EGU24-13255, updated on 09 Mar 2024
https://doi.org/10.5194/egusphere-egu24-13255
EGU General Assembly 2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Unifying endo-exo classification of episodic landslide movements

Qinghua Lei1 and Didier Sornette2
Qinghua Lei and Didier Sornette
  • 1Department of Earth Sciences, Uppsala University, Uppsala, Sweden (qinghua.lei@geo.uu.se)
  • 2Institute of Risk Analysis, Prediction and Management, Academy for Advanced Interdisciplinary Studies, Southern University of Science and Technology, Shenzhen, China (dsornette@ethz.ch)

Landslides, a widespread form of mass wasting, involve complex gravity-driven downslope movements developing over days to years before the final major collapse, which are commonly boosted by external events like precipitations and earthquakes. The reasons behind these episodic movements, characterised by alternating cycles of accelerating and decelerating creeps (marked by intermittent bursts of displacement followed by sustained periods of relaxation dynamics), and how these relate to the final instability, remain poorly understood. Here, we propose the new “endo-exo” classification of landslide bursts, based on the dynamical signatures of pre- and post-burst displacement rates. The underlying concept is based on the existence of cascades of triggered frictional slip and damage responses around a burst. The general theory of multiple cascades of triggered events predicts the existence of four classes of bursts: (i) exogenous non-critical, (ii) exogenous critical, (iii) endogenous non-critical, and (iv) endogenous critical, with respective displacement rates relaxing as power laws around the time tc of the burst respectively as (i) 1/(ttc)1+ϑ for t > tc, (ii) 1/(ttc)1–ϑ for t > tc, (iii) 1/ttc0, and (iv) 1/ttc1–2ϑ, thus depending on a single parameter ϑ. We test these predictions on the precursory and recovery signatures associated with bursts recorded in the long-term monitoring dataset of a rainfall-induced landslide at Preonzo, Switzerland, which exhibited significant episodic movements over many years prior to a catastrophic failure in 2012. Exogenous critical bursts (ii), provoked by external rainfall events, occur abruptly and relax gradually with a power-law exponent around 0.5. In contrast, for endogenous critical bursts (iv) that occur spontaneously under no external triggering, the landslide progressively accelerates prior to the burst and then slowly decelerates afterwards, showing a semi-symmetrical acceleration-deceleration behaviour governed by a small power-law exponent around 0.1. The longer-lived influence of an endogenous critical burst (as reflected by its small relaxation exponent) results from the precursory process that impregnates the system much more than its exogenous counterpart. Additionally, we document a unique exogeneous subcritical burst (i) triggered by the sudden collapse of a downslope sector; it is characterised by an immediate peak followed by a rapid recovery with a power-law exponent around 1.4, consistent with the absence of cascading failures. Endogenous non-critical bursts (iii) are largely driven by fluctuations and thus show no time-dependent recovery. The obtained power laws for these different burst classes are compatible with the existence of a single exponent ϑ ≈ 0.4±0.1, providing strong support for our theory. Our novel conceptual framework points at the existence of a deep quantitative relationship between episodic landslide movements, external triggering events (e.g. rainfall, snowmelt, and seismicity), and internal frictional slip, damage, and healing processes within the landmass.

How to cite: Lei, Q. and Sornette, D.: Unifying endo-exo classification of episodic landslide movements, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13255, https://doi.org/10.5194/egusphere-egu24-13255, 2024.

Supplementary materials

Supplementary material file

Comments on the supplementary material

AC: Author Comment | CC: Community Comment | Report abuse

supplementary materials version 1 – uploaded on 13 Apr 2024, no comments