The New Generation of Air Quality Modeling Systems
Pleim, Jon, David-C Wong, R. Gilliam, J. Herwehe, R. Bullock, C. Hogrefe, G. Pouliot, L. Ran, B. Murphy, D. Kang, W. Appel, R. Mathur, AND E. Hubal. The New Generation of Air Quality Modeling Systems. ENVIRONMENTAL MANAGEMENT. Springer-Verlag, New York, NY, , 1-6, (2018).
To support the implementation of the Clean Air Act (CAA) and to assess ecosystem impacts, this product supports development of an advanced model system with extended capabilities. The system will feature the flexibility to couple air quality specific processes to multiple meteorology/dynamics models. With this flexible design we will be able to couple with the new generation of global meteorology models that uses seamless grid refinement. The prototype system will include the capability to couple with the Model for Prediction Across Scales (MPAS), which is a global model with seamless mesh refinement that is developed and supported at NCAR.
The nature of air quality (AQ) problems in the US has evolved over the past several decades from intense local smog episodes in cities and industrial areas to widespread increasing background concentrations with less severe but persistent hotspots. While emissions of pollutants and their precursors in the US have declined significantly in recent decades, other parts of the world have seen substantial growth in air pollutants as their economies and industries have grown rapidly. Pollutants emitted near the Earth’s surface can be convectively lofted to the free atmosphere where strong winds efficiently transport them across continents and oceans. The recognition of the importance of atmospheric-process interactions among global, regional, and local scales inspired international cooperative research efforts such as the Task Force on Hemispheric Transport of Air Pollutants (TF-HTAP) and the third phase of the Air Quality Model Evaluation International Initiative (AQMEII3). Thus, the focus of AQ modeling has shifted over the years from limited area models which focus on urban areas to more regional domains and recently to multiscale systems that cover the entire globe. Typically, regional AQ models such as the Community Multiscale Air Quality (CMAQ) model1, the Weather Research and Forecasting model with Chemistry (WRF-Chem), Comprehensive Air Quality Model with Extensions (CAMx), and CHIMERE use lateral boundary conditions (LBC) from global AQ models such as the Goddard Earth Observing System with Chemistry (GEOS-Chem), Model for Ozone and Related Chemical Tracers (MOZART), Atmospheric Model version 3 (AM3), and Composition Integrated Forecasting System (C-IFS). However, recent studies2 have shown that LBCs derived from different global models have profound effects on simulated regional-scale ozone in the continental US (CONUS). It was found that biases inherent in global models propagate into the regional modeling domain. Also, additional uncertainty results from inconsistencies in process representations, chemical species mapping, and grid structures between the global and regional models. Therefore, the CMAQ development group at the EPA has developed a hemispheric application of the WRF-CMAQ system that has become the preferred practice for supplying LBCs to regional and finer scale WRF-CMAQ applications3. While using the same model to scale down from hemispheric minimizes many inconsistencies and sources of error, the multi-scale nesting approach, as shown in Figure 1, still includes spatial and temporal interpolation errors at each step of grid refinement.