Variations in Mid-Ocean Ridge CO2 Emissions Driven By Glacial Cycles

Friday, 19 December 2014
Jonathan M Burley, University of Oxford, Oxford, 0X1, United Kingdom, Richard F Katz, University of Oxford, Oxford, United Kingdom and Peter J Huybers, Harvard University, Cambridge, MA, United States
Glacial cycles impact continental volcanism through pressure changes associated with growth and retreat of ice sheets [e.g. Iceland - Jull, 1996]. Similarly, changes in sea level accompanying glacial cycles modulate mid-ocean ridge (MOR) volcanism by pressure changes and their influence on melt production [Crowley 2014; Lund 2011; Huybers 2009]. CO2 transport through the upper mantle is sensitive to mantle melting because CO2 partitions completely into the melt phase when present. Melt then transports CO2 to the ridge axis, where it enters the climate system. We present models of CO2 transport that investigate how sea level modulates the rate of CO2 emission from MORs. The total carbon reservoir in the mantle is circa 10^7 GtC [Dasgupta 2010], orders of magnitude more than the oceans (40,000 GtC) and atmosphere (600 GtC). Changes in the rate of CO2 emission from the solid Earth therefore have the potential to significantly affect the surface carbon system.

We have developed an analytical model of CO2 transport from the depth of first silicate melting (~60km) to the ridge axis, enabling a calculation of CO2 emission rate for a generic section of MOR. The model assumes homogeneous mantle and energy-conserving melt production from a simplified 2-component mantle; CO2 is taken as a perfectly incompatible trace element. Pressure variations modulate the depth of initial silicate melting and hence the flux of CO2 into the melting regime. The model can also be applied to any species that is strongly partitioned into the melt (eg. Uranium, Thorium, Niobium, Barium, Rubidium).

Results suggest that changing sea level over the past Myr could have altered the CO2 emissions from MOR by ~8%. The magnitude of variation in emissions is sensitive to the mantle permeability, the ridge spreading rate, and the rate of change of sea level. The travel time of melt through the mantle causes a delay between sea-level change and the CO2 response of the MOR. This delay is sensitive to plate spreading rate and mantle permeability.

Delayed CO2 response to ice-age-driven sea-level creates the possibility of climactic feedback. A dynamical-systems climate model indicates that for sufficiently large and delayed variations in emissions, such a feedback could pace the glacial/interglacial oscillations.