We have linked several models for use in evaluating the impacts of emission reductions on air quality including effects of feedback to meteorology. The models include modified versions of the MM5, CMAQ, and Streamer in addition to emission models and processors. The linked system provides a tool that is readily available for use by regulatory and environmental quality agencies with existing episodic meteorological data and emission inventories, as well as commonly used meteorological and photochemical models.

For each defined time interval during a simulation, the linked modeling system iterates through meteorological, emissions, photochemical, and radiative-transfer computations and provides radiative-forcing feedback to the meteorological model to initiate the next iteration loop. To date, our main concern for including an additional radiative-transfer calculation step following that for air-quality is to evaluate the impacts of perturbations in tropospheric ozone on radiative forcing and, thus, the corresponding feedback to meteorology. Scenarios of interest include emission control, e.g., during ozone air-quality action days, and associated anthropogenic-heat emission reductions.

We have applied the linked system to a July 24-29 (1998) ozone episode in the greater Portland Oregon region. While perturbations in ozone can have different impacts at different altitudes in the atmosphere, our sensitivity analysis suggests that perturbations in near-surface ozone can cause daytime heating/cooling rates in the order of 0.017 Wm-2 / ppbV. In complex and highly-dynamic scenes, this relationship is altered by the competing effects of gases, aerosols, and water vapor and their spatio-temporal variability, land-use/land-cover distribution in the area, surface albedo, and surface temperature. In such complex scenarios, we found that the impacts of changes in the radiative forcing field on ozone (via impact on meteorology and emissions) cause no changes in the general pattern of the simulated field (e.g., relative to simulations without forcing feedback), but that there can be some differences near urban areas. In those regions, the geographical area affected by some of the higher concentrations increases in size and there are changes of about 3 ppbV in peak ozone during these days of interest.

In the next stages of this study, we will further evaluate these results and sensitivities in light of further model improvements, sensitivity scenarios, and model performance evaluation.