— The model, developed by researchers at the Swiss Federal Institute of Technology Zurich (ETHZ) and the Institute for Snow and Avalanche Research (SLF), proved its worth during the landslides in Brienz and Blatten. It enables more accurate forecasts to be made of the course, height and propagation of alpine mass movements such as snow, ice and rock avalanches.
— After a first conclusive practical test in 2023 on the village of Brienz (Graubunden), which had been evacuated because of a landslide threat, the researchers carried out simulations in Blatten last spring. Their aim was to unofficially test their forecasting model in a more complex and unstable scenario than that of Brienz. At Blatten, in addition to the rocks and water, the ice also played a role and the terrain is very complex. As in reality, the results of the 3D simulation showed that most of Blatten was destroyed and that the hamlet of Weissenried was barely spared by the mass of collapsed rock and ice.
— The model also showed that the mass of the fall had spread by 1.2 kilometres on the south-west side of the valley and by 700 metres on the north-east side — very accurate values compared with the actual disaster. These results stunned the researchers at the time, but seemed rather implausible.
— Avalanche and debris flow modelling tools are generally based on two-dimensional "medium-depth" methods. They assume that the flow of rock and water is shallow and remains in constant contact with the ground, resulting in continuous friction. But the 3D model allows the particles to detach themselves from the surface, reducing friction on the ground and enabling the phases during which the material flies through the air to be accurately captured. This is a decisive factor in simulating the flow behaviour and propagation of flows in steep or complex terrain.
— Swissinfo: In the case of Blatten, the scientific simulations have not been shared with the Valais authorities, and are not part of the official studies currently under way, nor of risk management. The researchers' aim is not to replace existing 2D tools, but to offer a complementary solution when conventional models reach their limits.