The research, led by Gunnel Göransson and colleagues at the Swedish Geotechnical Institute, examines the Göta River, the Nors River, the Säve River and the Ångerman River, all of which flow through fine-grained or mixed sediments where landslide susceptibility is high. The five case studies, conducted between 2009 and 2022, were carried out within a national programme for assessing and mapping future landslide hazards along watercourses. Together they form one of the most sustained efforts anywhere in the world to connect climate-driven fluvial erosion with probabilistic geotechnical slope-stability analysis.

The methodological framework that emerged was iterative, refined case by case, but consistently followed seven steps: compiling previous studies and measurements, conducting hydroacoustic surveys and sediment investigations, running hydrodynamic models to derive erosion parameters, selecting future flow scenarios, modelling erosion, validating results and assessing uncertainty, and finally integrating erosion forecasts into geotechnical stability analyses. Multibeam echosounder surveys mapped bathymetry, side-scan sonar characterized bedforms, backscatter analysis classified sediments, and physical sampling determined how erodible the riverbed materials actually were.

Hydrodynamic modelling sat at the heart of each assessment. The teams used two-dimensional models, including Delft3D, TELEMAC-2D, MIKE 21 C and MIKE 21 FM, to simulate water levels, flow velocities and bed shear stresses under a range of hydrological conditions. Two-dimensional modelling was deliberately chosen as the best balance between computational demand, data requirements and accuracy; three-dimensional simulation, while more detailed, was not considered justified given the uncertainties inherent in forecasting erosion over nearly a century. Erosion estimates rested on the relationship between calculated bed shear stresses, critical shear-stress thresholds and sediment erodibility coefficients, with cohesive sediments handled through the Partheniades formulation within a GIS framework.

The choice of erosion model depended on the geology. In clay-dominated systems such as the Göta, Nors and Säve rivers, the GIS-based analyses were supplemented with the Bank Stability and Toe Erosion Model, known as BSTEM, which explicitly couples hydraulic toe erosion with geotechnical bank-failure mechanisms. The Ångerman River, by contrast, is dominated by frictional sediments such as silt, sand and gravel, so the team adopted a full morphodynamic approach using the MIKE 21 C multi-fraction model, representing sediment transport and channel evolution explicitly. Future flows were derived from downscaled RCP climate projections supplied by the Swedish Meteorological and Hydrological Institute, alongside alternative hydropower regulation strategies.

The erosion projections were then translated into future channel cross-sections that served as direct input to slope-stability calculations performed with Slope/W, following Swedish geotechnical practice. Both present-day and year-2100 conditions were analysed, accounting for projected changes in groundwater, pore-water pressures and erosion-modified slope geometry. To capture uncertainty, the team applied the Point Estimate Method, a computationally efficient alternative to Monte Carlo simulation, deriving failure probabilities from statistical distributions of shear strength, unit weight, pore pressure and geometry. These probabilities were classified into five classes and combined with consequence classes in a GIS-based risk matrix to produce landslide risk maps sensitive to climate change.

Among the most striking findings is the role of hydropower regulation. In heavily regulated rivers, erosion driven by operational flow management can exceed that caused by climate-related changes in discharge, potentially obscuring the climate signal altogether. On the Ångerman River, short-term regulation generated rapid fluctuations in water level and flow velocity that dwarfed projected climate-driven hydrological changes. The study also revealed that the temporal resolution of discharge data matters enormously: simulations based on 14-day averaged flows suggested reduced future erosion, while high-resolution data capturing short-duration flow peaks indicated the opposite, because those peaks contribute disproportionately to bed shear stress but vanish when flows are averaged.

The synthesis also clarified how erosion interacts with inherent susceptibility. In areas already highly prone to landslides, even minor erosion can significantly raise the probability of a catastrophic failure, whereas in stable terrain substantial erosion is needed before risk begins to climb. Sediment composition determines which climate-related driver dominates: in cohesive clay valleys such as those of the Göta, Nors and Säve rivers, fluvial erosion is the dominant climate-related trigger of slope instability, while in the coarser deposits of the Ångerman valley, changes in groundwater and pore-water pressure are likely to matter more.

The authors are candid about uncertainty. Scarcity of sediment-transport measurements, limited repeated bathymetric surveys, positional inaccuracies on steep underwater slopes and the sheer difficulty of simulating long-term fluvial geomorphology all constrain quantitative precision. Their response is a call for recurrent, systematic monitoring, particularly comprehensive bathymetric surveys, and for assessments that are treated as living documents, updated regularly and immediately after any landslide occurs, since such an event would fundamentally reshape river morphology. The projections are best read as semi-quantitative tools for identifying erosion-prone areas and prioritizing detailed investigations rather than as precise predictions.

Perhaps the most valuable export of the study is its transferable workflow. Because it is grounded in fundamental principles of hydrology, hydraulics, sediment transport and erosion, the framework can be adapted beyond Sweden by adjusting the hydrological forcing, whether the driver is monsoon rainfall, tropical cyclones, drought or rapid glacier retreat. The authors emphasize that success depends on interdisciplinary collaboration across geomorphology, hydrology, sediment transport, hydraulics and geotechnics, and on adaptive rather than static management. As extreme precipitation intensifies worldwide, the lesson from the Swedish rivers is clear: the ground beneath riverside communities is being reshaped now, and the only responsible way to plan for it is to model, monitor and adapt continuously.

Subject of Research: Projecting climate-induced river erosion to assess future landslide susceptibility in Swedish river valleys

Article Title: Projecting climate-induced river erosion for assessing future landslide susceptibility: methodological insights and lessons learned from five Swedish cases

Article References: Göransson, G., Odén, K., Bergdahl, K., & Bolin, P. (2026). Projecting climate-induced river erosion for assessing future landslide susceptibility: methodological insights and lessons learned from five Swedish cases. Environmental Earth Sciences, 85(15), Article 396. https://doi.org/10.1007/s12665-026-13128-4

Image Credits: AI Generated

DOI: 10.1007/s12665-026-13128-4

Keywords: river erosion, landslide susceptibility, climate change, slope stability, hydrodynamic modelling, sediment transport, bathymetry, hydropower regulation, geotechnical risk, Sweden, Göta River, morphodynamics