PREV ’AIR :An operational system for air quality monitoring and forecasting
www.prevair.org
presented by Laurence Rouïl
PREV ’AIR : www.prevair.orgA cooperative system for large scale air quality forecasting and mapping over Western Europe and France ; developed in 2003 upon an initiative of the Ministry of EcologyOperational purposes : in France, since 2004, public information related to pollution episodes (threshold exceedances) can be driven by forecasts (not only observations)
INERIS
ADEME
CNRS
Météo Franc
e
Public organization under the supervision of the Ministry of
Ecology
National Agency of Environment
National Research Centre
National Meteorological services
Builds up and hosts the prev ’air systemAQ expertise for the Ministry
Builds up and hosts the NRT databaseAQ expertise for the Ministry
Model development : CHIMERE
Model development : MOCAGE +meteorological forecasts
Architecture of the PREV’AIR Modelling System
3D Chemistry Transport Models...• CHIMERE / IPSL-CNRS-INERIS• MOCAGE / METEO-France• open system to other models...
… Driven by meteorological forecasts• AVN / NCEP global data (+ MM5
higher resolution forecasts)• ARPEGE, ALADIN
Other input data• Emissions inventory (EMEP)• BC (MOZART)• Land-use (GLCF data base)• NRT observations data from local AQ
monitoring organizations (AASQA) ; bilateral cooperation with other countries
Chemical forecasts
Low and high resolution meteorological forecasts
Numerical AQ forecast data
Forecasts (maps)
BASTER
Analysis (maps)
Observations (maps)
Emilia Romagna (ARPA-EMR) Swiss (NABEL/SAEFL) Germany (UBA)... A S QA A
The CHIMERE Model Set-Upin PREV’AIR
• Domains Over Europe Over France
Horizontal resolution: 0.5°x 0.5° 0.15° x 0.1° Vertical resolution: 8 levels from surface pressure up to 500 hPa
• Meteorological forecast data AVN / NCEP for initialisation and boundary conditions MM5 higher resolution forecasts ARPEGE
• Chemical scheme MELCHIOR (47 species, 125 reactions)
• Aerosol module 7 species (PPM, SOA, nitrate, sulphate, ammonium, Water Contents, dust) 25 reactions (aqueous and heterogeneous phase)
See http://euler.lmd.polytechnique.fr/chimere.
The MOCAGE Model Set-Upin PREV’AIR
Domains» Global Model / Over Europe / Over France
Horizontal resolution: 4°x 4° 0.5° x 0.5° 0.1° x 0.1° Vertical resolution: 47 levels from surface pressure up to 5hPa
Meteorological forecast data ARPEGE and ALADIN
Chemical scheme RACMOBUS (118 species, 381 reactions)
No aerosol module
• Available at D+0, 00 h LT
• Daily peak and averaged concentration maps
D+0, D+1, D+2
• Pollutants: O3, NO2 (France and Europe)particulate matter (Europe)
O3, peak forecast (µg/m3)
20040729 D+0
O3, peak forecast (µg/m3)
20040729 D+2
O3, peak forecast (µg/m3)
20040729 D+1
PREV’AIR Outputs : Daily Forecasts
Modelled concentrations corrected with observations: Observations collected in real-time for 150
stations Statistical adaptation using a kriging method Available at J+0 with updates if near real time
observations available Feasibility study for the PM
PREV ’AIR Outputs : Ozone Analyses for the Day Before and the Current Day
July 31, 2004
O3 simulation (µg/m3) O3 corrective field (µg/m3)
03 simulation corrected by observations (µg/m3)
PREV ’AIR Outputs : French maps of near real time observations
Daily mean concentrations of PM10 at surface measured on the 12 of September 2004
maps of daily max and mean
measured concentrations :
for O3, NO2, PM2.5 and PM10
using observations made
between 0h and t
hour by hour updating if data
available (pollution events)
for J and the 6 previous days
PREV’AIR Outputs : Scores
PM10 scores
PREV’AIR Outputs : Scores (ii)
PREV ’AIR Outputs : Annual BudgetsOzone, averaged peak (µg/m3)
during summer 2003Ozone, # of hours of exceedance of the
180 µg/m3 threshold during summer 2003
Ozone, # of hours of exceedance of the 240 µg/m3 threshold during summer
2003
PREV ’AIR Outputs : Numerical Forecast Data
• PREV’AIR numerical data Extraction over user-defined domains Available through the Internet for AQ related institutions (user accounts)
• Applications 3D data: BC for local AQ simulations : ESMERALDA (Ile-de-France area), AERES (PACA region), … 2D data: for local AQ monitoring and forecasting (interpolation, statistical modelling)
• Feed-back from PREV’AIR users Meetings with AQ networks for system improvement Day-by-day debriefing about forecasts
Towards a european framework….
• Bilateral cooperation for data exchange (NRT) and simulated results evaluation
• Involvement in the GMES programs:PROMOTE (ESA) : demonstrator (phase 1), definition of air
quality services related to forecasting and mapping (phase 2), definition of a cooperative framework (with other european AQ teams)
GEMS (UE) : Contribution to the Air Quality part
Developments planned (1)
July 2005 : Implementation of data assimilation methods of ozone ground level measurements in the PREV’AIR system Improvement of forecasts and mapping.
August 2005-2006 : feasibility study related to the operational acquisition of ozone data from LIDARs or of satellite data in the PREV’AIR system. Improvement of the whole system taking into account 3D information
End 2005 : Feasibility of a cooperative exchange network related to near real time AQ data as this is done for France in the PREV’AIR system Near real time air quality maps in other countries?
Developments planned (2)
Implementation of nested methods to couple the models MOCAGE (applied over the whole Europe) and CHIMERE (applied over the western part of Europe) in order to obtain more realistic boundary conditions for PREV’AIR forecasts.
Case study to generate « ensemble » forecasts over France and Europe issued from the results given by MOCAGE and CHIMERE (multi-models approach)
Conclusions on the possibility, relevance, and feasibility of the implementation of a common and european air quality forecasting system in cooperation with the other PROMOTE partners
www.prevair.org
Thank you for your attention!