A New Dispersion Model for Highly-Buoyant Plume Dispersion in the Convective Boundary Layer

This study was aimed at correcting and improving the modeling of highly-buoyant plume dispersion in the convective boundary layer as currently treated in AERMOD (Cimorelli et al., 2005). This situation typically occurs in the morning under relatively low mixed-layer depths and light winds. Previous work has shown that AERMOD tends to overpredict surface concentrations in this situation due to deficiencies in the ``indirect” and ``penetrated plume” models, which treat downward dispersion from the CBL top (indirect plume) or above the inversion layer (penetrated plume). The new indirect source model addresses the ``entrapment” process wherein a highly-buoyant plume rises to the CBL top and becomes temporarily trapped in the inversion layer. This model consists of sub-models for the initial entrapment, entrainment of plume material into the CBL based on a Richardson formulation, and subsequent dispersion. The new penetrated plume model has components for: 1) the plume fraction that penetrates the inversion and includes a temperature jump at the CBL top along with a deeper stable layer, and 2) penetrated plume dispersion which is handled by a fumigation model since that is the process occurring when the growing mixed layer intercepts the elevated plume. The models are compared with high-fidelity convection tank experiments (Weil et al., 2002) and show good agreement with the data. In addition, they are evaluated with power plant plume data around Maryland power plants and the Kincaid (Illinois) plant and also exhibit good skill.