Stochastic hydro‑sedimentary modelling of arsenic mobilisation and downstream propagation in a coupled slope–channel–lake system

The transfer of sediment‑bound contaminants from unstable hillslopes into fluvial and lacustrine environments is governed by the interaction between geomorphic processes, hydrological connectivity and sediment transport dynamics. This study develops a quantitative modelling framework to assess the mobilisation of arsenic (As) from a contaminated slope and its potential downstream propagation through an integrated slope–channel–lake system.

High‑resolution terrain data were used to parameterise slope geometry and derive section‑scale morphometric attributes relevant to sediment detachment and mass‑failure susceptibility (slope gradient, contributing area, profile curvature and cross-sectional geometry). Spatially distributed As concentration measurements were incorporated into a stochastic Monte Carlo model, which simulated 10,000 realisations of contaminant mass for each slope section using distribution-specific sampling to represent data variability. Mobilizable sediment volumes were estimated using geometrically constrained maximum‑failure envelopes, enabling derivation of event‑scale sediment yields.

Hydrological and sediment connectivity were conceptualised through a simplified source–pathway–receptor model. Collapse scenarios representing 10%, 20%, 30%, 50% and 100% slope mobilisation were propagated downstream assuming full sediment transfer efficiency and no attenuation processes such as channel storage, hyporheic exchange, settling velocity effects or precipitation–adsorption dynamics. This approach represents an upper-bound transfer model suitable for preliminary contaminant‑risk assessment.

Total As mass stored in the slope was estimated at approximately 458 kg. Model outputs indicate that even under complete slope failure, the resulting concentration in the receiving lake remains marginally below the commonly adopted 0.010 mg/L threshold for potable water, whereas partial‑failure scenarios yield concentrations an order of magnitude lower. Sensitivity analyses demonstrate that predictions are strongly influenced by bulk density assumptions, connectivity ratios and sediment pulse magnitudes, highlighting the importance of probabilistic approaches for representing parameter uncertainty.

These findings underscore the need to integrate hydro‑sedimentary modelling, geomorphic characterisation and stochastic uncertainty quantification when assessing contaminant transport in catchment‑scale systems. The methodology presented provides a transferable framework for evaluating contaminant propagation where legacy mining residues persist in erosion‑prone, hydrologically connected terrain.