Researcher (Postdoc) Deep learning for geophysical observables extraction and inversion: Toward characterizing the Dutch sub...

The global initiative to complete the energy transition by 2040 puts tremendous pressure on governments and industries. The need for any nation to know the potential of its subsurface to meet the future demand for energy and raw materials, as well as its storage capacity (e.g. CO2, H2) and associated hazards (e.g. induced seismicity, subsidence), is more critical than ever for informing exploration efforts, strategic economic planning, environmental policies, and decision-makers.

The Netherlands has acquired a vast amount of geophysical data over the past 30 years. However, the formal integration and inversion of all available and complementary data have not been yet fully exploited for obtaining unifying, multi-parameter models of the subsurface. This project will develop novel ways of characterizing the physical state of the subsurface beneath the Netherlands via the formal integration of multiple geophysical (mainly seismological), geochemical/petrological and petrophysical data into a single physics-based framework. Specifically, we will establish a workflow to extract seismological observables from large seismological data sets using both existing and new Deep Learning (DL) techniques to be developed within the project. The retrieved seismological observables will be integrated with other available data sets (e.g. potential fields) and existing models to establish a common database to be inverted using an innovative multi-observable probabilistic platform. The new DL methods and workflow to extract high-quality seismic information will apply to similar data sets in other regions, opening new opportunities for analysing and using seismic records worldwide. Results will be disseminated through high-quality scientific publications, conference presentations and reports for a wide range of partners. Involvement in teaching and/or supervision of PhD projects related to the project is also expected.

You will collaborate closely with researchers at Utrecht University, KNMI, and Delft University of Technology. You will be a member of a large and dynamic geophysics/computational geodynamics team of PhD students and post-doctoral fellows.

Your project will be embedded in the Applied Earth Sciences department at the Faculty for Geo-Information Science and Earth Observation at the University of Twente. You will work together with the partners in the DeepNL INTEGRATION project to develop and generate new tools and models.

The department of Applied Earth Sciences combines earth scientific knowledge with dynamic modelling and advanced remote sensing, to analyse earth systems and processes in space and time. Our goal is to contribute to global challenges concerning future demands for earth resources and to help reduce disaster risk and the impact of natural hazards on communities living in changing environments.