History of carbonate ion concentration over the last 100 million years II: Revised calculations and new data

In an earlier contribution to this journal, we provided a reconstruction of seawater carbonate ion concentration over the last 100 million years (Tyrrell, T. and Zeebe, R.E., Geochim. Cosmoch. Acta, 68, 3521-3530, 2004; TZ04 hereafter). Since then, multiple new and more robust data sets on past ocean carbonate chemistry, atmospheric CO2, and major ion seawater composition have emerged, which prompt new CO2 system reconstructions. In addition, we have gained new insight into the effects of past major ion seawater composition on equilibrium constants affecting CO2 system calculations — most notably due to sulfate. Here we present new reconstructions of past ocean carbonate chemistry and atmospheric CO2 based on new data and revised calculations, including error analysis. We also provide simple corrections for past equilibrium constants, supported by experimental data and well-suited for numerical models and observational studies on multi-million year time scales. Our updated result for just the seawater carbonate ion concentration (∼2.3 to 4-fold lower 100 Myr ago) is similar to TZ04, indicating that our core approach is robust. However, all revised reconstructions using new alkenone and boron data now suggest that long-term ocean inventories of total dissolved inorganic carbon (DIC) and total alkalinity (TA) were similar to modern over the Cenozoic. This result contrasts strongly with one of TZ04’s scenarios, which featured high Paleocene-Eocene DIC/TA inventories and was based on boron-derived pH values that have recently been revised. Because the carbonate system has two degrees of freedom, consistency checks can be made when three or more parameters are determined. Overall, our estimated long-term trends in CO2 system parameters across the Cenozoic appear consistent, regardless of whether we combine our carbonate ion concentration with alkenone-derived pCO2 or boron-derived pH. Our results suggest convergence towards a consistent picture of Cenozoic atmospheric CO2 and seawater chemistry. Finally, we identify changes in past seawater sulfate as a conceptual and practical problem for seawater pH reconstructions