Alaska, several frozen debris lobes (FDLs) – or slow-moving landslides in permafrost – are approaching critical infrastructure. FDL-A, the largest and closest of these features, was 32.3m from the toe of the Dalton Highway as of October 2016. Here we present the analysis of nearly three years of data from a MEMS-based in-place inclinometer installed within FDL-A. Analysis of the strain within the active layer indicates that it is closely tied to air temperature and water, suggesting that cooling and/or draining the FDL may be effective mitigation techniques. Within the lobe body, strain rates are comparable to those measured within rock glaciers. Using the cyclical pattern and phase lag identified within the strain data, and surface movement rates 1) derived from historic imagery analysis and InSAR data, and 2) measured using a differential GPS system, we developed and vetted a predictive function for surface movement of FDL-A. The predictive function indicates that FDL-A moves at an average rate of 4.9myr−1, reaching a maximum velocity of 7.9myr−1 during the fall, and falling to 1.9myr−1 in the early spring. We hypothesize that the seasonal movement pattern is due to the infiltration of snow melt, and the subsequent reduction of effective stress within the shear zone. Using the predictive function and the measured October 2016 distance, we predict that FDL-A will reach the current Dalton Highway embankment during early 2023.