The advection–diffusion of a scalar field arises in a wide range of physical contexts. In combustion processes, the advected scalar typically represents the mixture fraction between the various injected components. In heat transfer problems, the scalar field corresponds to temperature. In respiratory system simulations, the scalar represents pollutants or drugs. In contrast, for non-diffusive scalars, the advection of a level-set function remains a challenging problem and continues to be the focus of extensive research. Most numerical studies reported in the literature solve the advection–diffusion system using classical Eulerian frameworks, subjected to a temporal stability constraint, commonly referred to as the CFL condition. Lagrangian particle methods have been shown to offer a key advantage in this context, as computations are performed only on the support of the scalar field. Semi-Lagrangian methods [1] provide an effective compromise between these approaches. Being grid-based, they combine high-order accuracy and good conservation properties with significantly relaxed stability constraints compared to the classical CFL condition. We present recent developments in the coupling of a fluid solver with a scalar transport solver using a multicode approach. The Navier–Stokes equations are discretized and solved with the finite element method, while scalar transport is handled using the previously described hybrid Lagrangian method. The framework is applied to the simulation of pollutant transport in urban environments. Both are solved with Alya simulation software [2]. [1] Sonnendrücker E., Roche J., Bertrand P. and Ghizzo A., The semi-Lagrangian method for the numerical resolution of the Vlasov equation, J. of computational physics 149, 2, 201-220, 1999. [2] Vázquez M., Houzeaux G. et al., Alya: Multiphysics engineering simulation toward exascale, J. of Computational Science 14, 15-27, 2016.