1. Modelling of Soluble Iron Formation Transport and Deposition to the North Pacific. Anthropogenic impacts. F. Solmon (1,2), P. Chuang (1), N.Meskhidze (3) (1) University of California Santa Cruz, CA (2) Laboratoire d’aérologie, Toulouse, France (3) North Carolina State University, Raleigh, NC

2. Why study atmospheric iron cycle ? Atmospheric deposition of dust has been considered as the main source of iron for open ocean. At the source, dust iron (~ 3.5 % in mass) is mostly unsoluble i.e not bioavailable At the source, the Dissolved Iron Fraction : DIF ~ 0 - 1 % Measurments at remote sites show that DIF increases during atmospheric transport : DIF ~ 1-30 % with a large variability. Modeling effort to better to characterize DIF (e.g Fan et al., 2006; Luo et al., 2007) Possible anthropogenic influences on soluble iron: Atmospheric processing of dust Released by combustion activities North Pacific Ocean Iron is an important nutrient for marine ecosystems. In High Chlorophyll Low Nurient regions, iron input stimulates biological productivity and CO 2 pumping => Climate / Paleoclimate studies.

3. Acidic attack H 2 SO 4, HNO 3, HCl, carbonates H2OH2O Alkaline compounds, NH 3 2: Scavenging of soluble gaz species by dust Uptake coeff Therm. Equilibrium (HNO3, NH3), ISORROPIA Iron dissolution modelling Dust alkaline Meskhidze et al., 2005 Fe 3 : Mineral dissolution (production of dissolved iron) Kinetics processes (calcite, …, hematite) pH sensitive 4 : Chemical speciation, Thermo. Equilibrium, pH Specific mechansims : eg Ca 2+ / CaSO 4 ISORROPIA GEOS-CHEM Anthro. aerosols O 3 1: Assume an initial mineral composition for the dust

4. Dissolved iron modelling 11 new tracers in GC representing mineral species in the dust mode : (Fe, Ca, Al, Na, Sil, K, Mg, SO 4 2-, NO 3 -, NH 4 ) aq, (CaCO 3 ) s Dissolved iron FEDI (oxydation III) One mode representative of dust (aggregation of bins 1 and 2) Tracers are transported and removed (wet and dry dep) as dust in GC At the source : FETOT = 3.7 % * DUST ; DIF = FEDI / FETOT = 0.45 %

5. Test Case Simulation (2 x 2.5, Full chemistry) : MARCH-APRIL 2001 Week 2 Week 3Week 4 « HIDU » « LODU » High dust regimeLow dust regime DUST (< 1µm) vert av. (µg.m-3) Anthropogenic SO4 vert av. (ppb) Validation of GC aerosol fields Heald et al., 2006

6. Gas scavenging by dust SO 2 HNO 3 GC-ref GC-Fe « LODU » vert av. (ppb) SO4DINO3DI Dust mode Results in line with e.g Jordan et al. 2003, Song and Carmichael 2001 Dust mode / Anthro mode aerosol partition 30 %80 %

7. Dust mode pH evolution > OD Lagrangian Box model Dust mode pH Dust (bin 1) 8 4 1 µg.m-3 « LODU » vert av.

8. Soluble iron formation HIDU LODU DUST µg.m -3 FEDI DIF % 3.5 % Large dust event are not necessarily the most FEDI productive (consistency with the 0D scheme, Meskhidze et al., 2005) (ppt) vert av.

9. DIF FEDI FETOT Cor = 0.4 3.5 % 6 ng.m-3 400 ng.m-3 Chen, 2005; remote pacific site 9-26 April 2001 : Soluble Iron highly correlated with dust Comparison with surface concentration observations Consistency of model results with obs. Underestimation of FEDI and DIF Transport issues Missing processes ? photoreduction of FeIII promoted by organic acids

10. Chuang et al., 2005 (ACE-Asia) Comparison with surface concentration obs. : Kosan April 1-30, 2001 Model FEDI vs measurments ? No significant correlation with total iron carried by dust Data analysis ; FEDI shows : Good correlation (R=0.67) with BC particles (anthro. combustion) Cor = -0.1 !

11. FEDI (vert. av.ppt) FEbc (vert. av.ppt) Contribution of combustion soluble iron Simple approach : FEbc = BC x 0.02 Slope obtained by Chuang et al., 2005

12. Wet deposition of soluble iron ng.m -2.d -1 FEbc FEDI 70 « HIDU »« LODU »

13. Importance of chemical buffering effects : low intensity events (more frequent) are more efficient to produce soluble iron compared to big storms. =>Validation and further development of dust/anthro heterogeneous chemistry and aerosol µ-physic in GEOS-CHEM is an important issue for iron modelling. Potential importance of continuous anthropogenic emission of soluble iron (Luo et al., 2007). Experimental characterisation of combustion iron and processing is an issue. Impact of anthropogenic pollution on soluble iron carried by dusts in the East Asian outflow. Longer term simulations, further validations (global) and sensitivity studies are required. How will soluble iron deposition and ecosystem response evolve in the future ? Other mechanisms for dust iron processing and DIF increase (iron III photoreduction / dissolution promoted by organic acids). Mechanisms not fully understood yet. Conclusions

14. Thank you

15. Transpacific transport and simulated DIF Week 2 (high dust)Week 3Week 4 (low dust) DIF ~ 0.15 – 0.6 %DIF ~ 0.6 – 1.1 %DIF ~ 2– 2.7 % FeDI Vertical distribution 4 km 6 km 2 km

16. Importance of iron in High Nutrients Low Chlorophyll regions Annual average Nitrate concentration in surface water (levitus ocean atlas) Regional Iron Fertilisation experiment in different HNLC: e.g / SOIREE expriment Bloom of biological activity and carbon sequestation A 30 to 90  atm drawdown in surface pCO 2 Kohfeld et al., 2005; Archer et al., 2000; Mahowald et al.,1999 … Paleoclimate : ‘The Iron hypothesis’ (Martin) Climate change mitigation (!)