Microseismic monitoring illuminates phases of slope failure in soft soils

Yfantis, G. and Pytharouli, S. and Lunn, R.J. and Carvajal, H.E.M. (2021) Microseismic monitoring illuminates phases of slope failure in soft soils. Engineering Geology, 280. 105940. ISSN 0013-7952 (https://doi.org/10.1016/j.enggeo.2020.105940)

[thumbnail of Yfantis-etal-EG-2020-Microseismic-monitoring-illuminates-phases-of-slope-failure]
Preview
 Text. Filename: Yfantis_etal_EG_2020_Microseismic_monitoring_illuminates_phases_of_slope_failure.pdf
Accepted Author Manuscript
License: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 logo
Download (3MB)| Preview

Abstract

The role of microseismic monitoring in rock slope stability has been long established: large microseismic events associated with rock failure can be detected by seismometers, even at distances of a few kilometres from the source. This is a favourable characteristic for the monitoring of mountainous areas prone to failure. We show that microseismic monitoring, using short-period arrays and a sufficiently high sampling rate, can also record weak precursory signals, that could represent early phases of a larger scale slope failure in soft soils. We validate this hypothesis with field observations. We find that, even in high attenuation material such as clays, it is possible to record and detect in the frequency domain, soil failures at source-to-receiver distances up to 10 m for crack formation/propagation to more than 43 m for small (less than 2.5 m3) events. Our results show for the first time, an extended frequency range (10 Hz to 380 Hz) where small soil failures can be detected at short monitoring distances, even at sites with high background noise levels. This is the first published study focusing on ground-truthed only, slope failure induced seismic signals in soft soils at field scale and within the seismic frequency range (1–500 Hz). We suggest that microseismic monitoring could complement existing monitoring techniques to characterize the response and structural integrity of earth structures, such as embankments, where the monitoring distances are a few 10s of metres, with the potential to detect any material deterioration at the very early stages. This study does not focus on automatic classification of slope failure signals, however, our observations and methodology could form the basis for the future development of such an approach.

CORE (COnnecting REpositories)