Code and datasets to accompany Evans et al. (2026): The major ion chemistry of seawater was closely coupled to the long-term carbon cycle during the Cenozoic

This upload contains: 1. The MATLAB code (Evans_et_al_NaCa_PNAS.m) and datasets required to perform the data analysis and generate the key figures of Evans et al. (2025) The major ion chemistry of seawater was closely coupled to the long-term carbon cycle during the Cenozoic. The folder 'datasets' contains both the new data reported in the publication as well as datasets utilised in the data analysis or shown for comparison in the figures. The folder 'DeepCarb_output' contains the results of running the carbon cycle box model DeepCarb (https://github.com/dbjevans/DeepCarb/) as described in the publication. The MATLAB script CO2SYSpalaeo.m is a modified version of co2sys (Lewis & Wallace, 1998; Sharp et al., 2020) which accounts for changes in the carbon system dissociation constants as a result of changes in the major ion composition of seawater, following Zeebe & Tyrrell (2019). 2. The key primary datasets reported in the publication: i) Evans_et_al_Cenozoic_Ca_Mg_data.xlsx provides the individual seawater [Ca] and [Mg] reconstructions based on foraminifera Na/Ca (see also the SI Appendix tables) ii) Evans_et_al_composite_Cenozoic_Ca.xlsx provides a Cenozoic composite seawater [Ca] reconstruction with propagated uncertainties and 95% CI (see main text Fig. 1), interpolated at 0.1 Myr intervals.