Using computational fluid dynamics, we analyze the error trends of an analytical shear stress distribution model used to drive aeolian transport for coastal dunes, which are an important line of defense against storm-related flooding hazards. We find that compared to numerical simulations, the analytical model results in a net overprediction of the landward migration rate. Additionally, two data-driven approaches are proposed for reducing the error while maintaining computational efficiency.

Using a numerical model in 2D, we study how two identical dunes interact with each other when exposed to reversing winds. Depending on the distance between the dunes, they either repel or attract each other until they reach an equilibrium distance, which is controlled by the wind strength, wind reversal frequency, and dune size. This process is controlled by the modification of wind flow over dunes of various shapes, influencing the sediment transport downstream.

Our study explores permafrost–glacier interactions with a focus on their implications for preparing or triggering high-volume rock slope failures. Using the Bliggspitze rock slide as a case study, we demonstrate a new type of rock slope failure mechanism triggered by the uplift of the cold–warm dividing line in polythermal alpine glaciers, a widespread and currently under-explored phenomenon in alpine environments worldwide.

The erosion of rocky coasts occurs episodically through wave action and landslides, constituting a major natural hazard. Documenting the factors that control the coastal retreat rate over millennia is fundamental to evidencing any change in time. However, the known rates to date are essentially representative of the last few decades. Here, we present a new method using the concentration of an isotope, ¹⁰Be, in sediment eroded from the cliff to quantify its retreat rate averaged over millennia. 

We investigated the dynamic behavior of a rock arch to understand how fractures influence its stability. Using geophones, we measured its modes of vibration and used numerical modeling to replicate them. We found that higher-order resonance modes are the most sensitive to fractures, indicating their potential as early indicators of structural damage. Therefore, monitoring these higher-order modes could provide a more accurate tool to assess the structural integrity of natural rock landforms.

Channel belts comprise the area affected by a river due to lateral migration and floods. As a landform, they affect water resources and flood hazard, and they often host unique ecological communities. We develop a model describing the evolution of channel-belt area over time. The model connects the behaviour of the river to the evolution of the channel belt over a timescale of centuries. A comparison to selected data from experiments and real river systems verifies the random walk approach.

We have developed a new numerical model to represent the formation of duricrusts, which are hard mineral layers found in soils and at the surface of the Earth. We assume that the formation mechanism implies variations in the height of the water table and that the hardening rate is proportional to precipitation. The model allows us to quantify the potential feedbacks they generate on the surface topography and the thickness of the regolith/soil layer.

Short-lived, high-magnitude flood events across high mountain regions leave substantial geomorphic imprints, which are frequently triggered by excess precipitation, glacial lake outbursts, and natural dam breaches. These catastrophic floods highlight the importance of understanding the complex interaction between climatic, hydrological, and geological forces in bedrock catchments. Extreme floods can have long-term geomorphic consequences on river morphology and fluvial processes.

This study presents a novel convolutional-neural-network approach for detecting instream large wood in rivers, addressing the need for flexible monitoring methods across diverse data sources. Using a database of 15 228 fully labelled images, the model achieved a weighted mean average precision of 67 %. Fine-tuning parameters and sampling techniques can improve performance by over 10 % in some cases, offering valuable insights into ecosystem management.

Analyzing the geometry and the dynamics of riverine bedforms (so-called dune tracking) is important for various fields of application and contributes to sound and efficient river and sediment management. We developed a workflow that enables a robust estimation of bedform characteristics and with which comprehensive sensitivity analyses can be carried out. Using a field dataset, we show that the setting of input parameters in bedform analyses can have a significant impact on the results.