Evaluation of the CMAQ and GMI Model-Simulated Shape Factors with DISCOVER-AQ Observations with Implications for Retrievals

The NASA Earth Venture-1 DISCOVER-AQ (Deriving Information on Surface conditions from Column and Vertically Resolved Observations Relevant to Air Quality) project took place over four years and was comprised of four deployments: the Baltimore-Washington metropolitan region of Maryland during July 2011: the San Joaquin Valley of California during January-February 2013; the Houston, TX, metropolitan area during September 2013; and the last in the Front Range region of Colorado during July-August 2014. The P-3B aircraft provided in situ vertical profiles of meteorological quantities, trace gases, and aerosols over surface air quality monitoring sites during flight days for each deployment. A major goal of DISCOVER-AQ is to understand the processes linking column abundances to surface concentrations for O3 and NO2, which includes understanding the typical profile shapes and the atmospheric conditions influencing these profile shapes. Additionally, these data may be used to evaluate the ability of air quality models to replicate the observed profile shapes. This has implications for the global models currently used in remote sensing retrievals of vertical column abundance as well as for the use of regional air quality models within retrievals. The regional CMAQ and global GMI atmospheric chemistry and transport models are evaluated against the observed P-3B lower tropospheric shape factors (the fraction of the column within each vertical layer) for the clusters of typical profile shapes obtained through an agglomerative hierarchical clustering technique for O3 and NO2. The CMAQ model captured the shape factors for O3, and moderately well captured the NO2 shape factors, for the conditions associated with the Maryland campaign, suggesting that a regional air quality model may adequately specify a priori profile shapes for remote sensing retrievals for the conditions associated with this deployment. Coarser vertical resolution in the GMI model affected the ability of that model to reproduce the observations. These results demonstrate the importance of resolution for accurate representation of pollutant profiles as a priori information within satellite retrievals, and for the ability to relate column abundances to surface concentrations.