Aerosol simulations over a Euro/Med/African domain with RegCM

1. Aerosol simulations over a Euro/Med/African domain with RegCM F. Solmon, A. Zakey, F. Giorgi, C. Liousse, A. Konare … and the ICTP RegCM team … • Introduction / motivations • Tracers and aerosols in RegCM • Simulation over Eur-Med-Af domain • Some comparisons with data … • Specific studies over West Africa / AMMA • Conclusion

2. Aerosols … Feb 2004, MODIS

3. Tropospheric aerosols : Importance of regional scale processes • First step : simulation of sources, physico-chemical nature and atmospheric content of aerosol • Second step : radiative forcing, climatic and ecological feedback

4. General approach Tracer model / RegCM3 (from Giorgi et al., Qian et al.) Primary Emissions Removal terms Physico – chemical transformations Transport Strongly dependent on the nature of the tracer Aerosols in RegCM3 Particles and chemical species considered.

5. χi= χcu Transport of tracers χ1 Advection / diffusion Cloud mmr ( mm5 options) χi Wet removal by large scale rainfall: χ1 (Giorgi et al., 1989) Wet removal by cumulus convective rainfall: w,cum~ 20min Aerosols processes Convective transport Based on Grell scheme mass fluxes Simple mixing hypothesis Dry deposition : 1 prescribed deposition velocities ( nature tracer / surface) 2 on-line settling and dry deposition.

6. Sulfur Aerosol Model(Kasibhalta et al., 1997, Qian et al., 2001) Gas phase: Aqueous phase: [OH] : constant profile + diurnal evolution (max for cos = 1) Aging of carbon aerosol (Cooke et al.,1999) hydrophobic hydrophilic Simple approach : aging= 1,15 days Deposition(dry, wet) Optical properties CCN Aerosols processes

7. Application Domain Grid : 147*165 Res : 60 km Rotated Mercator DJF 1999-2000 LS forcing : NNRP2 JJA 2000

8. SO2 sources 1995 EDGAR, 1 deg, annual, Fossil fuel + 2000 EMEP emission over Europe a SOx (ng m-2 s-1) SOx ED (ng m-2 s-1)

9. Fossil fuel : 1997, 1 deg, annual BC OCtot • Hydrophobicity at the emisson Biomass burning : 80-90’s 3.75*5 deg, monthly • Forest • Savanna • Agriculture • Cleaning • Domestic fires 80 % 50 % 20 % 50 % Injection height Carbonaceous aerosol (Liousse et al., 1996) • BC Updated emission factors • OCtot= OCprim + OCsec

10. BC-OC sources Jan Jul BC OC

11. Soil texture (12 types, USDA) Soil erodible dry agregates distribution Land surface properties (BATS) ( roughness, soil humidity, cover fractions) Regcm atmospheric variables ( surface wind, air temperature, air density) Aerosol dust source and deposition parameterization in RegCM Input parameters DUST emission scheme See talk Dr. A. Zakey Saltation (Marticorena et al. 1995) Roughness and humidity correction Sand-blasting (Alfaro et al., 1997, 2001) Dust flux distribution (3 log-normal emission modes) 0.01 -1 μm 1 - 2.5 μm 2.5 - 5μm 5 - 20 μm 4 Transport bins Size dependant settling and surface déposition AOP / radiation

12. Ex of transport : BC Dec 1999

13. DUST emission and transport, Feb 2000

14. Surface concentration validation: SO2/SO4 SO2 RegCM vs. EMEP Estimation over West Africa ? SO2 Abidjan: 2 ppb Dakar: 6 ppb ( Yoboue, Ndiaye, Lacaux P.C. ) RegCM ~ 0.3 ppb ( max ~ 3 ppb over Lagos) SO4 Underestimation Measurements / Emission Inventories over African urban areas DJF JJA

15. Surface concentration validation: BC and OC BC OC Estimation over West Africa ? Very few surface concentrations measurements points Urban emissions inventories DJF JJA RegCM vs Bilbliographic data base (Mostly Europe)

16. RegCM / prognostic aerosol fields Remote sensing measurments Phys_chem nature, mass and size distribution the of aerosols Optical properties “Optical properties ” Aer. Opt. Prop. in RegCM Assuming Log-normal dry size distribution • Extinction coefficient • Single scattering albedo • Assymetry parameter Correction for RH Off-line Mie calculations External mixing !

17.   For 16 CCM3 spectral intervals BChl , OChl RH effect

18. RegCM3 / observations Monthly Aerosol Optical Thickness (AOD) (integration of vertical extinction) AERONET Ground based photometer network DATA MODIS MISR Satellite ( since march 2000) Data in free download access

19. RegCM AOD (550nm) vs. AERONET AOD Sahel : Complex mixture / interaction dust / biomass burning aerosol

21. Over West Africa • AERONET ( few stations) + MODIS JJA : Spatial patterns show consistency but underestimation of magnitudes specifically in biomass burning dominated regions. • Underestimation of burdens  Underestimation of biomass burning emissions (other anthropogenic sources might be underestimated here). • biogenic SOA aerosols ( tropical region) • Aerosols properties ( hygroscopicity/wet removal, optical properties) To go further in the validation Surface and vertical concentrations measurments Wet deposition. Use of different wavelenghts to discriminate aerosol e.g extension to optical parameters ω, g / remote sensing (diffusive or absorbtion properties of an aerosol layer  chemical nature)

22. Exemples of applications / specific studies Budgets Perspetives

24. Region B

25. Further interest of budgets E.G differentiate and quantify the effect of different sources contributing to the actual aerosol content over a given area Indicator for experimentation (flight)

26. Perspectives

27. Simulated JJA dust distribution (yellow isocontour) and cloud distribution JJA Wet deposition (cumul) here linked to convective clouds SAL/ICT South

28. Simulated JJA OC distribution (yellow isocontour) and cloud distribution Validation with IDAF data

29. Conclusions / Perspectives • RegCM3 : sulfates carbonaceous dust aerosols Over the African part of the domain • Consistent results obtained in the simulation of AOD fields : • Underestimation in biomass burning regions / use of updqted and finer resolution inventories • More validations : AMMA Land use change / Aerosols .climatic feedback studies • Introduction sea salt and biogenic SOA • Log-normal or bins model + gazeous chemistry

30. Thank you

31. Ex 1 : SO2 and SO42- DEC1999 Column Burden evolution (mg.m-2) Isocontour = SO2 Average SO42- concentration field (μg.m-3) and column burden