Data assimilation for predicting pollution maps under realistic environmental conditions

Topic description

Controlling pollutant emissions is one of today’s most pressing environmental challenges. Greenhouse gases (GHGs) and industrial emissions such as methane (CH₄) and hydrogen (H₂) directly impact air quality, safety, and climate change. Yet predicting pollutant dispersion in complex environments like industrial sites remains difficult due to fluctuating wind conditions and obstacles.
This PhD project offers a unique opportunity to develop innovative tools combining high-fidelity simulations based on a Lattice Boltzmann Method (LBM) CFD code with advanced data assimilation techniques, notably the Ensemble Kalman Filter (EnKF). By integrating both fixed and mobile sensors, the candidate will design methods to improve turbulence models, reduce uncertainties, and deliver reliable real-time forecasts. The ambition is clear: support better control of industrial emissions and reduce health and environmental risks.
Bringing together fluid mechanics, high-performance computing, and data science, this project fosters a dynamic, collaborative setting supported by strong academic and industrial partnerships. The candidate will rely on real data from industrial measurement campaigns and contribute to cutting-edge advances in data assimilation. A distinctive feature will be the adaptive control of mobile sensors to optimize data collection and further reduce uncertainties.
This PhD strives for strong scientific impact, contributing to global emission monitoring and decarbonization efforts. Results will be shared through high-level publications and international conferences, ensuring excellent visibility within both academy and industry.

Starting date

-11-02

Funding category

Public funding alone (i.e. government, region, European, international organization research grant)

Funding further details