Sea-level driven Pleistocene decline in atmospheric oxygen concentrations after the MPT

Ice core records show a sigificant decline in atmospheric oxygen concentrations during the last 800ka (Stolper et al., 2016). Recent data from blue ice seem to show similar oxygen concentrations from around 800 ka to 1500 ka (Yan et al 2021) and this change has been suggested as one possible driver of the Mid Pleistocene Transition (MPT) from 40ka (one obliquity cycle) to 80 or 120 ka glacial cycles (two or three obliquity cycles). We have shown that not only shallow but also deep terrestrial pyrite oxidation and potentially oxidation of organic substance on shelves exposed during sea-level lowstands in glacials is a significant source of both CO₂ and a significant sink for O₂ (Kölling et al 2019). Since these reduced phases are only recharged in the upper meters of the inundated shelf sediments during interglacials, the reservoir of the CO₂ producing and oxygen consuming material is gradually declining over the Pleistocene. In our model, the MPT is the time, when the sea-level decline during one 40ka obliquity cycle is not enough to expose significant amounts of reduced shelf sediments anymore. We suggest that the CO₂ release through subareal oxidation of pyrite and organic substance on exposed shelves might be a substantial ingredient adding to the obliquity driven increase in high latitude insolation to terminate a glacial phase. The lack of this additional CO₂ contribution might thus be a cause of glacial cycles failing to terminate whithout any significant change in orbital conditions. In our model, the oxygen demand by pyrite oxidation causes a step decline to 21.16 vol% in the beginning of the Pleistocene at 2.6 Ma, a gradual decline to 21.11 Vol% at 1.2 Ma (MIS 36) and a steeper decline after the MPT. The major step declines in oxygen concentrations driven by pyrite oxidation are modelled in MIS 20 (-0.02 vol%), MIS16 (-0.04 Vol%) and MIS 12 (-0.03 Vol%). In our model, the Pleistocene decline in atmospheric oxygen is a side effect of the terrestrial oxidation of pyrite and organic substance in glacially exposed shelf sediments. We think, there is no "100 ka" driver for longer glacial cycles, but the magnitude and timing of this CO₂ producing and O₂ consuming process might be crucial to terminate glacials and might thus be responsible for failed terminations /missing interglacials specifically after the MPT.