Volcanic eruptions liberate large amounts of carbon from the crust and mantle into Earth’s atmosphere. Understanding how this release occurs is a crucial factor in constraining the deep carbon cycle. Until now, many models did not reflect direct observations of CO2 degassing, significantly underestimating how much carbon was released. In a new study, published in the May 2014 of Nature Geoscience , DCO’s Jay Ague and Stefan Nicolescu (Yale University) propose a new model, which could better account for the levels of CO2 observed at arc volcanoes.
Carbon dioxide (CO2) hosted by carbonate minerals in subducted lithosphere can be released during metamorphism or buried to great depths in Earth’s mantle. Most models consider simple devolatilization reactions that produce CO2 and consume or produce H2O. These models predict substantial deep CO2 retention, and that the bulk of the CO2 release occurs in the shallower parts of subduction zones (forearc) before lithosphere reaches the depths of arc magma genesis (subarc). Nonetheless, arc volcanoes emit substantial CO2, strongly suggesting that other CO2 release mechanisms also operate.
This study investigates carbonate mineral dissolution and its role in releasing CO2 from subducted lithosphere. The researchers studied chemical, mineralogical, and oxygen and carbon isotopic changes in metamorphosed carbonate rocks adjacent to syn-subduction veins or fold hinge fluid conduits on Tinos and Syros islands (Greece), part of the exhumed Cycladic subduction complex. As the distance to the conduits decreases, calcite (after aragonite) drastically decreases in abundance, whereas silicate minerals (e.g., omphacite, glaucophane, epidote, chlorite) increase. The reactions involve open-system addition of Al, Si, Na, Fe, and Mg, and loss of Ca, Sr, Ba, Rb and volatiles. The d18O of carbonate minerals is 5 to 10 ‰ lighter than expected for typical metacarbonate rocks, demonstrating infiltration of fluids from surrounding metasediments and metavolcanics (Tinos) or ultramafic mélange (Syros).
The authors propose a model in which dissolution of CaCO3 resulting from fluid infiltration at high pressures added CO2 to the fluids, lowering water activity and, thus, the solubilities of silicate minerals. This approach would link carbonate dissolution and silicate precipitation and account for the observed chemical and mineralogical alteration. The coupled dissolution and precipitation release as much as 60–90 % of the rock’s CO2, which greatly exceeds the amounts predicted by simple devolatilization models. The observed dissolution is consistent with a growing body of experimental and fluid inclusion evidence that requires that carbonate minerals have substantial solubilities at high or ultrahigh pressures in subduction zones. Carbonate mineral dissolution should operate in subarc regions as well as forearcs, and could thus play a crucial role in balancing the global carbon cycle.
Craig Manning, DCO’s Extreme Physics and Chemistry Community Chair, contributed a News and Views article regarding this paper to the same edition of Nature Geoscience , which highlights the importance of this work to understanding Earth’s deep carbon cycle.
The Reservoirs and Fluxes Community of the Deep Carbon Observatory is thanked for their support for a workshop on Tectonic Fluxes of Carbon. Funding provided by the US National Science Foundation Directorate of Geosciences (EAR-0105927, EAR-0744154) is gratefully acknowledged.
Image credit: Jay J. Ague
Cover Caption: The balance between carbonate subduction into the deep Earth and CO2 release through degassing at volcanoes is critical for the carbon cycle. Geochemical analyses of an exhumed subduction zone complex in Greece show that fluidmediated reactions could liberate significant amounts of carbon from the subducting slab for later release at arc volcanoes. The image shows crystals of epidote, several millimetres in length, in crosspolarized light. The crystals are from a quartz vein on Tinos island, Greece, that facilitated fluid infiltration and carbonate mineral dissolution.