Coupled carbon and silica cycle perturbations during the Marinoan snowball Earth deglaciation

Abstract

The snowball Earth hypothesis proposes that if polar ice sheets were to advance equatorward of a mid-latitude threshold, runaway ice-albedo effects would lead to a stable, globally ice-covered climate state that would require extremely high atmospheric pCO2 levels (supplied by volcanic degassing over millions of years) for deglaciation. Geologic evidence, including globally distributed and low-latitude glacial deposits, suggests that two such global glaciations occurred during the Neoproterozoic. We model the coupled carbon and silica cycles through a snowball Earth event, including the extremely high pCO2 and dramatically accelerated chemical weathering of its aftermath. The enhanced delivery of dissolved weathering products to the ocean induces elevated sedimentary burial of CaCO3 (deposited as “cap carbonates”) and SiO2. Uncertainty in the relative importance of carbonate versus silicate weathering allows a wide range of possible CaCO3 burial magnitude, potentially dwarfing that of SiO2. However, total SiO2 burial is insensitive to weathering strengths, and is set by the amount of CO2 required for deglaciation (~1019 mol). Chert associated with Marinoan post-glacial cap carbonates in Africa and Mongolia corroborate modeled predictions of elevated SiO2 burial. INTRODUCTION Synchronous, globally distributed, low-latitude glacial deposits (Evans and Raub, 2011; Kirschvink, 1992) suggest that near-global glaciation occurred at least twice (the Marinoan and Sturtian glaciations) in the Neoproterozoic. The snowball Earth (SE) hypothesis (Hoffman et al., 1998; Kirschvink, 1992) attributes such glaciations to the equatorward advance of polar ice sheets across a mid-latitudinal threshold, beyond which runaway ice-albedo effects led to a stable, globally glaciated Earth (see Hoffman et al. [2017] for a summary of SE events). The high-albedo, low-temperature SE climate state would require extremely high atmospheric pCO2 (Caldeira and Kasting, 1992; Le Hir et al., 2008b), supplied by volcanic degassing in the absence of silicate weathering (Si-weathering) for millions of years, for deglaciation (Walker et al., 1981). The high pCO2 required to overcome a SE state (~0.12 bar [Caldeira and Kasting, 1992] or higher [Pierrehumbert, 2004]) would significantly accelerate the rate of continental chemical weathering (Berner et al., 1983; Walker et al., 1981) following deglaciation. The Marinoan and Sturtian cap sequences are distinct, but both likely formed in stratified water masses during intervals of enhanced chemical weathering in the high-pCO2 SE aftermath (Higgins and Schrag, 2003; Rooney et al., 2014; Yang et al., 2017). Ca and Mg isotopes (Huang et al., 2016; Kasemann et al., 2014) support elevated chemical weathering rates of both silicates and carbonates due to high pCO2 (Bao et al., 2008) following the Marinoan, while Os and Sr isotopes support elevated post-Sturtian weathering (Rooney et al., 2014). Elevated weathering would enhance the delivery of weathering products (dissolved silicate [DSi], dissolved inorganic carbon [DIC], and cations, which contribute alkalinity) to the oceans via the generalized reactions (Urey, 1952): CaCO3 + CO2 + H2O → 2HCO3 + Ca2+ ; (1) CaSiO3 + 2CO2 + 3H2O → 2HCO3 + H4SiO4 + Ca2+. (2) The addition of alkalinity and DIC to seawater in the above 1:1 ratio (the products of Reactions 1 and 2) elevates the saturation state of carbonate (Ω), favoring its precipitation and burial in sediments: 2HCO3 + Ca2+ → CaCO3 + CO2 + H2O . (3) Widespread evidence for significantly elevated post-SE carbonate burial comes in the form of “cap carbonates,” globally ubiquitous 10-m-scale layers of calcite and dolomite overlying glaciogenic deposits and glaciated surfaces (Hoffman et al., 2007; Hoffman and Schrag, 2002) with sedimentological features indicating rapid, inorganic precipitation (James et al., 2001). The elevated DSi flux from (from Equation 2) should also promote the precipitation and burial of SiO2: H4SiO4 → SiO2 + 2H2O , (4) which in the Precambrian was accomplished by inorganic precipitation of cristobalite, tridymite, or microcrystalline quartz and preserved in sediments as their diagenetic product, chert (Maliva et al., 1989; Siever, 1992). While the link between elevated post-SE weathering and cap carbonates is established (Higgins and Schrag, 2003; Hoffman and Schrag, 2002), the effects on the ocean’s Si cycle and any connection to SiO2 burial has received little attention. MODELING POST-SNOWBALL WEATHERING To explore the geochemical implications of elevated chemical weathering during SE deglaciation, we perform a suite of simulations using a model of the coupled Precambrian C and Si cycles (PreCOSCIOUS, the PreCambrian Ocean Silica Carbon Inorganic Ocean Underwater Sedi ment model, described in the GSA Data Repository1 and in Figure 1), constructed for this purpose by combining elements of previous models. PreCOSCIOUS uses the three-box architecture (atmosphere, shallow ocean, and deep ocean) of previous SE geochemical models (Higgins and Schrag, 2003; Le Hir et al., 2008b). Carbonate chemistry and weathering feedbacks are adapted from the carbon cycle model LOSCAR (Zeebe, 2012), and a Si cycle was modified from Penman (2016) to include the 1GSA Data Repository item 2019107, detailed description of the PreCOSCIOUS model, snowball Earth sensitivity experiments, and MATLAB code for reproducing the runs described in the text, is available online at http:// www .geosociety .org /datarepository /2019/, or on request from editing@ geosociety .org. CITATION: Penman, D.E., and Rooney, A.D., 2019, Coupled carbon and silica cycle perturbations during the Marinoan snowball Earth deglaciation: Geology, v. 47, p. 1–4, https:// doi .org /10 .1130 /G45812.1 Manuscript received 13 June 2018 Revised manuscript received 7 January 2019 Manuscript accepted 11 January 2019 https://doi.org/10.1130/G45812.1 © 2019 Geological Society of America. For permission to copy, contact [email protected]. Published online XX Month 2019 Downloaded from https://pubs.geoscienceworld.org/gsa/geology/article-pdf/doi/10.1130/G45812.1/4653545/g45812.pdf by Yale University, Donald Penman on 02 March 2019 2 www.gsapubs.org | Volume 47 | Number 4 | GEOLOGY | Geological Society of America Precambrian abiotic SiO2 precipitation (Maliva et al., 1989; Siever, 1992). Technical details (equations, constants, references, and MATLAB code) are included in the Data Repository. For each time step (~0.01 yr), input and output fluxes of four tracers in each reservoir are calculated: carbon (atmospheric CO2 and seawater DIC), alkalinity, Ca2+, and DSi. The atmospheric CO2 budget includes volcanogenic input, removal by carbonate and Si-weathering fluxes, (Equations 1 and 2, the rates of which are functions of pCO2 after Walker and Kasting [1992] and scaled to modern fluxes), and air-sea gas exchange with the surface ocean based on the degree of pCO2 disequilibrium (after Zeebe, 2012). pCO2 determines temperatures (using a simple climate sensitivity with a decay function, after Zeebe [2012]), which affect the equilibrium constants used to calculate full seawater carbonate chemistry from [DIC] and total alkalinity (TA). The dissolved products of chemical weathering (DIC, DSi, and TA) are delivered to the surface ocean, where they fuel precipitation and burial of CaCO3 (at a rate governed by Ω) and SiO2 (governed by [DSi]) from the surface ocean, consistent with Precambrian modes of SiO2 (Maliva et al., 1989; Siever, 1992) and CaCO3 (Zeebe and Westbroek, 2003) burial. Both CaCO3 and SiO2 precipitation occur only above a critical threshold of oversaturation, resulting in a “Strangelove ocean” (Zeebe and Westbroek, 2003) with very high Ω and [DSi]. The surface and deep oceans are linked by vertical mixing applied to all tracers, and by a DIC flux from the surface to deep representing the formation, sinking, and remineralization of organic carbon. Before experiments, the model was run for millions of model years until it reached a stable equilibrium, characterized by constant concentrations of all tracers in all reservoirs, carbon input by volcanic degassing perfectly balanced by Si-weathering and CaCO3 burial. A simple SE forcing halts carbonate and Si-weathering and organic carbon export and sets ocean temperatures to 0 °C. A complete SE experiment is shown in Figure 2, using a favored set of constants and boundary conditions (default run). Sensitivity tests to explore different configurations are detailed in the Data Repository. When the SE forcing is applied, the supply of weathering-derived alkalinity to fuel carbonate burial is shut off, and subsequently DIC removal quickly ceases and CO2 begins to accumulate in the atmosphere at ~18,000 ppm/m.y. Dissolution of CaCO3 (delivered by physical erosion and transport by glacial action; Hoffman and Schrag, 2002) also occurs whenever Ω < 1, which maintains carbonate saturation and imparts an increase in seawater [Ca2+] during glaciation. The importance of [Ca2+] rise for Ω and sensitivity tests that omit this feature are explored in the Data Repository. DSi inputs are significantly reduced without Si-weathering, but hydrothermal DSi input continues and the Si-cycle reaches a stable state with lower [DSi] and SiO2 burial rate. The default run takes ~6.7 m.y. for pCO2 to reach the 0.12 bar deglaciation threshold (used in the default run for consistency with previous SE modeling, e.g., Higgins and Schrag, 2003; higher thresholds are explored in the Data Repository). Using different configurations (detailed in the Data Repository), the SE duration ranges from 5 to 16 m.y., in agreement with the 5–14 m.y. duration of the Marinoan glaciation suggested by geochronology (Hoffman et al., 2017). Upon deglaciation, the surface rapidly warms due to high pCO2 from 0 to ~35 °C and weathering rates immediately accelerate (to ~11× equilibrium rates in the default run; differing weathering sensitivities are compared