BSC researchers win HPC Innovation Excellence Award for improving air quality using CFD simulations

14 December 2022

A work carried out by researchers from the Computer Applications in Science and Engineering (CASE) Department of the Barcelona Supercomputing Center-Centro Nacional de Supercomputación (BSC-CNS)  has received an HPC Innovation Excellence Award in a virtual ceremony organized by Hyperion Research.

The research, developed within the FF4EUROHPC iBAM experiment in collaboration with air quality mapping company Bettair, focused on mitigating the effects of air pollution on local communities by means of advanced Computational Fluid Dynamics (CFD) and Machine Learning (ML) techniques. The objective was to develop a low-computational-cost high-accuracy meteorological downscale and dispersion solver to help predict the distribution of pollutants in urban environments. This new tool will overcome the need for HPC in the simulations by mimicking the wind dynamics using CNNs, a well-known neural architecture in ML.

As a result of the project, BSC and Bettair have generated two types of numerical datasets: the first one is an array of high-precision wind simulations inside 30 different European cities that relate mesoscale meteorological conditions with the urban wind field for a particular geometry. Then, a second dataset relates pollutant dispersion within the urban area for 450 different emission scenarios and the meteorological conditions commented above. This is done using ALYA, a CFD software of the BSC that runs on HPC. Then, an experimental dataset was created from a measurement campaign in El Prat de Llobregat. Bettair complemented the two already existing air quality reference stations of the city with seven Bettair air quality monitors, five 2D sonic anemometers and a 3D sonic anemometer that are scattered strategically to collect validation data in relevant locations. This data has been used to validate the performance of the trained models with real data collected throughout the whole year. The final model trained with the first data set and validated with the second, has been able to predict both the wind characteristics and pollution dispersion with sufficient accuracy and significant time to solution to respect traditional CFD techniques.

“We are very happy to be part of the team that received this award,” said Oriol Lehmkuhl, team leader of the Large-scale Computational Fluid Dynamics Group in the CASE Department and lead partner of the FF4EUROHPC project from BSC. “Air pollution is the largest environmental health risk in Europe and is a major cause of premature death and disease. Knowing how air pollutants spread could help us find solutions to contain them or mitigate their effects.”