Master & SIE Projects

Description:

• Helmut will be the future French/Swiss flagship land surface model and is currently developed between CEN (MeteoFrance) and SLF Davos. The numerical core has been developed and the work will be to execute first model simulations and compare to the state of the art model SNOWPACK. Work will be at SLF Davos.

Description:

• The rapid change of the Antartic Ice Sheet due to global warming has strong implications to the global climate system. In order to better understand the dynamics in Antartica, CRYOS set up a tower equipped with several sensors at the Princess Elisabeth Station (PE), East Antartica. The main objective of this proposal is to use measurements from the 2024 campaign to investigate how turbulent structures in the atmospheric boundary layer are affected by stability, wind direction and blowing snow.
• In addition, we aim to add the effect of stability into the empirical formulation of Davenport for the analysis of vertical coherence in the atmospheric boundary layer.

Description:

• Documentation of geophysical models is often lacking, with model updates outpacing the speed of documentation updates, or legacy code remaining undocumented. In this thesis, the ability of a Generative AI to automatically translate code changes into documentation changes will be explored.
• The student will train an existing large language model (LLM) on the code base of the atmospheric model HICAR to serve as an assistant to both model users and developers. This LLM can then answer questions about model use or structure using natural language, improving the user/developer experience. Ultimately the student will test the accuracy of the LLM relative to a human and existing, generalized agentic AI coding assistants.
• This work will explore a novel method of code documentation which promises to free up developer time and provide users with more up-to-date information. The possibility of gathering additional model training data from model-user interactions may also be explored if time permits.
• This thesis topic is flexible in its implementation, and motivated students are encouraged to come to the project with their own ideas.

Description:

The rapid change of the Antarctic Ice Sheet due to global warming has strong implications to the global climate system. In order to better understand the dynamics in Antarctica, CRYOS group set up a tower at the Princess Elisabeth Station (PE), East Antarctica, equipped with sensors to capture meteorological variables which are analyzed in terms of wind direction, blowing snow events and stability. It was demonstrated that the Monin-Obukov similarity theory parametrization underestimates latent and sensible heat exchange during blowing snow events which, in turn, has consequences for numerical model applications. Using data from the 2024 and 2025 campaigns at PE, this project focuses on :

• assessing whether the average of upward and downward fluxes over several months reduces the underestimation effect;
• investigating heat turbulent fluxes at higher elevations in the atmosphere;
• answering the question of whether the Monin-Obukov theory can still be applied for blowing snow events by calculating the fluxes with an atmospheric level above the blowing snow cloud.

Measurement data will also be coupled with high-resolution numerical simulations using the mesoscale model CRYOWRF.

Description:

The turbulent fluxes of sensible and latent heat between the surface of the Earth and the Atmosphere are main contributors to the total energy and mass balance and determine boundary layer dynamics and surface properties for diverse land covers from vegetation to snow and ice.

Based on the recent observation that current models heavily underestimate latent heat fluxes, this master thesis investigates the hypothesis whether radiation penetration into the snow can explain this. The figure shows a temperature profile, for which penetration of shortwave radiation into the snow causes the maximum temperature to be below the snow surface. In this case, most water vapor will be produced at the depth of the maximum temperature and the vapor gradient based on the (lower) surface temperature will be underestimated. The master thesis will explore this mechanism using a combination of numerical modelling and data analysis. Own measurements of temperature profiles in the snow will validate the modelling efforts. Modelling will be done with SNOWPACK and with OpenFOAM, trying to explicitly capture vapor transport through the snow – atmosphere interface.

Description:

• The tasks consist on deploying a LIDAR system in the Flüela valley during winter and use the data of the new Tschuggen station to study the effects of mesoscale circulations (katabatic winds, rotors, etc.) on the snow in an Alpine Valley.

Description:

• The work conducts validation studies with the new simplified snow model NIX as part of the TERRA land surface scheme in ICON. The task consists in simulating snow cover dynamics over the Alps and validating against observations. 

Description:

• The task consists on observing and mapping the snow redistribution using drones. After, the event will be simulated using the coupled snow-atmospheric model CRYOWRF to compare it with the observational dataset.

Description:

• The tasks will be compiling heat fluxes by different datasets in Antarctica and analyze them to study the uncertainty of reanalysis and climate models in Antarctica. Also set simulations with CRYOWRF to understand the importance of grid resolution.

Description:

• Snow transport by wind is a key process in snow-covered regions, significantly influencing local snow distribution and stratigraphy. In Antarctica, high wind speeds result in frequent drifting snow events that strongly impact the local surface mass balance and contribute to large-scale redistribution of snow. In alpine regions, wind-driven snow transport leads to the formation of wind slabs, which can cause unstable snowpacks and contribute to avalanche risks.
• The aim of this thesis is to validate recent improvements in a physics-based snowpack model for simulating drifting snow conditions. Model output of drifting snow mass fluxes will be compared against field measurements of drifting snow. Additionally, the study will assess how the simulated snow transport affects the resulting snow stratigraphy. These results will provide insights in the information delivered by snowpack models regarding the presence of drifting snow, and the properties of formed wind slabs, enhancing our understanding of the surface processes and improve model tools used for avalanche forecasting.

Description :

• Accurate representation of snow profiles in models is an important ingredient for the Earth’s surface energy and mass balance but has not been achieved yet. SNOWPACKFoam is a coupled modeling framework that links the multi-layer snow model SNOWPACK with the CFD solver OpenFOAM to simulate natural convection and water vapor transport in snow covers. SNOWPACK accounts for snow processes such as settling, melting–refreezing, and metamorphism, while OpenFOAM resolves the two-dimensional vapor circulation within the snow matrix and feeds back the convective vapor fluxes and resulting density changes to SNOWPACK.
• As demonstrated in Jafari and Lehning (2025), convection can strongly modify snow properties when wind slab formation is absent, producing pronounced vertical density and temperature variations. Moreover, tundra vegetation such as shrubs and herbs decreases basal snow density and creates spatial heterogeneity in snow density, which may significantly boost the onset of convection. The random distribution of these roughness elements can lower the local snow density, enabling convective motion at Rayleigh numbers below classical thresholds.
  The objective of this Master’s project is to extend SNOWPACKFoam to incorporate spatial heterogeneity in snow density caused by vegetation effects (Jafari and Lehning, 2023) and to compare simulations with and without such heterogeneity. The student will quantify how these small-scale variations influence convective structures, vapor fluxes, and overall snowpack evolution

Requirements:

• Background in atmospheric, environmental, or computational science

• Experience in C++ and OpenFOAM is advantageous but not required

• Interest in snow physics, heat and mass transport, and numerical modeling

References:

• Jafari, M. & Lehning, M. (2025). Simulating the effect of natural convection in a tundra snow cover. EGUsphere, link
• Jafari, M. & Lehning, M. (2023). Convection of snow: when and why does it happen? Front. Earth Sci. 11:1167760, link