West Florida Ozone Study (WFOS): An Integrated Meteorological and Air Quality Data Analysis and Modeling Study for West Florida

The West Florida Ozone Study (WFOS) was designed to provide detailed analysis of 8-hour ozone issues for Pensacola and other areas located along the Florida panhandle. WFOS is an extension of the Gulf Coast Ozone Study (GCOS) and includes the integrated analysis of observed data and modeling results for the purpose of developing a detailed conceptual description of 8-hour ozone of the area. The modeling analysis also included the use of a new technique, the Ozone and Precursor Tagging Methodology (OPTM), for assessing the contribution of emissions sources to simulated ozone concentrations, process-based analysis of the modeling results to examine the mechanisms contributing to high simulated ozone concentrations, and application of the EPA 8-hour ozone attainment demonstration procedures with consideration of meteorological representativeness.

In this study, we used a combination of data analysis and modeling to examine the causes of high ozone along the Florida panhandle, and to investigate the effectiveness of various control measures in reducing future-year ozone concentrations and maintaining the 8-hour ozone standard.

The meteorological and air quality data suggest that a key factor contributing to the incidence of high 8-hour ozone concentrations is the “gulf breeze”. By prolonging the period over which the ozone concentrations are relatively, yet not extremely, high the gulf breeze can contribute to a high 8-hour average ozone concentration. This finding is also supported by model-based process analysis. The sustained relatively high levels of ozone during the daytime hours are due to a combination of photochemistry, vertical diffusion, and horizontal advection; the contributions from these processes are timed such that relatively high ozone levels are sustained over eight or more hours. Process analysis results also suggest that transport of ozone into the area is a key factor in the exceedances for certain coastal areas.

For the WFOS modeling analysis, the UAM-V Ozone and Precursor Tagging Methodology (OPTM) was used to examine the contributions from selected emission source categories and source regions to simulated ozone for a 2007 baseline simulation. Source categories that were explicitly tagged include point, on-road mobile, non-road, and area sources within the Pensacola two-county area, as well as electric generating unit (EGU) and non-EGU sources located in southern Alabama and Mississippi. Several source regions were also explicitly tagged. The tagging results indicate that, on average, the high simulated 8-hour ozone concentrations in the Pensacola area are the result of local, regional, and extraregional (including background) contributions. For NOx emissions, the relative contributions from these three general categories represent roughly 40, 20, and 40 percent of the average simulated maximum 8-hour ozone concentrations. For VOC emissions, local anthropogenic emissions contribute about 10 percent, while regional and biogenic emissions comprise the majority of the source-category/source-area VOC contributions.

Future-year estimated design values (EDV) for the WFOS areas of interest were calculated for the future-year baseline and control-strategy simulations. Of interest, is the use of a “meteorologically adjusted” design value for the Pensacola area. The adjustment is based on frequency of occurrence of certain types of meteorological conditions, rather than severity of these conditions. The results indicate that a meteorologically adjusted design value is much more stable than the observation-based design value. For the 1996-2003 analysis period, the high design values appear to be attributable to more persistent than usual ozone conducive meteorological conditions. Similarly the low design values are attributable, using this methodology, to fewer than usual days with ozone-conducive conditions. The 2001 design value was used in the WFOS modeling analysis as the basis of the modeled attainment test and is representative of the meteorologically adjusted value for that year and the full analysis period. In this regard, it provides a satisfactory, indeed best-case, basis for the application of the modeled attainment test.