![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081013072904im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Performance Evaluation of the Community Multiscale Air Quality (CMAQ) Model
Research Programs
In support of the CMAQ releases, a series of model evaluation analyses will be conducted for an annual simulation over the continental U.S. domain. Model evaluation serves both the purpose of establishing model performance and identifying areas where model improvements or further testing are needed. These are typically referred to as performance evaluation and diagnostic evaluation approaches, respectively. For the performance evaluation, comparisons against observational networks are needed to provide data across the continental U.S. domain for the entire year. CMAQ predictions will be compared against speciated aerosol data from the Clean Air Status and Trends Network (CASTNet), Speciated Trends Network (STN), and the Interagency Monitoring of Protected Visual Environments (IMPROVE) network; precipitation chemistry data from the National Atmospheric Deposition Program (NADP) network; and ozone data from the USEPA Aerometric Information Retrieval System (AIRS). To quantify these comparisons, statistical measures will be calculated including the correlation coefficient, root mean square error, mean bias, normalized mean bias, and normalized gross error. These metrics will be calculated on all data within the year combined, separately for each season or month, and for subsets of the modeling domain (e.g., East versus West). Spatial maps will also be generated of both the concentration fields and specific comparative measures or statistics to provide information about the model performance across the domain. To provide additional spatial information about the model predictions for aerosols, aerosol optical depth (AOD) values will be calculated from the CMAQ fine particulate matter (PM2.5) predictions and compared to AOD satellite product from the Moderate Resolution Imaging Spectroradiometer (MODIS). For the 2005 annual release, one area of diagnostic evaluation that will be included is an analysis of nitrate replacement. Nitrate replacement refers to the scenario where reduction of SO2 emissions leads to less sulfate aerosols; however, more ammonia would then become “free ammonia” that can react with nitric acid to form nitrate aerosols. In this scenario, the total PM2.5 concentration may not be substantially reduced because you have increased the concentration of nitrate aerosols while decreasing the sulfate aerosol concentration.