Landslides exhibit intermittent gravity-driven downslope movements developing over days to years before a possible major collapse, commonly boosted by external events like precipitations and earthquakes. The reasons behind these episodic movements and how they relate to the final instability remain poorly understood. Here, we develop a novel “endo-exo” theory to quantitatively diagnose landslide dynamics, capturing the interplay between exogenous stressors such as rainfall and endogenous damage/healing processes. We predict four distinct types of episodic landslide dynamics (endogenous/exogenous-subcritical/critical), characterized by power law relaxations with different exponents, all related to a single parameter ϑ . These predictions are tested on the dataset of the Preonzo landslide, which exhibited multi-year episodic movements prior to a catastrophic collapse. All its episodic activities can be accounted for within this classification with  ϑ≈0.45±0.1 , providing strong support for our parsimonious theory. We find that the final collapse of this landslide is clearly preceded over 1-2 months by an increased frequency of medium/large velocities, signaling the transition into a catastrophic regime with amplifying positive feedbacks.

Fundamentally related to the ultraviolet (UV) divergence problem in Physics, conventional wisdom in seismology is that the smallest earthquakes, which are numerous and often go undetected, dominate the triggering of major earthquakes, making accurate forecasting of the latter difficult if not inherently impossible. Using the general class of epidemic type aftershock sequence (ETAS) models and rigorous pseudo-prospective experiments, we show that ETAS models that feature a specific magnitude correlation between triggered and triggering earthquakes and a magnitude-dependent Omori kernel, significantly outperform simpler ETAS models, in which these features are absent. Using the best forecasting model, we then show that large earthquakes are preferentially triggered by large events. These findings have far-reaching implications for short-term and medium-term seismic risk assessment, as well as for the development of a deeper theory without UV cut-off that is locally self-similar.

We present numerical simulations of elastic wave propagation, scattering and attenuation in two-dimensional fractured media. Natural fracture systems following a power law length scaling are modeled by the discrete fracture network approach for the geometrical representation of fracture distributions and the displacement discontinuity method for the mechanical computation of fracture-wave interactions. The model is validated against analytical solutions for wave reflection, transmission and scattering by single fractures, after which we apply it to solve the spatiotemporal wavefield evolution in various synthetic fracture networks. We find that the dimensionless angular frequency ῶ plays a crucial role in governing wave transport. When ῶ is smaller than the critical frequency ῶc (≈ 5), waves are in the extended mode, either propagating (for small ῶ) or diffusing by multiple scattering (for intermediate ῶ); as ῶ exceeds ῶc, the wave energy becomes trapped, entering either the Anderson localization regime (kl* ≈ 1) in well-connected fracture systems or the weak localization regime (kl* > 1) in poorly-connected fracture systems, where k is the incident wavenumber and l* is the mean free path length. Consequently, the inverse quality factor Q-1 scales with ῶ obeying a two-branch power law dependence, showing significant frequency dependence when ῶ < ῶc and almost frequency independence when ῶ > ῶc. In addition, when ῶ < ῶc, the wavefield exhibits a weak dependence on fracture network geometry, whereas when ῶ > ῶc, the fracture network connectivity has an important impact on the wave behavior such that strong attenuation occurs in well-connected fracture systems.