Iron and sulfur cycling in the cGENIE.muffin Earth system model (v0.9.21)

Abstract

The coupled biogeochemical cycles of iron and sulfur are central to the long-term biogeochemical evolution of Earth's oceans. For instance, before the development of a persistently oxygenated deep ocean, the ocean interior likely alternated between states buffered by reduced sulfur (“euxinic”) and buffered by reduced iron (“ferruginous”), with important implications for the cycles and hence bioavailability of dissolved iron (and phosphate). Even after atmospheric oxygen concentrations rose to modern-like values, the ocean episodically continued to develop regions of euxinic or ferruginous conditions, such as those associated with past key intervals of organic carbon deposition (e.g. during the Cretaceous) and extinction events (e.g. at the Permian–Triassic boundary). A better understanding of the cycling of iron and sulfur in an anoxic ocean, how geochemical patterns in the ocean relate to the available spatially heterogeneous geological observations, and quantification of the feedback strengths between nutrient cycling, biological productivity, and ocean redox requires a spatially resolved representation of ocean circulation together with an extended set of (bio)geochemical reactions. Here, we extend the “muffin” release of the intermediate-complexity Earth system model cGENIE to now include an anoxic iron and sulfur cycle (expanding the existing oxic iron and sulfur cycles), enabling the model to simulate ferruginous and euxinic redox states as well as the precipitation of reduced iron and sulfur minerals (pyrite, siderite, greenalite) and attendant iron and sulfur isotope signatures, which we describe in full. Because tests against present-day (oxic) ocean iron cycling exercises only a small part of the new code, we use an idealized ocean configuration to explore model sensitivity across a selection of key parameters. We also present the spatial patterns of concentrations and δ56Fe and δ34S isotope signatures of both dissolved and solid-phase Fe and S species in an anoxic ocean as an example application. Our sensitivity analyses show that the first-order results of the model are relatively robust against the choice of kinetic parameter values within the Fe–S system and that simulated concentrations and reaction rates are comparable to those observed in process analogues for ancient oceans (i.e. anoxic lakes). Future model developments will address sedimentary recycling and benthic iron fluxes back to the water column, together with the coupling of nutrient (in particular phosphate) cycling to the iron cycle.