Late Pleistocene carbon cycle revisited by considering solid Earth processes

The importance of volcanic CO2 release, continental weathering and coral reef growth on the global carbon cycle has been highlighted by several different studies. Based on these independent approaches we here revisit the last 800 kyr with the box model BICYCLE, which has been extended by a process based sediment module to be able to address these solid Earth contributions to the carbon cycle in detail. We show, that the volcanic outgassing of CO2 as function of sea level change from mid ocean ridges and hot spot island volcanoes cannot be the generic process that leads during phases of falling obliquity to a sea level-CO2 decoupling as has been suggested before. The combined contribution from continental and marine volcanism, if both lagging sea level change by 4 kyr, might have added up to 13 ppm to the glacial/interglacial CO2 rise. The shallow water carbonate sink related to coral reef growth as suggested by an independent model are dur- ing glacial terminations about an order of magnitude too high to be reconciled with meaningful carbon cycle dynamics. Global riverine input of bicarbonate caused by silicate and carbonate weathering is suggested to have been stable over Termination I. However, if weathering fluxes are changed by up to 50% in sensitivity experiments the corresponding bi- carbonate input might contribute less than 20 ppm to the deglacial atmospheric CO2 rise. The overall agreement of re- sults with the new process-based sediment module and the previously applied time-delayed response function to mimic carbonate compensation gives confidence in the results obtained in previous applications of the BICYCLE model without solid Earth processes.