Glacial-interglacial variability of atmospheric CO2 due to changing continental silicate rock weathering: A model study

An 11-box model of the oceanic carbon cycle including sedimentary processes is used to explore the role chemical weathering of continental silicate rocks might play in driving atmospheric CO2 levels on glacial-interglacial timescales. Histories for the consumption of CO2 by silicate rock weathering processes are derived from the marine Ge/Si record. Taking the major uncertainties in the knowledge of the Ge and Si cycles into account, several histories for the evolution of the riverine dissolved silica fluxes are calculated from this record. The investigation of the systematics between riverine dissolved silica and bicarbonate fluxes under different weathering regimes leads us to the tentative conclusion that although there is no correlation between dissolved silica and total bicarbonate concentrations in the major rivers, there may exist a negative correlation between weathering intensity and the ratio of dissolved silica to bicarbonate derived from silicate weathering alone. With this correlation as a working hypothesis, it is possible to interpret the dissolved silica fluxes in terms of equivalent CO2 consumption rates. The calculated histories indicate that glacial rates of CO2 consumption by chemical silicate rock weathering could have been twice, and possibly up to 3.5 times, as high as they are today. When used to force the carbon cycle model, they are responsible for glacial-interglacial pCO2 variations in the atmosphere of typically 50–60 ppm and up to 95–110 ppm. These variations are superimposed to a basic oscillation of 60 ppm generated by the model, mainly in response to coral reef buildup and erosion processes. The total pCO2 signal has an amplitude of about 80–90 ppm and up to 125–135 ppm. Although these large amplitudes indicate that silicate weathering processes should be taken into account when studying glacial-interglacial changes of CO2 in the atmosphere, it also raises new problems, such as too high CO2 levels during the period from 110–70 kyr B.P., requiring further study