A Data-constrained Estimate of the Global Ocean Iron Cycle: Budgets, Timescales, and Iron Limitation

Abstract:

The oceanic iron cycle is estimated by optimizing a simple steady-state model based on a data-assimilated global circulation, with a prescribed optimized phosphorus cycle and a prescribed aeolian source pattern. Key biogeochemical parameters are determined by minimizing a suitably weighted quadratic misfit between the model’s dissolved iron concentration and a global data set of sparse measurements. The global dissolved iron inventory is estimated to be (7.1±0.1)×10¹¹ mol Fe, of which (6.9±0.1)×10¹¹ mol Fe is bound to organic ligands and hence bioavailable, while the remainder is “free” iron. The aeolian iron input rate is estimated at (3.3±0.5)×10⁹ mol Fe/year, corresponding to a bulk residence time for bioavailable iron of 215±40 years, comparable to the bulk biological cycling timescale estimated at 246±24 years. Iron limitation is quantified in terms of the difference [Fe^∗] between the actual iron concentration and that needed to utilize the available phosphate. The optimized model captures the observed high-nutrient, low-chlorophyll regions of the ocean as iron-limited regions with [Fe^∗]<0. We define an iron age, Γ_Fe, as the mean time since iron at a given point was last injected from the atmosphere and compute Γ_Fe using an equivalent linear formulation of the model. In the euphotic zone, Γ_Fe ranges from a few decades or less in regions of high aeolian input to ∼1800 years in the Southern Ocean. The patterns of Γ_Fe show that iron is supplied to the Southern Ocean euphotic zone primarily from depth rather than being advected within the thermocline following deposition from continental dust plumes. Because [Fe^∗] is negative in the deep southern oceans, upwelling waters maintain Southern Ocean iron limitation.