Abstract:Discrepancies in grid structure, dynamics and physics packages in the offline coupled NWS/NCEP NAM meteorological model with the U.S. Environmental Protection Agency Community Multiscale Air Quality (CMAQ) model can give rise to inconsistencies. This study investigates the use of three vertical mixing schemes to drive chemistry tracers in the National Air Quality Forecast Capability (NAQFC). The three schemes evaluated in this study represent various degrees of coupling to improve the commonality in turbulence parameterization between the meteorological and chemistry models. The methods tested include: (1) using NAM predicted TKE-based planetary boundary height, h, as the prime parameter to derive CMAQ vertical diffusivity; (2) using the NAM mixed layer depth to determine h and then proceeding as in (1); and (3) using NAM predicted vertical diffusivity directly to parameterize turbulence mixing within CMAQ. A two week period with elevated surface O-3 concentrations during the summer 2006 has been selected to test these schemes in a sensitivity study. The study results are verified and evaluated using the EPA AIRNow monitoring network and other ozonesonde data. The third method is preferred a priori as it represents the tightest coupling option studied in this work for turbulent mixing processes between the meteorological and air quality models. It was found to accurately reproduce the upper bounds of turbulent mixing and provide the best agreement between predicted h and ozonesonde observed relative humidity profile inferred h for sites investigated in this study. However, this did not translate into the best agreement in surface O-3 concentrations. Overall verification results during the test period of two weeks in August 2006, did not show superiority of this method over the other 2 methods in all regions of the continental U.S. Further efforts in model improvement for the parameterizations of turbulent mixing and other surface O-3 forecast related processes are warranted.

Impact of Consistent Boundary Layer Mixing Approaches Between Nam and Cmaq