Key points are not available for this paper at this time.
Landslide hazard in seismically active mountainous regions is a dynamic variable influenced not only by seismic intensity but also by the structural setting of slopes and by the modes of energy release from seismic sources. It is further conditioned by the spatial distribution of past slope instabilities recorded in the region's seismic history. This study thoroughly explores this complex interaction, proposing a novel framework for near-real-time forecasting of rockfall hazard, tested on the 2022 Ms. 6.1 Lushan earthquake. The approach integrates seismic anisotropies in wave propagation and polarization with topographic amplification and introduces a new conceptual model of pre-induced resilience, emerging from the interplay between fragility and antifragility as phenomena controlling residual resilience. Back-analysis of observed rockfalls, avalanches, and flows distributions reveals that sectors previously affected by slope instabilities, particularly following the 2013 Ms. 7.0 event, exhibit signs of induced antifragility, which in some cases has limited or even prevented the triggering of new failures during the 2022 event. In contrast, areas with fault-parallel slope orientations and transverse topographic amplification display a significant concentration of large-scale failures, especially where long-wavelength seismic loads interact with deep fractured zones. These findings underscore the key role of dynamic shear transmission and structural slope conditions in modulating landslide hazard. The proposed framework offers a physics-based, scalable method for forecasting slope failures in seismically active mountain regions, with strong implications for early warning systems and risk mitigation strategies. • Seismic anisotropy by directionality and polarization can be relevant to Specific Hazards. • A novel hazard framework, based on fragility-antifragility, models the rockfall distribution. • Fragility-antifragility is introduced as a dynamic concept of adaptive response. • The computational framework is validated on the 2022 Ms. 6.1 Lushan earthquake.
Luo et al. (Sat,) studied this question.