The PIXIL (Pyrenees Imaging eXperience: an InternationaL network) project seeks to contribute to the innovation and competitiveness of geothermal energy in the cross-border area between Spain and France, and thereby generate regional conditions and opportunities to move towards energy transition.
PIXIL proposes to develop innovative algorithms to solve geophysical imaging problems applied to geothermal energy. In this sense, the Barcelona Supercomputing Center (BSC) and the Basque Center for Applied Mathematics (BCAM) are joining forces to improve electromagnetic modeling and inversion tools within the framework of the PIXIL project. Specifically, it seeks to integrate deep learning algorithms implemented by the BCAM to the electromagnetic modeling tool developed by the BSC. This aims to bring together the best of both worlds to obtain a state-of-the-art, robust inversion kernel that contributes to PIXIL's goals. It worth mentioning that the main criterion for evaluating the efficiency of these algorithms will be their potential to improve the competitiveness of the regional industry.
The BCAM-BSC collaboration is fundamental for the Action 3 of the PIXIL project that focuses on the development of new numerical algorithms for geophysical imaging. These kinds of algorithms are able to reproduce the response of different materials to external excitation, electromagnetic or mechanical, in order to analyze the real responses (data) and infer models of the subsurface as correct as possible without the need to punch out. The response of the subsoil is modeled by numerical algorithms that allow reproducing the behavior of the system in a computer.
One of the main objectives of the project is to produce highly differentiating and innovative knowledge in the context of geothermal energy. Therefore, working in developing an HPC code for electromagnetic inversion based on Deep Learning techniques represents an important step towards this goal.
