PhD on machine learning-based flood forecasting at national scale

With this PhD position, you have the opportunity to develop cutting-edge modelling systems to provide key hydrological information to the Danish society. The real-time forecasting system to be developed will leverage information on flood inundation risk to increase resilience towards hydrological extremes.

The PhD project is part of a large project that involves collaboration with the Danish Meteorological Institute (DMI), the Danish Agency for Data Supply and Infrastructure (SDFI) and the Danish Environmental Protection Agency (MST).

Job description
Within the PhD project you will conduct cutting-edge research within large-scale flood inundation modelling utilizing a suite of modelling tools, i.e., hydrological models, hydrodynamic models, machine learning (ML) and hybrid models. The latter benefit from combining the best of the alternative modelling approaches. You will work with satellite-derived (e.g. Sentinel SAR and optical data) inundation maps, observed water levels in streams, precipitation data, and high-resolution digital elevation models to further develop and contrast the alternative modelling approaches. Your project will be transdisciplinary spanning over hydrology, remote sensing and ML, in particular spatial Deep Learning approaches.

The specific objectives include (1) the development of a novel ML-based flood inundation modelling framework, (2) the inter-comparison of alternative methods (ML-stand alone, hydrodynamic models and different degrees of novel hybrid models) to model inundated areas, (3) transferring the methods from Denmark to different geographic settings.

As a PhD at GEUS, you will contribute to the further development of the ongoing efforts to operationalize the national coupled groundwater surface water model (DK-Model) to provide real-time and forecast information on the status of groundwater and surface water to the Danish society. You will work closely with a Post Doc who will be hired on the same project starting in January 2024. You will be enrolled at the PhD school at the Faculty of SCIENCE at the University of Copenhagen. At the Department of Geosciences and Natural Resource Management at University of Copenhagen, you will also interact and collaborate with early career researchers from related projects, e.g., the recently launched Global Wetland Center.

You should expect to broaden your existing field by covering elements of hydrological modelling, remote sensing, machine learning, data analysis and data management. Your research findings should be published in leading journals within the field. You are expected to contribute to the internationalization of the project by attending international conferences and going on an external research stay during the course of your PhD project.