ShallowLandslider: a hybrid Landlab component for predicting regional distributions of coseismic landslides

Earthquakes in mountainous regions can trigger tens of thousands of shallow landslides, reshaping hillslopes and amplifying disaster impacts. Predicting their spatial distribution remains challenging because most existing models are empirical, event-specific, and lack physical interpretability. We present ShallowLandslider, a new component within the open-source Landlab framework that integrates deterministic mechanics with probabilistic and empirical elements for regional-scale prediction of coseismic landslides. The model extends the classical Newmark sliding block approach to three dimensions, incorporating transient seismic accelerations, slope geometry, and variable properties of mobile regolith, such as cohesion, internal friction, and moisture content, on structured grids. Instability is assessed using critical acceleration thresholds, complemented by a probabilistic selection scheme to represent natural variability in failure occurrence. To improve geometric realism, the component partitions unstable regions using an empirical distribution of observed landslide length-width ratios from the study area.
We validate ShallowLandslider against landslide inventories from two subregions affected by the 2015 Mw 7.8 Gorkha earthquake in Nepal.  Performance is evaluated using distributional metrics across landslide area, elevation, slope, and aspect. Results highlight that mobile regolith depth, parameterised by local elevation and planform curvature, strongly controls predicted landslide distributions and size. In addition, moderate cohesion values (10–15 kPa) proved critical for matching observed clustering of landslides on hillslopes and limiting unrealistically large failures. While pixel-level prediction remains impractical, the model captures first-order spatial and statistical patterns of coseismic landsliding, offering a reproducible, physically grounded tool for hazard assessment. Its modular design enables coupling with other Earth-surface process models, paving the way for integrated simulations of landscape response to seismic forcing and cascading hazards. We are extending ShallowLandslider beyond earthquake-specific triggers to support rainfall-induced failures, creating a multi-trigger framework that also links fault mechanics and slope stability through coupling with a dynamic rupture model. These developments aim to enable more holistic simulations of shallow landslide distributions and support next-generation approaches for regional and global landslide risk assessment.