A new hybrid particle-puff approach to atmospheric dispersion modeling, implemented in the Danish Emergency Response Model of the Atmosphere (DERMA)

Abstract

The Danish Emergency Response Model of the Atmosphere (DERMA) is a Lagrangian puff model originally developed for long-range dispersion modeling, at distances longer than roughly 50 km from the source. The model is used operationally as part of Danish emergency preparedness for the prediction of atmospheric dispersion in case of nuclear accidents, airborne spread of animal diseases, and ash from volcanic eruptions. To be able to simulate dispersion on shorter spatial scales, a new description of turbulent diffusion has been developed and implemented in DERMA, combining a stochastic particle approach with a classic puff model. Furthermore, updates have been made to the parameterizations of the turbulent wind fluctuations and Lagrangian timescales, the boundary layer height, and the initial plume rise due to heat release. These improvements allow a more realistic description of turbulent diffusion near the release location, while an updated version of the existing turbulence description is used at longer distances. The new version of DERMA is evaluated against three different tracer gas experiments: the European Tracer Experiment (ETEX), the Øresund experiment, and the Kincaid experiment. The results indicate that the new particle-puff hybrid approach gives more accurate predictions, especially on shorter spatial scales, while a small improvement is also observed for long-range dispersion.