Inside Switzerland’s giant 3D model for predicting landslides

A 5.4-metre model of Blatten is giving Swiss researchers a controlled way to recreate mountain mass movements. In a former military bunker near Davos, the WSL Institute for Snow and Avalanche Research SLF has assembled a detailed physical copy of the Valais landscape at a scale of 1:577. The work aims to improve the computer models that cantonal and municipal authorities will use when assessing risks to homes, roads and other infrastructure.

The timing is significant for an Alpine country where settlements often occupy narrow valleys and steep slopes. Blatten has become a focus of scientific attention after the region’s major collapse, which researchers are still working to understand. The new model cannot reproduce every geological process in the mountain environment, but it allows scientists to isolate how a moving mixture behaves as it crosses a known terrain.

The SLF describes the project as a long-term investment in research. Its findings are intended to strengthen hazard assessments, giving public authorities better evidence when they review land-use plans, protective measures and infrastructure exposure in vulnerable areas. The experiments were announced on September 8, 2026.

Fifty-six printed sections make up the model, which took around 100 days to produce. The components reproduce the landscape around Blatten in a form large enough for researchers to observe moving material across slopes, channels and changing contours. The physical scale also gives the team room to place instruments over the model and track the movement in detail.

The model is housed near Davos, far from the Valais terrain it represents. That separation turns a complex mountain environment into a repeatable test system. Researchers can introduce the same mixture under controlled conditions, adjust the setup and compare the resulting flow paths and deposits. The approach complements computer simulations by supplying measurements against which those models can be checked.

The SLF says the model represents a long-term research investment. Its value lies in repeated observation rather than a single demonstration. Each test can generate information about how material accelerates, spreads and settles as it moves through a miniature version of the landscape. Those observations can help researchers identify where digital forecasts reproduce real behaviour and where they need refinement. The result is a bridge between detailed terrain data and the decisions made by authorities responsible for safety in Alpine communities.

Water, sand and clay become a laboratory landslide when researchers channel them over the miniature terrain. The mixture is designed to create an observable mass movement across the model. Cameras, lasers and sensors then measure how quickly it travels, how much material passes through a given point and where the moving sediment is deposited.

These measurements allow the team to study the flow as it interacts with the landscape. A narrow channel may guide material, while a bend or a change in slope can alter its speed and spread. The researchers are paying particular attention to the shape of the ground. “We are currently particularly interested in the influence of terrain curvature on flow dynamics,” said Johan Gaume, head of the SLF’s Alpine Landslides research group.

Physical experiments offer a direct record of movement that can be compared with computer predictions. The tests do not recreate every trigger or underground condition involved in a real Alpine collapse. They focus on the behaviour of flowing material once it moves across the terrain. By measuring speed, discharge and deposits together, the SLF can examine how small changes in topography affect the reach and intensity of a mass movement.

The model’s practical test will come when its measurements reach planning offices in Valais and other Alpine cantons. The SLF says the experiments are intended to improve the reliability of computer models used by cantonal and municipal authorities. Those tools help officials assess potential threats to settlements and infrastructure, including the areas where moving material may travel and accumulate.

Better modelling could inform decisions on where construction is permitted, which roads and facilities require protection, and how authorities prioritise monitoring. It could also help planners compare the exposure of existing settlements with the risks faced by new projects. The model will not decide those questions. It will provide more detailed evidence for officials working within Switzerland’s established system of local and cantonal responsibility.

The wider Alpine context remains complex. A related Swissinfo report noted that small landslides have increased in the Swiss Alps as the climate changes, while the link between warming and large natural disasters is less straightforward. Researchers studying Blatten are also examining how the collapse occurred and how similar events might be anticipated. The new physical model adds one experimental tool to that effort, with results that could influence future hazard maps and infrastructure planning.