The middle Pleistocene transition: characteristics, mechanisms, and implications for long-term changes in atmospheric pCO 2

Abstract

The emergence of low-frequency, high-amplitude, quasi-periodic (∼100-kyr) glacial variability during the middle Pleistocene in the absence of any significant change in orbital forcing indicates a fundamental change internal to the climate system. This middle Pleistocene transition (MPT) began 1250 ka and was complete by 700 ka. Its onset was accompanied by decreases in sea surface temperatures (SSTs) in the North Atlantic and tropical-ocean upwelling regions and by an increase in African and Asian aridity and monsoonal intensity. During the MPT, long-term average ice volume gradually increased by ∼50 m sea-level equivalent, whereas low-frequency ice-volume variability experienced a 100-kyr lull centered on 1000 ka followed by its reappearance ∼900 ka, although as a broad band of power rather than a narrow, persistent 100-kyr cycle. Additional changes at 900 ka indicate this to be an important time during the MPT, beginning with an 80-kyr event of extreme SST cooling followed by the partial recovery and subsequent stabilization of long-term North Atlantic and tropical ocean SSTs, increasing Southern Ocean SST variability primarily associated with warmer interglacials, the loss of permanent subpolar sea-ice cover, and the emergence of low-frequency variability in Pacific SSTs and global deep-ocean circulation. Since 900 ka, ice sheets have been the only component of the climate system to exhibit consistent low-frequency variability. With the exception of a near-universal organization of low-frequency power associated with marine isotope stages 11 and 12, all other components show an inconsistent distribution of power in frequency-time space, suggesting a highly nonlinear system response to orbital and ice-sheet forcing. Most hypotheses for the origin of the MPT invoke a response to a long-term cooling, possibly induced by decreasing atmospheric pCO 2. None of these hypotheses, however, accounts for the geological constraint that the earliest Northern Hemisphere ice sheets covered a similar or larger area than those that followed the MPT. Given that the MPT was associated with an increase in ice volume, this constraint requires that post-MPT ice sheets were substantially thicker than pre-MPT ice sheets, indicating a change in subglacial conditions that influence ice dynamics. We review evidence in support of the hypothesis that such an increase in ice thickness occurred as crystalline Precambrian Shield bedrock became exposed by glacial erosion of a thick mantle of regolith. This exposure of a high-friction substrate caused thicker ice sheets, with an attendant change in their response to the orbital forcing. Marine carbon isotope data indicate a rapid transfer of organic carbon to inorganic carbon in the ocean system during the MPT. If this carbon came from terrigenous sources, an increase in atmospheric pCO 2 would be likely, which is inconsistent with evidence for widespread cooling, Apparently rapid carbon transfer from terrestrial sources is difficult to reconcile with gradual erosion of regolith. A more likely source of organic carbon and nutrients (which would mitigate pCO 2 rise) is from shelf and upper slope marine sediments, which were fully exposed for the first time in millions of years in response to thickening ice sheets and falling sealevels during the MPT. Modeling indicates that regolith erosion and resulting exposure of crystalline bedrock would cause an increase in long-term silicate weathering rates, in good agreement with marine Sr and Os isotopic records. We use a carbon cycle model to show that a post-MPT increase in silicate weathering rates would lower atmospheric pCO 2 by 7–12 ppm, suggesting that the attendant cooling may have been an important feedback in causing the MPT.