Towards Reconciling Iron Supply and Demand in the Southern Ocean

Alessandro Tagliabue1,2

J-B Sallée3, P.W. Boyd4, A.R. Bowie5, M. Lévy6, S. Swart2

1University of Liverpool, UK2CSIR, South Africa

3British Antarctic Survey, UK4University of Otago, New Zealand5University of Tasmania, Australia

6LOCEAN-IPSL, France

a.tagliabue@liverpool.ac.uk

Outline

• Importance of physical processes• The Ferricline• Methods• Results– Ferricline distributions and relation to MLD– Estimating Fe inputs

• Generalised View of Seasonal Fe Cycle• Summary and Conclusions

Importance of physical processes• Southern Ocean productivity is

Fe limited• Variability in production should

be connected to changing degrees of Fe limitation

• Much attention on external supplies

• Large reservoirs below the mixed layer

• Physical processes crucial in mediating transfer of Fe to the biota

Boyd and Ellwood, 2010

Tagliabue et al., 2010

Importance of physical processes

• Two main physical mechanisms:

– Winter Entrainment– Diapycnal Diffusion

• Fe stock down to MLDMAX

• Some ‘detrained’ during shallowing

• dFe/dz at MLD• Kz (±10-5-10-4, m2 s-1)

Importance of physical processes

• Two main physical mechanisms:

– Winter Entrainment– Diapycnal Diffusion

• Sensitive to different processes

• Fe stock down to MLDMAX

• Some ‘detrained’ during shallowing

• dFe/dz at MLD• Kz (±10-5-10-4, m2 s-1)

Buoyancy vs momentum

• Relative Roles unknown, implies that we don’t well know the climate sensitivity of Fe vertical supply

The Ferricline

• Key control on the vertical input of dFe

• Similar to the ‘nitracline’• dFe has– Longer remineralisation

length scale– Particle Scavenging– Variable biological

demand• Relation to MLD at basin

scale unknownKlunder et al. (2011)

ZFe

MLD

Methods

• 3 complementary datasets:– New compilation of dFe measurements– ARGO co-location– Satellite estimates of iron utilisation

Results

Ferricline Depths

328±198m

Ferricline Depths

-Strong latitudinal trend-Modification to isopycnals drives much variability in ZFe

s0 at ferricline

Relation to MLDZFe – MLD (m)

236±200mZFe <MLD in 11 (8%)

Or 4-19 cases at ±2s

Vertical Gradients

• The “ferricline” is the largest Fe source• Gradients sharper in the Atlantic Basin• Some regions do show some vertical gradient at the MLD

Diapycnal Diffusionmmol m-2 yr-1

Across ±2s and Kzestimates:2-10 nmol m-2 d-1

OR0.6-7.7 mmol m-2 yr-1

Winter Entrainment

• MLDs deepen in winter– ARGO provides us this information

• But ZFe determinations generally from summer

• Assume conservation of density• Use Fe, s measured to ‘project’ Fe onto s profile at

time of MLDMAX

Relation to MLDMAXZFeW – MLDMAX (m)

~210mZFe <MLD in 22 (17%)

Or 9-40 cases at ±2s

Entrainmentmmol m-2 yr-1

Across ±2s:9.1-30.2 mmol m-2 yr-1

(0.6-7.7 mmol m-2 yr-1) “detrainment” dFe stock during shallowing accounted for from ARGO

Supply versus Demand?

• Iron utilisation includes recycled Fe– ‘fe ratios’ can be as

low as 0.1• Range of different Kz

estimates• Sensitivity to

Detrainment of winter dFe stock Boyd et al. 2012

Supply vs Demand?

Only

- when kz is very high- fe-ratio very low

can diapycnal supply meet demand at >50% of locations

Supply vs Demand?

Only

- when detrainment losses are very high- fe-ratio very high

Does entrainment NOT meet demand at >50% of locations

Seasonal Cycling

Summary and Conclusions

• The ferricline is robustly decoupled from the MLD by ~200m

• The ferricline depth has a strong relation to density• Only entrainment during winter is able to supply

appreciable amounts of Fe over much of the S.O.• Low diapycnal inputs during summer result in a large

reliance on recycled Fe in many locations• PP likely sensitive to processes that modulate winter

mixing rather than summer stratification

a.tagliabue@liverpool.ac.uk

Density and the Ferricline