Lagrangian exposure and residence times for urban street-canyon models

The residence time of a pollutant is most naturally calculated in the Lagrangian frame. Typically, however, the residence time is estimated from Eulerian data. Using large-eddy simulation and a Lagrangian stochastic model, we calculate the Lagrangian residence time (and the closely related exposure time, which allows for re-entrainment) for a single street canyon. It is shown that the residence and exposure timescales are much shorter than typical Eulerian estimates of the mean age. It is also shown that the PDFs of the Lagrangian timescales have long exponential tails. The role of the Lagrangian stochastic parameterisation will also be discussed.