Topics Master in Earth Sciences ETH

We are looking for two motivated Master's students to join our exciting research projects exploring microbial communities in Alpine ecosystems and groundwater environments. You will gain hands-on experience in molecular techniques, microbial culturing, and field sampling in the Alps, contributing to critical research on microbes and ecosystem functioning. If you're curious, enjoy outdoor fieldwork, and have a passion for microbiology and ecosystem exploration, we'd love to hear from you! For more information: https://tinyurl.com/geobiology2025

Hoang at htrannguyen@ethz.ch Prof. Dr. Magnabosco at cmagnabosco@ethz.ch

Accurate modelling of glacier mass balance requires the representation of wind-driven snow processes. Yet, the physical representation of these processes is computationally too expensive to include in standard mass balance models and, therefore, requires simplification via parameterization schemes.

This thesis will compare existing but not widely used parameterizations for different climatic regions (continental Alpine glaciers, maritime glaciers in New Zealand, and seasonal snow cover). The aim is to consolidate existing parameterizations for applications in standard energy and mass balance models.

For further information please contact Dr. Ruzica Dadic

Antarctic ice streams feed approximately 90% of Antarctica’s ice into the world’s oceans. These ice streams flow to the ocean through a combination of internal deformation and by sliding over and deforming the rock and sediment that underlies them. Processes at the base of the ice are particularly important to understand as they can result in rapid changes in ice flow velocity at time scales of hours to days and years to centuries. Many of these processes emit acoustic signals which can be recorded at the ice surface making passive-source seismology an ideal tool with which to study them. This project will use existing data to examine the transition where the grounded Kamb Ice Stream crosses into the Ross Ice Shelf and contributes to global sea level.

Project aims include identifying, locating, and characterising seismicity from sources at the base of the ice stream and within the ice shelf. These sources include water flow in a large subglacial drainage channel, and crevasse formation due to flexure of the floating ice shelf as it rises and falls with the tide. To achieve the aims you will use existing tools (e.g. obspy) to process and analyse the data. Basic programming skills, or the desire to develop them, are required. The ideal candidate would also have an interest in glaciology and applied geophysics. Should time allow, the results will be compared with other complementary data (e.g. Global Navigation Satellite System positioning) and model output (e.g. subglacial water routing.)

For further information please contact Dr. Huw Horgan (horganh@ethz.ch)

**Context and objectives** Anthropogenic climate warming is causing rapid melting and retreat of mountain glaciers worldwide, which is of particular concern for regions that depend on them for water resources (Immerzeel et al., 2020). While the processes influencing melt at the surface of glaciers are fairly well constrained, there remains some major uncertainties related to key processes occurring within these ice masses, including: - Basal melt from subglacial streams and conduits which lead to collapse features and the disintegration of stagnating glacier tongues (Egli et al., 2023; Kneib et al., 2023). As of today, this melt contribution has never been measured directly and is still unknown for the vast majority of glaciers. - The advection of englacial rock debris within the ice, which when it emerges at the surface of the glacier in the ablation zone leads to the formation of a layer of supraglacial debris that strongly influences the ice melt. There exists almost no englacial debris concentration measurements, which hinders the representation of debris evolution and its effect on ice ablation in current glacier models (Miles et al., 2021; McCarthy et al., 2022; Compagno et al., 2022). We have developed a robot that can travel in subglacial and englacial glacier conduits, and that transports a small LiDAR to map the 3D geometry of these features (Polzin & Hughes, 2024). In this project, you will build a processing workflow to derive englacial volume changes and englacial debris concentrations from repeated LiDAR acquisitions taken over the summer 2025 on various glaciers of the Swiss Alps. **Methods** The processing workflow will include the following processing steps and leverage several point cloud processing functions (Kneib et al., 2022): - Point cloud acquisitions with the LiDAR in the field - Pre-processing of each individual point cloud for internal consistency - Coregistration of the point clouds using reference points and/or stable terrain - Uncertainty and volume change estimation - Processing of englacial debris concentrations using pictures from the robot's camera - Development of a report in journal manuscript form that describes changing ice melt patterns and the concentration of englacial debris **Application** We are looking for an enthusiastic candidate in the first or second year of a Master's degree in earth sciences, remote sensing or any other related field. The work is expected to last ~6 months and start in June 2025. **Requirements** Good maths and physics background (fluid mechanics, algebra), Quantitative data analysis and programming (Python, Matlab, R), Experience in the use of Geographic Information Systems (e.g. QGIS), photogrammetry, robotics and/or remote sensing is a plus. **Conditions** Location: The project is conducted in collaboration between EPF Lausanne, Université de Lausanne (UNIL) and ETH Zürich, the successful candidate can therefore be based either at UNIL or ETHZ and will make several visits to the other institutions. Fieldwork: Based on the interest and skills of the successful candidate, fieldwork in one to three days periods in the Alps will be possible, spread over the whole summer (June-October). Conditions: 2250 CHF/months salary. Please send us a CV and a one-page cover letter describing your motivation and experience for the position. **References** Compagno, L., Huss, M., Miles, E. S., McCarthy, M. J., Zekollari, H., Dehecq, A., Pellicciotti, F., & Farinotti, D. (2022). Modelling supraglacial debris-cover evolution from the single-glacier to the regional scale: an application to High Mountain Asia. The Cryosphere, 16(5), 1697–1718. https://doi.org/10.5194/tc-16-1697-2022 Egli, P. E., Belotti, B., Ouvry, B., Irving, J., & Lane, S. N. (2021). Subglacial Channels, Climate Warming, and Increasing Frequency of Alpine Glacier Snout Collapse. Geophysical Research Letters, 48(21). https://doi.org/10.1029/2021GL096031 Kneib, M., Fyffe, C. L., Miles, E. S., Lindemann, S., Shaw, T. E., Buri, P., McCarthy, M., Ouvry, B., Vieli, A., Sato, Y., Kraaijenbrink, P. D. A., Zhao, C., Molnar, P., & Pellicciotti, F. (2023). Controls on Ice Cliff Distribution and Characteristics on Debris‐Covered Glaciers. Geophysical Research Letters, 50(6). https://doi.org/10.1029/2022GL102444 Kneib, M., Fyffe, C. L., Miles, E. S., Lindemann, S., Shaw, T. E., Buri, P., McCarthy, M., Ouvry, B., Vieli, A., Sato, Y., Kraaijenbrink, P. D. A., Zhao, C., Molnar, P., & Pellicciotti, F. (2023). Controls on Ice Cliff Distribution and Characteristics on Debris‐Covered Glaciers. Geophysical Research Letters, 50(6). https://doi.org/10.1029/2022GL102444 McCarthy, M., Miles, E., Kneib, M., Buri, P., Fugger, S., & Pellicciotti, F. (2022). Supraglacial debris thickness and supply rate in High-Mountain Asia. Communications Earth & Environment, 3(1), 269. https://doi.org/10.1038/s43247-022-00588-2 Miles, K. E., Hubbard, B., Miles, E. S., Quincey, D. J., Rowan, A. v, Kirkbride, M., & Hornsey, J. (2021). Continuous borehole optical televiewing reveals variable englacial debris concentrations at Khumbu Glacier, Nepal. Communications Earth & Environment, 2(1), 12. https://doi.org/10.1038/s43247-020-00070-x M Polzin, J Hughes, “Into the ice: Exploration and data capturing in glacial moulins by a tethered robot”, Journal of Field Robotics 41 (3), 654-668

For further information, please contact Marin Kneib kneibm@ethz.ch, Ian Delaney ianarburua.delaney@unil.ch or Max Polzin max.polzin@epfl.ch

Seismologists record seismic waves generated by earthquakes to monitor seismic activity and understand earthquake dynamics. However, earthquake recordings are often contaminated by a range of noise signals, which complicate the analysis and interpretation of earthquake signals, especially in automated processing workflows. With the rapid growth of earthquake data, reliable denoising techniques are becoming increasingly important. Effectively separating earthquake signals from noise is a game-changer for automated workflows, allowing for faster and more accurate signal processing across numerous recordings with minimal latency. You will develop state-of-the art methods to effectively “clean” time series of earthquake signals by building on existing deep learning models [e.g.,1,2,3] and incorporating your own creative solutions to design an optimal approach. The denoising performance will be evaluated using metrics relevant to seismic data analysis, benchmarked against conventional methods, and demonstrated on earthquake sequences. **Related Literature** - [1] https://doi.org/10.1029/2022JE007503 - [2] https://doi.org/10.1029/2024JH000179 - [3] https://doi.org/10.48550/arXiv.2309.05975

- Assess existing denoising models, identify their limitations, and propose improvements. - Design real-time monitoring solutions for earthquake early warning applications. - Integrate denoising methods into operational earthquake monitoring with continuous data streams. - Contribute your own ideas to advance earthquake monitoring! **Requirements** - You are excited to bring your strong data science and machine learning skills to advance earthquake monitoring through a Master’s thesis. - You enjoy programming in Python, TensorFlow / PyTorch, and are passionate about developing and implementing machine learning approaches to solve complex problems **Hosted by:** Nikolaj Dahmen, Men-Andrin Meier (Department of Earth and Planetary Sciences) **In cooperation with:** Simon Dirmeier (Swiss Data Science Center)

nikolaj.dahmen@eaps.ethz.ch

The thesis focuses on applying a glacier mass balance (MB) model based on machine learning, driven by climate variables and topographical features. The model is part of the Mass Balance Machine (https://github.com/ODINN-SciML/MassBalanceMachine), a collaborative project that aims to create an open-source machine learning MB modelling package. While the Mass Balance Machine is currently being applied to the Swiss Alps and Norwegian glaciers, the thesis aims to investigate its transferability to the French Alps. The project requires programming knowledge (Python), and while it is not restricted to computer science students, it expects some familiarity with data science and machine learning methods.

Package development: participate in developing the Mass Balance Machine package by tailoring it to the French Alps and producing unit tests, potentially also trying out new ML architectures. Science: make predictions of glacier-wide and regional mass balance for the French Alps. This will encompass data gathering, pre-processing, and the development of a testing and validation pipeline.

For further information, please contact Marijn van der Meer