Comparison of water column [CO2aq] with sedimentary alkenone‐based estimates: A test of the alkenone‐CO2proxy

Abstract

The stable carbon isotopic compositions of alkenones have been used to interpret the long‐term history of the partial pressure of atmospheric carbon dioxide (pCO2). Although extensive water column and culture studies document the potential utility and limitations of this approach, to date the accuracy ofpCO2values derived from sedimentary alkenones remains untested. For this study we establish Holocene‐aged, alkenone‐based CO2aqestimates ([CO2aq]alk) from 20 sites along a central Pacific Ocean transect and compare them against both observed modern water column CO2aqand estimated preindustrial concentrations at the depth of alkenone production at each site. Although the [CO2aq]alktrack measured water column values, they are conspicuously lower than modern values across the subtropics. This offset likely reflects the contributions of anthropogenic CO2in modern surface waters relative to preindustrial concentrations at the time of alkenone production. When a model‐based estimate of anthropogenic CO2is removed from the modern observed values, a majority (84%) of [CO2aq]alkfalls within 20% of modeled preindustrial values. Consistency between the modeled and alkenone‐based estimates of preindustrial CO2levels points to the relative accuracy of the alkenone‐CO2method across a wide range of ocean and biogeographic regimes, provided that phosphate concentrations, at the depth of haptophyte production, are reasonably constrained. It further suggests that light‐limited growth and/or active carbon uptake, if they occur, have a negligible effect on reconstructed [CO2aq].