3.1 Implementation of Real-Time Bias-Corrected O_3 and PM_(2.5) Air Quality Forecast and Their Performance Evaluations During 2008 over the Continental United States

摘要

Real-time bias-corrected O_3 and PM_(2.5) forecast systems are implemented using the Kalman Filter, combining observations from AIRnow and outputs from the NOAA/EPA's NAM/CMAQ air quality forecast model. Bias-corrected O_3 and PM_(2.5) forecasts are created at locations of the AIRNow monitoring network where report hourly concentrations of these species. Observations and model outputs from two previous consecutive days are required to produce bias-corrected model forecasts. The performance of these systems is examined on a daily basis using O_3 and PM_(2.5) observations and the results are compared with raw model forecasts. The overall performance of the Kalman filtering technique and its capability to produce a real-time bias correction to improve the day-to-day forecast from the NAM-CMAQ modeling system during 2008 is investigated. Performance evaluation trough detailed time-series analysis and regional analysis will be presented. The ability of the technique in improving the prediction of daily 8-hr maximum O_3 and daily mean PM_(2.5) as well as its impacts on false-alarms will be examined through the use of statistical categorical metrics.