Air quality forecasts in the mid-Atlantic region: Current practice and Benchmark skill

Air quality forecasts for the mid-Atlantic region (including the metropolit an areas of Baltimore, Washington, D.C., and Philadelphia) began in 1992. T hese forecasts were issued to the public beginning in 1995 and predict dail y peak O-3 concentrations (1-h average) within each metropolitan area. The purposes of the forecasts are to warn sensitive populations of concentratio ns that are in excess of the National Ambient Air Quality Standard (NAAQS) for O-3 and initiate voluntary control programs (''ozone action days") desi gned to reduce pollution. Ozone is a photochemical pollutant whose concentr ations reach a maximum during the summer months when day length is long and solar zenith angle low. Forecasts are issued daily from mid-May to mid-Sep tember at approximately 1900 UTC and are valid the Following day. The forec asts are based on statistical models that use primarily meteorological pred ictors. Output from the statistical model is used as guidance and modified by the forecasters to account for features not resolved by those methods. T he forecast is issued to the public in me form of a color code with "code r ed" indicating unhealthy conditions. The range of peak O-3 concentrations i n the mid-Atlantic region during a given season is typically 30-180 parts p er billion by volume (ppbv). For the period of this study (1995-98) the Bal timore forecast area observed seasonal mean peak O-3 of 85.7 ppbv. Median a bsolute forecast error far the 1995-98 seasons was 9.0 ppbv and root-mean-s quare error was 14.8 ppbv. This represents a 40% increase in skill over sim ple persistence forecasts. Nine of the 10 most severe cases during this per iod were correctly forecast code red with the remaining case forecast "code orange" (O-3 watch). Currently, real-time photochemical models are being d eveloped to forecast O-3. The results presented here represent benchmark sk ill from which to judge improvements occasioned by numerical models or othe r forecasting techniques.