Clouds are one of the most difficult meteorological inputs to air quality modeling applications to accurately replicate. The widely varying horizontal and vertical scales of clouds are often ill-suited to Eulerian grid modeling applications. Errors and uncertainties in this modeling input can translate to large errors and uncertainties in the air quality modeling outputs.

As part of two separate regional ozone and fine particulate modeling applications over the eastern U.S. for the years of 2001 and 2002, this analysis attempts to answer three questions with regard to clouds and their attenuation effects on air quality: 1) can the effects of cloud attenuation on air quality be isolated and quantified; 2) how and why do various air quality models differ in their treatment of cloud attenuation; and 3) what uncertainty bounds are inherent in air quality modeling applications solely from the treatment of cloud attenuation.

Both the Community Multi-scale Air Quality modeling system (CMAQ) and the Comprehensive Air Quality Model with Extensions (CAMx) have been applied to a large eastern U.S. domain for two model years. In both cases, the raw meteorological modeling input data were based on simulations with the Penn State University / National Center for Atmospheric Research mesoscale model known as MM5. Several model simulations were conducted with the effects of cloud attenuation off or otherwise modified. These runs showed that even relatively optically-thin clouds could have large suppressing effects on ozone (up to 40 ppb) and other photochemically activated pollutants. This analysis also showed that the two air quality models that are primarily used to support regional air quality planning differ greatly in the way they simulate the attenuation effects of clouds. CMAQ uses an empirical formula for cloud optical depth based on data from Stephens (1978) where the key parameters are mean cloud liquid water and cloud depth. CAMx uses a cloud optical depth equation from the Regional Acid Deposition Model (RADM) that is essentially a linear relationship between the same parameters: mean cloud liquid water and cloud depth. However, the varying equations result in quite different estimates of cloud optical depth with CMAQ generating stronger-attenuating clouds than CAMx for most combinations of cloud liquid water path. Assuming that both model schemes are equally valid representations of cloud optical depth, it is possible to quantify the uncertainty in air quality outputs resulting from the different approaches.