1. Chemical regimes over Europe – long term, seasonal and day to day variability Matthias Beekmann LISA University Paris 7 and 12, CNRS Créteil, France Thanks to Robert Vautard for discussions and MM5 data

2. Definitions VOC sensitive regime: VOC emission reductions more favorable for ozone reduction Near intense anthropogenic emission sources Low VOC / NOx emission ratio Weak actinic flux Radical chain termination NO 2 + OH -> …. RCOO 2 + NO 2 -> …. RO 2 + NO -> …. NOx sensitive regime : NOx emission reductions more favorable for ozone reduction Rurales areas High VOC / NOx emission ratio Strong actinic flux Radical chain termination HO 2 + HO 2 -> …. RO 2 + HO 2 -> …. INTRODUCTION Environmental conditions Chemical cycles

3. PLAN Many older studies, but no comprehensive work addressing: Differences in chemical regimes with respect to target Interannual, seasonal variability Day to day variability, which forcings ? Long term variability (decades) Dependency to model configuration

4. Developed by: IPSL, LISA /CNRS, INERIS [http://euler.lmd.polytechnique.fr/chimere] Model domain : 10°W-23°E, 35°-58°N Horizontal resolution : 0.5° x 0.5° Vertical resolution: 8 layers in hybrid pressure coordinates P k =a k p top + b k p surf ; p top =500 mbar Chemical mechanism: reduced MELCHIOR ( 44 species, 120 reactions) here use of gas phase only Meteorology: MM5 driven by NCEP Reanalysis or Forecast Emissions: EMEP NO x,VOC, CO, SO 2, 10 SNAP sectors; res. 50x50 km Biogenic emissions of isoprene, pinene and NO Boundary conditions: LMD-Z/INCA, MOZART, GOCART (PM) CHIMERE CHEMISTRY-TRANSPORT MODEL

5. Results Reference simulation Daily O3 maximum Surface = 0 – 50 m Average over May – August 2001 2002 2003

6. Results (2) Emission scenarii

7. Chemical Regime = NOx -30 % minus VOC -30% emission scenario

8. Chemical Regime Emissions Industrial NOx emissions 10 10 mol. cm -2 s -1 => Strong dependence of chem. regime on emission strength, modulated by meteorology

9. Daily O 3 max. => basic pollution indicator Daily OX max. OX = O 3 + NO 2 => normalises out O 3 titration by NO AOT60  hours max [O 3 – 60 ppb, 0] => climatological health index AOT90  hours max [O 3 – 90 ppb, 0] => pollution peak indicator SUMO 35  days max [daily max O 3 8h – 35 ppb, 0] => climatological health index Chemical regimes for different targets

10. Chemical regimes for different targets maximum daily OX concentration May – August 2001- 2003 average

11. Chemical regimes for different targets SUMO35 May – August 2001 – 2003 integration/ average

12. Chemical regimes for different targets AOT60 May – August 2001 – 2003 integration/ average

13. Chemical regimes for different targets AOT90 May – August 2001 – 2003 integration/ average

14. Chemical regimes for different targets Conclusions General structure keeps unchanged irrespective of target : NW EUVOC sensitive SE EUNOx sensitive North ItalyNOx sensitive with exceptions Shipping trackVOC sensitive

15. Are these model results robust ? Model has been evaluated with ozone measurements over Western Europe => Small bias, RMSE ~ 20%, R ~ 80% over WE But no garantee, that sensitivity of P O3 to E VOC and E NOx is correct Ideal solution: Monte Carlo simulations => give global uncertainty Here sensitivity tests * EVOC + 40 % * other chemical mecanism: extended Melchior (~80 compounds, 320 reactions) instead of reduced one (40 compounds, ~120 reactions, operator approach) * look at model layer 2 (50 – 200 m) instead of layer 1 (0-50m) => regions with complex vertical structure

16. Robustness test O3max Many structures absent

17. Robustness test AOT90

18. Interannual variability O3max

19. Seasonal variability O3max

20. The heatwave case August 2003 O3max AOT90 O3max and AOT90 : NOx sensitive regime more extended

21. Day to day variability of chemical regimes O3max Integration over the strongly VOC sensitive region in NW – Europe From May to August 2001

22. Day to day variability of chemical regimes O3max  Chemical regime is always VOC sensitive over North-Western Europe  No dependency on ozone levels How can day to day variability be explained ?

23. Day to day variability of chemical regimes O3max Dependence on emission accumulation (NOy) Dependence on classical chemical regime tracers O3 / NOz ratio

24. Chemical regime for 1980 – 2020 emissions Is the chemical regime expected to change for future emission scenarii ? Did it change in the past ? =>Use EMEP 1980, 1990, 2010 and 2020 expert emissions in addition to 2002 one’s => No change in boundary conditions, trends not clear

25. Chemical regime for 1980 – 2020 emissions O3max 1980 1990 20102020

26. Chemical regime for 1980 – 2020 emissions AOT90 Year 1980 Year 2020

27. Conclusions Chemical regime is (in CHIMERE European CTM) VOC sensitive over North Western Europe, NOx sensitive over Mediterranean region and Eastern Europe This is robust with respect to : * target * model uncertainty * interannual, seasonal, day to day variability Northern Italy : NOx sensitive, with exceptions (Milan area, …. exceptions not so robust) Shipping emissions : VOC sensitive, but complex vertical structure 1980 to 2020 emission changes drive system to more NOx sensitive in NW EU, in North Italy