|
Abstract:
|
The goal of this project is to more accurately model the aerosol size distribution with the Community Multiscale Air Quality modeling system (CMAQ). The approach is to compare its performance to observations in the Pacific Northwest and to make improvements to the model science.The official CMAQ v4.4 underpredicts the total particle number concentration by 1-2 orders of magnitude. The bias is consistent throughout the day and across the urban-influenced region. It becomes progressively worse for smaller sizes and is not associated with any particular chemical species, emissions source, or air mass aging. Errors in total aerosol loading, meteorology, and gaseous aerosol precursors do not show a pattern consistent with the number underprediction. Of all the aerosol processes that create and destroy particles in the urban environment, the nucleation of new particles and the emission size distributions in the official CMAQ v4.4 attract attention for their scientific deficiencies and their tendency to produce the observed errors.The latest mechanisms for ternary NH3-H2SO 4-H2O nucleation and nucleation mode scavenging and growth are added to CMAQ, and the emission size distribution is updated to reflect modern mesoscale measurements. Modeled particle concentrations increase substantially, but they are still underpredicted by up to an order of magnitude. Nucleation changes are responsible for most of the increase but are also responsible for spatially inconsistent performance. Emissions updates increase the number of particles smoothly across urban-influenced areas by a factor of 2-4. The modeled size distributions, especially in the ultrafine range, are a better match to observations, although errors in the accumulation mode remain. The fact that these changes make a noticeable improvement in results adds weight to the premise that regional nucleation occurs regularly during the Puget Sound summer and that emissions of Aitken mode particles are an important component to the ambient aerosol size distribution. Although the updates to CMAQ represent only our incomplete understanding of aerosol pollution, they are able to reduce the underprediction of aerosol number concentrations and produce size distributions with the appropriate major features. |