ATMOSPHERE-OCEAN CO2 DISEQUILIBRIUM AND THE HYDROLOGIC RESPONSE TO CLIMATE CHANGE DURING THE PETM: AN ARCTIC PERSPECTIVE

Recently recovered IODP cores from the central Arctic Ocean provide the first opportunity to evaluate the polar response to global warming during the PETM. Here we present δ¹³C and δD compositions of n-alkanes recovered from these sedimentary cores. The δ¹³C values from higher plant n-alkanes record a PETM carbon isotope excursion (CIE) that is ~3‰ more negative than the ~-2.5‰ CIE recorded in benthic foraminifera, but similar to that recorded in soil carbonates from mid-latitudes. Given that humidity and soil moisture are close to saturation in the Arctic, we propose that the CIE recorded by higher plant n-alkanes accurately reflects the CIE of atmospheric CO₂ during the PETM. If valid, the δ¹³C of atmosphere CO₂ was consistently offset by 2 to 3‰ from that of dissolved inorganic carbon in the deep ocean over the length of the PETM. Maintenance of this atmosphere-ocean disequilibrium demands that the atmosphere directly received an influx of ¹³C-depleted carbon during PETM at a flux that could compete with the rapid exchange with the surface ocean. The δD values of n-alkanes from terrestrial plants average ~-215‰ before the PETM, and shift to -160‰ during the event. Modern studies suggest a hydrogen isotope fractionation of ~-130 to -150‰ between n-C₂₉ and source water. Accordingly, we estimate that spring precipitation before and after the PETM had δD values of ~-70 to -95‰, and increased to -30 to -55‰ during the onset of the PETM. Modern precipitation in this isotopic range occurs primarily in subtropical regions. Thus, our δD data indicate a substantial reduction in the degree of D depletion of high-latitude precipitation during the latest Paleocene-early Eocene relative to the Present. This effect is enhanced during the early portion of the PETM. D-enriched precipitation could result from changes in evaporative sources, including the Arctic Ocean. Additionally, it could result from a decrease in either the meridional or vertical temperature gradients, both of which would act to reduce rain-out of subtropical water vapor and decrease the isotopic fractionation between vapor and condensate. We prefer the latter scenario of increased water vapor transport because this mechanism is supported by the prevalence of low salinity surface waters and high seasonal run-off in the Arctic during the PETM.