Osmium, and stable carbon and oxygen isotope record accross the PETM

In the latest Paleocene an abrupt shift to more negative d13C values has been documented at numerous marine and terrestrial sites (Bralower et al., 1997, doi:10.1130/0091-7613(1997)025<0963:HRROTL>2.3.CO;2; Cramer et al., 1999; Kaiho et al., 1996, doi:10.1029/96PA01021; Kennett and Stott, 1991, doi:10.1038/353225a0; Koch et al., 1992, doi:10.1038/358319a0; Stott et al., 1996; Thomas and Shackleton, 1996, doi:10.1144/GSL.SP.1996.101.01.20; Zachos et al., 1993). This carbon isotope event (CIE) is coincident with oxygen isotope data that indicate warming of surface waters at high latitudes of nearly 4°-6°C (Kennett and Stott, 1991, doi:10.1038/353225a0) and more moderate warming in the subtropics (Thomas et al., 1999, doi:10.1029/1999PA900031). Here we report 187Os/188Os isotope records from the North Atlantic and Indian Oceans which demonstrate a >10% increase in the 187Os/188Os ratio of seawater coincident with the late Paleocene CIE. This excursion to higher 187Os/188Os ratios is consistent with a global increase in weathering rates. The inference of increased chemical weathering during this interval of unusual warmth is significant because it provides empirical evidence supporting the operation of a feedback between chemical weathering rates and warm global climate, which acts to stabilize Earth's climate (Walker et al., 1981). Estimates of the duration of late Paleocene CIE (Bains et al., 1999, doi:10.1126/science.285.5428.724; Bralower et al., 1997, doi:10.1130/0091-7613(1997)025<0963:HRROTL>2.3.CO;2; Norris and Röhl, 1999, doi:10.1038/44545; Röhl et al., 2000, doi:10.1130/0091-7613(2000)28<927:NCFTLP>2.0.CO;2) in conjunction with the Os isotope data imply that intensified chemical weathering in response to warm, humid climates can occur on timescales of 104-105 years. This interpretation requires that the late Paleocene thermal maximum Os isotope excursion be produced mainly by increased Os flux to the ocean rather than a transient excursion to higher 187Os/188Os ratios in river runoff. Although we argue that the former is more likely than the latter, we cannot rule out significant changes in the 187Os/188Os ratio of rivers.