Large-eddy simulation and stochastic modelling of Lagrangian particles for footprint determination in stable boundary layer

Large-eddy simulation (LES) and Lagrangian stochastic modelling of passive particle dispersion were applied to the scalar flux footprint determination in stable atmospheric boundary layer. The sensitivity of the LES results to the spatial resolution and to the parameterizations of small-scale turbulence was investigated. It was shown that the resolved and partially resolved "subfilter-scale" eddies are mainly responsible for particle dispersion in LES, implying that substantial improvement may be achieved by using recovery of small-scale velocity fluctuations. In LES with the explicit filtering this recovering consists of application of the known inverse filter operator. The footprint functions obtained in LES were compared with the functions calculated with the use of first-order single particle Lagrangian stochastic models (LSM), zeroth-order Lagrangian stochastic models – the random displacement models (RDM), and analytical footprint parameterisations. It was observed that the value of the Kolmogorov constant C0 = 6 provided the best agreement of the one-dimensional LSMs results with LES, however, also that different LSMs can produce quite different footprint predictions. According to presented LES the source area and footprints in stable boundary layer can be substantially more extended than those predicted by the modern analytical footprint parameterizations and LSMs.