8-months period, Nov 96 – Jun 97

1. Estimation of the contribution of mineral dust to the total aerosol depth: Particular focus on Atlantic Ocean G. Myhre, A. Grini, T.K. Berntsen, T.F. Berglen, J.K. Sundet, I.S.A. Isaksen • Use Oslo CTM2, which is an offline global chemical-transport model driven by meteorological data from ECMWF • Satellite retrievals of aerosols • Information from AERONET stations

2. 8-months period, Nov 96 – Jun 97

3. Oslo CTM2 with sea salt, mineral dust, carbonaceous (fossil fuel and biomass burning), and sulfate included. (Max AOD 1.2) November • Five satellite retrievals • One channel AVHRR • Two channel AVHRR • POLDER • OCTS • TOMS

4. Modis 2001 annual mean AOD Modelled 1996 annual mean AOD

5. Sulfate, BC, OC, Mineral dust, Sea salt are included • Model calculations performed in T63 (1.875x1.875) degrees • Dust emissions are modeled using the Dust Entrainment and Deposition model (DEAD) (Zender et al., 2003). • 8 size bins from 0.03 – 25 µm, with no hygroscopic growth included • Refractive index from the SHADE campaign (n550nm=1.47 – 0.0012i) Model description

6. SHADE Reproduced a dust storm and calculated a radiative impact of –115 Wm-2. Model AOD (550 nm) MODIS AOD (550 nm)

7. Modelled AOD compared to AERONET measurements at Cape Verde AERONET MODEL Correlation coefficient 0.74

8. Mineral dust -> Generally a cooling effect • Biomass burning aerosols -> Cooling under clear sky condition and warming above stratocumulus clouds • Internal mixture of mineral dust and biomass burning aerosols -> ??

9. Biomass burning aerosols much smaller than mineral dust

10. Ångstrøm Exponent for MODIS, POLDER and for the model MODIS shown for annual mean 2001, POLDER for Nov 1996, and the model for annual mean 1996 and Nov 1996 MODIS MODEL POLDER

11. Estimate of mineral dust fraction of AOD from satellite retrievals • Observation of Ångstrøm exponent at Cape Verde as low as 0.15 and in South Africa during the main biomass season of up to 1.85 • Modelling gives slightly lower Ångstrøm exponent for mineral dust and Ångstrøm exponentfor aerosols from biomass burning around 1.85 • Based on the difference in Ångstrøm exponent a fraction of large to small particles can be estimate • Use information from the model of sea salt to estimate fraction mineral dust

12. Fraction mineral dust contribution to total AOD MODIS MODEL Annual mean POLDER MODEL Nov

13. Summary • An estimate of the fraction of mineral dust to total AOD in the region over ocean with highest AOD • Difficult as large uncertainties in satellite retrievals and in aerosol models • We see a large gradient in the mineral dust fraction from 20N to 10S • Large potential for mixing from 10N to 0N • Potential larger warming effect from biomass burning aerosols and mineral dust if internal mixing than separate particles

16. Oslo CTM2 with sea salt, mineral dust, carbonaceous (fossil fuel and biomass burning), and sulfate included. (Max AOD 1.1) December • Five satellite retrievals • One channel AVHRR • Two channel AVHRR • POLDER • OCTS • TOMS

17. Modelled aerosol optical depth from 5 aerosol components December Mineral dust

18. Modelled aerosol optical depth from 5 aerosol components November Mineral dust

19. Ångstrøm eksponent (low values for large particles and high values for small particles)

20. MODIS

21. MODIS MODEL POLDER