PREV'AIR: An Operational Forecasting and Mapping System for Air Quality in Europe.

The current state of the art in three-dimensional chemistry-transport models allows them to be considered as mature and reliable enough to be combined with observations networks for implementing integrated air quality monitoring systems over large territories. A cooperative initiative of tour research and operational organizations in France has led to the creation of an integrated air quality platform providing near-real-time and forecasted information using last model developments. Since summer 2003, this system, Prévisions et Observations de la Qualité de l'Air en France et en Europe (PRBV AIR; available online at www.prevair. org), has been delivering daily air quality observations short-term (0-3 days) forecasts, and real-time analyses of ozone, nitrogen dioxide, and particulate matter (PM) at the global and European scales, with a focus over France. It has become a powerful scientific and communication tool to inform the public, sensitive population about the potential occurrence of air pollutant concentrations exceeding regulatory thresholds. The architecture of the operational system is described in this paper. Its setup allows the partners to cope with stringent operational constraints regarding the timely availability of the forecasts and the reliability and the quality of the products delivered. The system relies on a chain of numerical models and methods, and on real-time air quality observations, it uses National Centers for Environmental Predictions (NCEP) and Météo-France weather forecasts. Methods used for producing air pollutant concentration forecasts and analyses are described, as well as the final products, with case examples. A synthesis of the operational system skill results obtained over three consecutive years of operation (2003-06) is given. Lessons learned are presented and issues related to the use of the PREV'AIR system as a complementary tool for air quality management are discussed. [ABSTRACT FROM AUTHOR]