Significant emissions of anthropogenic nitrogen oxides (NOx) and sulfur oxides (SOx) are released at various levels into the atmosphere from individual elevated point sources. These primary species are important precursors for a variety of secondary pollutants, in particular, ozone and aerosols. Notable aspects of the plumes emitted from these major point sources include the initial, small physical dimensions in the horizontal and vertical and their finite growth rate. Depending on meteorological conditions, pollutant plumes often require up to several hours to grow to a typical regional model grid cell size. This diffusion-limited nature of plumes is in contrast to the traditional method applied in Eulerian grid modeling, which has been to uniformly and instantly mix point source emissions into the entire volume of each model grid cell. Since the typical regional model grid size is generally 20 km or greater, a real-world point source plume may remain a subgrid scale feature for a substantial time period and for a significant distance downwind after release.
The inadequate treatment of plume dynamic processes by Eulerian grid models also has important implications on the photochemical process. The effect of instantaneous overdilution of point source emissions into rather large grid cells initiates rapid photochemical production of ozone, particularly when volatile organic compounds (VOCs) are simultaneously available. The subsequent spatial and temporal distortions of the chemical processes contribute to model uncertainty. Consequently, it has been recognized that a realistic, subgrid scale modeling approach is needed to treat the relevant physical and chemical processes impacting this notable class of large point source emissions.
A multi-year, cooperative research and development effort was undertaken to design and implement a plume-in-grid (PinG) technique in the Models-3 CMAQ Chemical Transport Model (CCTM). The PinG approach has been designed to be suitable for the largest NOx point sources for regional model grid resolutions where the subgrid scale error in the representation of these sources is expected to be greatest. A description of the key modeling components will be provided. A Plume Dynamics Model (PDM) serves as a processor program to perform plume rise, and to generate plume dimensions, plume position and related parameter data needed by the PinG module. A Lagrangian Reactive Plume Model is the principal PinG algorithm which has been fully integrated and coupled with the CCTM. Each Lagrangian plume cross-section consists of a contiguous array of "plume cells". The PinG module simulates the dispersion, transport, chemistry, and removal processes impacting each plume section for multiple sources. The PinG module operates simultaneously during a CCTM grid model simulation and it applies the same chemical mechanisms (RADM2, CB4) and solvers as the CCTM. The CCTM grid concentrations provide boundary values for the entrainment process in the PinG plumes, and an important feedback occurs when the plume pollutants are transferred to the grid system at an appropriate time and grid cell. An overview of technical aspects of the PinG module will be given and initial simulation results versus observed plume data will be described.
Symposium on Interdisciplinary Issues in Atmospheric Chemistry