Research projects

What mechanisms control the spontaneous rearrangement of ice motion in ice sheets? To answer this important question and shed light on better understanding the formation of ice streams in Greenland and Antarctica, we got granted 5'000'000 GPU-hours (equivalent to 80'000'000 core-hours) on the external pageLUMI-G machinecall_made, the European flagship supercomputer and number 3 in the World (Top500 list). The STREAM project represents a unique opportunity to unleash the Julia language by running the next generation full-Stokes and thermo-mechanically coupled ice flow solvers on more than 3’000 AMD MI250x GPUs. The simulations aim at resolving parts of Greenland ice flow close to meter-scale resolution in 3-dimensions. Regarding the technical aspects, the project will allow to run Julia at scale and use its powerful GPU stack, in particular external pageAMDGPU.jlcall_made, to target LUMI-G's AMD MI250x GPUs. Find more info on external pagehttps://ptsolvers.github.io/GPU4GEO/streamcall_made.

As glaciers thin and retreat, they leave behind eroded basins. This has important implications for the stability of valley walls. Unstable slopes in remote places may not be particularly hazardous to humans, but as tidewater or lake-terminating glaciers retreat, the risk increases that a landslide could fail into water, generating a tsunami and significantly enlarging the impact area. While several studies have identified a link between glacier retreat and slope stability, the role of the glacier specifically is still unclear.

This project aims to better understand to what extent landslides can be (re-)mobilized by retreating glaciers. More specifically, it will investigate the glacial, thermal, and hydrological processes controlling slope motion at different stages of glacier retreat. Additionally, we will investigate why certain sites are more susceptible to instability or even failure. We focus on southern Alaska, an area currently experiencing some of the highest glacier mass loss rates in the world. We use remote sensing, in-situ observations, and modeling. Fieldwork provides site-specific information at Portage Glacier, Alaska, where a landslide is currently being debuttressed. Our multi-year investigations will illuminate how the landslide responds to the glacier and other environmental factors.

This work will improve our understanding of glaciers’ role in initiating or exacerbating slope failures. Its results will be of broad relevance for any rapidly deglacierizing region on Earth.

Glacier water pockets (GWPs) are elusive bodies of liquid water that can form within glaciers. The term is most often used in conjunction with glacier floods, which can happen when GWPs suddenly release their waters (so-called "GWP-ruptures"). Being invisible from the surface and very difficult to detect, GWPs pose a real hazard for the areas downstream. An influential analysis of the 1980s estimated that about one third of all floods observed from Swiss glaciers originate from GWP-ruptures. Although this estimate seems remarkably high today, GWPs remain amongst the least understood features known to exist within glaciers.

This project sets out to fundamentally advance our understanding of GWPs, and to answer some long-standing questions including: How often do GPW-ruptures occur and what are the magnitudes of the related floods? Which morpho-climatic factors determine the presence of GWPs? How can the locations of GWPs be determined, and which morphologies do they have? Which mechanisms control the temporal evolution of GWPs, and can their ruptures be predicted?

To answer these questions, the project will combine (i) information from historical data and reports, (ii) extensive field observations acquired by former projects, (iii) in-depth geophysical surveys for selected locations, and (iv) numerical modelling of the relevant physical processes. We anticipate the project findings not only to be of strong interest to the broader geosciences community, but also to be of direct relevance for public and private actors in charge of managing glacier-related natural hazards and downstream infrastructures.

The Himalayan mountain range is characterized by extreme topographic gradients and a substantial glacier cover consisting of several ten thousands of glaciers of different sizes. These glaciers represent a crucial element of the hydrological cycle, as they feed the large streams of the Indus, Ganges and Brahmaputra. Past, ongoing and future changes in climate conditions have major consequences for Himalayan glaciers, thus raising concerns about water security across the Indian subcontinent.

This research project sets out to shed light on key aspects related to small Himalayan glaciers, as well as to their hydrological relevance. Focusing on glaciers smaller than 2 km2 , it aims at better characterizing their present state based on improved observational techniques at a benchmark site, their future evolution using enhanced glaciological process-based models, and their effect on local to regional hydrology with a special focus on future changes in interannual variability. The project builds upon a combination of in situ and remote observations, as well as upon process-based numerical analysis that includes both glaciological modelling at high spatial resolution and streamflow modelling at local to regional scales. The combination will result in the first detailed study on the present and future state of small glaciers in the Himalayas and on the impact of 21st century glacier melt on water availability.

The so-called “Spezialbefliegungen” are a data set of very high-resolution aerial photographs and digital elevation models covering about 100 Swiss glaciers and proglacial areas. Since 2012, this data has been acquired at yearly intervals by the Federal Office of Topography in the frame of the Glacier Monitoring Switzerland (GLAMOS) programme. Although both the quality of the data and its spatio-temporal resolution are unique, the resource has been strongly underused so far.

This project sets out to change this, and includes four main activities focusing glaciological and periglacial research questions, as well as data visualization for public outreach. More specifically, the project will (1) advance the extrapolation of glacier mass balance from local measurements to the regional scale, (2) assess the abundance and formation of surface-collapse features on Alpine glacier tongues, (3) quantify periglacial sediment dynamics at the multi-catchment scale, and (4) perform 3D visualization of annual aerial imagery and terrain models for public outreach. These activities will unlock the full potential of a unique data set documenting one decade of climate-related changes in the Swiss Alps.

Knowing the ice thickness of glaciers is essential for a number of applications, ranging from the detection of potential ice-core drilling sites to constraining the glacier’s and ice sheet’s ultimate contribution to sea level rise. In terms of measurement techniques, Ground Penetrating Radar (GPR) is clearly the method of choice. This project – funded by the external pageSwiss Polar Institutecall_made and conducted in close collaboration with ETH’s Institute of Geophysics and Professorship of Space Geodesy – sets out to re-design the “Airborne Ice Radar of ETH Zurich (AIR-ETH)“, a self-build GPR that is amongst the handful of systems existing worldwide for the airborne surveying of high-altitude glaciers. Weight minimization, inclusion of a fully autonomous positioning system, increases in signal quality, and the boosting of the system’s versatility are the main targets. We envision that the project will result in a system that is deployable both from the air as from the ground in virtually any glacierized region on Earth.

Due to ongoing climate change, drought-related extremes will lead to significantly negative ecologic, economic, and human-health related effects within many different areas. So far, these have mostly been investigated separately. Focusing on the hydro-meteorological aspects, for example, the external pagewww.drought.chcall_made platform has been developed for informing end-users about dryness in Switzerland, and has recently been enhanced as to include sub-seasonal forecasts. Despite the experience of probabilistic drought forecasting as well as expertise in modelling impacts like forest fire, glacier alterations, bark beetle developments, or biodiversity, there is a lack of understanding of the benefits of forcing such impact models with sub-seasonal (monthly) weather and hydrological forecasts.

The project external pageMaLeFiXcall_made, led by WSL's group external pageHydrological Forecstscall_made and including a consortium of WSL and ETH scientist, aims at closing this gap between disciplines. It aims at doing so by combining diverse models in a user-friendly and optimal way, and by using state-of-the-art machine learning methods. Our contribution is related to the projection of water yields from glaciers, and of related indicators. The outcome should provide end users with a decision tool, explained in a language readily accessible to stakeholders thanks to the help of results from the social sciences, which allows to detect and prepare for drought-related extremes weeks in advance. The ultimate goal is to gain time for avoiding and mitigating environmental and socio-economic impacts.

Climate change is rapidly altering mountain environments, increasing the need for reliable tools for the prediction, monitoring, and early warning of alpine mass movements. In this project we integrate satellite-based InSAR data and optical images to monitor slope instabilities, evaluate how meteorological conditions drive slope deformation, and to assess how these driving factors are expected to change under future climate scenarios. Complementing the coarse temporal resolution of the satellite data, we also leverage seismic and infrasound measurements to characterize the subsurface processes driving the instability at Spitzer Stein (BE), catalogue the mass movements at this key location, and train an intelligent detector to identify them in real time. Leveraging the seismic data, as well as automatic avalanche detection systems, this project further exploits recent advances in machine learning to improve the spatio-temporal resolution of avalanche forecasting in Switzerland and assess changes in avalanche hazard due to climate change. This project is in support to the external pageWSLcall_made initiative external pageClimate Change Impacts on Alpine Mass Movements (CCAMM)call_made.

Computational Earth sciences rely on modelling to understand and predict the evolution of complex physical systems. Ice sheet dynamics and geodynamics modelling are, despite their apparent differences, two domains that build upon similar physical models and ensuing computational challenges: Extremely large simulations which are not possible with the state of the art in high-performance computing (HPC), thus requiring next-generation supercomputers. Those large-scale simulations are crucial when assessing ice flow dynamics over Greenland and Antarctica, understanding the dynamic motion of Earth’s interior or performing three-dimensional simulations.

The goal of the external pageGPU4GEO projectcall_made is to propose software tools which provide a way forward in these two domains by exploiting two powerful emerging paradigms in HPC: massively parallel iterative solvers, and HPC with Julia on graphical processing units (GPUs). Solvers based on pseudo-transient relaxation show great promise for solving a wide range of mechanical multi-physics problems in geoscience, at scale and on GPU-accelerated supercomputers. We use Julia as the main language because it features high-level and high-performance capabilities and performance portability amongst backends (e.g. multicore CPUs and AMD, NVIDIA).

The 4D-Antarctica project focuses on creating and consolidating an unprecedented dataset composed of ice-sheet wide hydrology and lithospheric products, Earth Observation datasets, and state of the art ice-sheet and hydrology models for the whole of Antarctica. VAW-GL is a Co-PI on this international project lead by Noel Gourmelen (U. Edinburgh, UK) with the responsibility to develop new subglacial water flow models and apply them to the whole of Antarctica. The aim is to enhance our understanding of the subglacial hydrology and its impact on the future development of Antarctica.

The ETH Zurich is part of this a large European project funded in the frame of Horizon 2020. The overarching scientific objective of PROTECT is to assess and project changes in the land-based cryosphere with quantified uncertainties, to produce robust global, regional and local projections of SLR on a wide range of timescales. The project will place particular emphasis on the low-probability, high impact scenarios of greatest interest to coastal planning stakeholders. A novelty in PROTECT is the strong interaction between these stakeholders and the sea-level scientists (ranging from glaciologists to coastal impact specialists) to identify relevant risks and opportunities from global to local scales and enhance European competitiveness in the provision of climate services. The contribution of the ETH Zürich focuses on the global modelling of glaciers both in the past and in the future using GloGEMflow, thus building up on our previous work and other running projects. The focus is on further model development, e.g. regarding ice-flow dynamics and the effect of debris cover, and on a new calibration-validation scheme making use of the full breath of glacier-specific geodetic mass balances worldwide.

For till-bedded glaciers, subglacial flow and till dynamics are intrinsically connected. While subglacial flow is a feature of glacier drainage, till dynamics can be understood as the temporal evolution of the underlying sediment due to water flow. Subglacial water flow has major impacts on the hydrographs of proglacial streams, on basal sliding and glacier motion, as well as on the dynamics of glacial lake outbursts. Together with till dynamics, the processes affect sediment uptake and evacuation from glacierised environments. Directly accessing the subglacial environment is very difficult and thus, corresponding observations remain scarce. As a consequence, numerical models that aim at representing subglacial processes are poorly tested, and often rely on decades-old parameterisations.

The goal of this project is to overcome these shortcomings by investigating the hydrological and mechanical conditions beneath a till-bedded glacier with physical experiments. For this end, a new laboratory flume will be constructed, in which pressure is applied to the till simulating the weight of a glacier. At the same time, water is allowed to flow along the till-ice boundary. In the early stages of the experiment, we use an ice substitute in order to avoid the otherwise necessary cooling of the entire setup. The new flume will allow for current theories related to till dynamics and the subglacial drainage under till-bedded glaciers to be tested in a controlled environment, and to precisely constrain some key parameters of the related processes.

How much are Swiss glaciers melting right now? The CRAMPON project investigates the potential of an operational modeling tool to nowcast and predict mass balance and runoff of all Swiss glaciers. We assimilate both field data and near real-time remote sensing observations into the workflow, and use ensemble approaches for uncertainty assessment. The operational tool resulting from this work aims to be one of the first near-real time information web-platforms for glacier mass balance and runoff. This is anticipated to enable better management of hydropower and water resources. In the context of climate services, the tool supports decision making related to electricity production and water allocation.

Taking basic parts of the external pageOpen Global Glacier Model (OGGM)call_made infrastructure, we set up an ensemble of mass balance models including snow cover processes and albedo aging. Model calibration for each of the approximately 1500 Swiss glaciers is done exploring the suitability of repeated Digital Elevation Models (DEMs) from the Swiss National Forest Inventory (LFI). Model ensemble validation is done with field measurements from the Glacier Monitoring Switzerland (GLAMOS) program and in situ melt observations with on-ice cameras (see also Operational near-real-time glacier monitoring).
Further, the added value of data assimilation for operational glaciological modeling will be investigated. We use extents of melt areas derived from satellite radar data and the glacier snow-covered fraction derived from optical imagery. Different assimilation techniques are compared and their performance is evaluated.

Eventually, the calculations will be extended to deliver runoff from glacierized catchments.