Technical Note: Evaluation of the WRF-Chem "Aerosol Chemical to Aerosol Optical Properties" Module using data from the MILAGRO campaign

Abstract

A comparison between observed aerosol optical properties from the MILAGRO field campaign, which took place in the Mexico City Metropolitan Area (MCMA) during March 2006, and values simulated by the Weather Research and Forecasting (WRF-Chem) model, reveals large differences. To help identify the source of the discrepancies, data from the MILAGRO campaign are used to evaluate the "aerosol chemical to aerosol optical properties" module implemented in the full chemistry version of the WRF-Chem model. The evaluation uses measurements of aerosol size distributions and chemical properties obtained at the MILAGRO T1 site. These observations are fed to the module, which makes predictions of various aerosol optical properties, including the scattering coefficient, <i>B</i>_scat; the absorption coefficient, <i>B</i>_abs; and the single-scattering albedo, &piv;₀; all as a function of time. Values simulated by the module are compared with independent measurements obtained from a photoacoustic spectrometer (PAS) at a wavelength of 870 nm. Because of line losses and other factors, only "fine mode" aerosols with aerodynamic diameters less than 2.5 μm are considered here. Over a 10-day period, the simulations of hour-by-hour variations of <i>B</i>_scat are not satisfactory, but simulations of <i>B</i>_abs and &piv;₀ are considerably better. When averaged over the 10-day period, the computed and observed optical properties agree within the uncertainty limits of the measurements and simulations. Specifically, the observed and calculated values are, respectively: (1) <i>B</i>_scat, 34.1&plusmn;5.1 Mm⁻¹ versus 30.4&plusmn;3.4 Mm⁻¹; (2) <i>B</i>_abs, 9.7&plusmn;1.0 Mm⁻¹ versus 11.7&plusmn;1.2 Mm⁻¹; and (3) &piv;₀, 0.78&plusmn;0.05 and 0.74±0.03. The discrepancies in values of &piv;₀ simulated by the full WRF-Chem model thus cannot be attributed to the "aerosol chemistry to optics" module. The discrepancy is more likely due, in part, to poor characterization of emissions near the T1 site, particularly black carbon emissions.