PhD position on machine learning based predictions of river flow and water temperature in a changing climate

The Land-Climate Dynamics group at ETH Zürich is looking for a Doctoral candidate to leverage machine learning to investigate changes in Swiss water temperatures and river flow in support of energy system science.The Land-Climate Dynamics group at ETH Zürich is looking for a Doctoral candidate to leverage machine learning to investigate changes in Swiss water temperatures and river flow in support of energy system science.

Both river flow volumes and water temperature are essential factors in the energy system as they directly influence generation of hydropower and impact the cooling efficiency of thermal power plants. In Switzerland, anthropogenic climate change is affecting both water quantity and temperatures in rivers and streams thereby raising legitimate questions on associated risks to the energy system. To advance our understanding of climate change impacts on water flow volumes and temperature, the goal is to build on recent advances in using machine learning for hydrological modelling to develop spatially contiguous data-driven models of daily water temperature and river flow in Switzerland. These will be linked with climate projections to investigate to what extent climate change is increasing the risk of riverine heatwaves - relevant to thermal power plants - and hydrological droughts – which may be a threat to hydropower production.

The research will be conducted within the RECIPE (Resilient Infrastructure for the Swiss Energy Transition) project that is funded by the SWiss Energy research for the Energy Transition (SWEET) program in collaboration with the National Centre for Climate Services (NCCS).

As a doctoral student you will conduct quantitative and theoretical research at the interface between hydrological and climate sciences and investigate approaches that allow for a joint prediction of flow volumes and water temperatures at ungauged locations in Switzerland. This involves applying deep learning methods to in-situ observations and to use the resulting models in together with climate model simulations to explore past and future extremes of flow volumes and water temperatures across Switzerland.

• You hold a master’s degree in climate/environmental science, hydrology, physics, statistics, applied math, computer science or a related field (or you will graduate soon).
• You are a creative personality and enjoy solving scientific problems independently as well as in team efforts.
• You have a strong interest in terrestrial climate variability and hydrological extremes.
• You are passionate about using quantitative methods with focus on statistical data analysis and/or large-scale physical modelling.
• You have a good command of scripting languages suitable for data analysis (e.g. Python), knowledge of machine learning libraries (e.g. PyTorch) is beneficial.
• You have good command of spoken and written English.

• ETH Zurich is a family-friendly employer with excellent working conditions. You can look forward to a competitive working conditions, an exciting working environment, cultural diversity, and attractive offers and benefits 
• Stimulating research environment within the group for Land-Climate Dynamics lead by Sonia I. Seneviratne at the Institute of Atmospheric and Climate Science 
• Supervision by Sonia I. Seneviratne and Lukas Gudmundsson towards a Doctorate at the Department of Environmental System Science, ETH Zürich