Planktonic foraminiferal Mg/Ca ratios from top of different sediment cores from the Atlantic

The accuracy of the Mg/Ca paleothermometer is contested over the influences of secondary environmental parameters such as salinity and pH. Recent calibrations based on compiled sediment trap and laboratory culture data suggest moderate influences from salinity and pH. Core-tops are the best analogues to downcore samples used for paleoceanographic reconstruction with well-constrained environmental parameters and thus can be used to validate sediment trap and laboratory culture calibrations. Here, we calibrate new core-top Mg/Ca data in Globigerinoides* ruber* (white) and *Trilobatus sacculifer *(without final sac-like chamber) with sea surface temperature, salinity, and pH. Part of these core-tops were previously used to argue for a large salinity effect on G. ruber (w) Mg/Ca [Arbuszewski et al., 2010]. Our new G. ruber (w) Mg/Ca data are on average 12% lower than the previous results. Our calibrations yield Mg/Ca-temperature sensitivities of 8.1 ± 0.7 %/°C for G. **ruber (w) and 6.6 ± 0.8 %/°C for *T. sacculifer *(w/o sac), and Mg/Ca-salinity effects of 4.7 ± 2.4 %/‰ for G. **ruber (w) and 5.5 ± 2.3 %/‰ for *T. sacculifer *(w/o sac). These results agree well with culture experiments but discount the large salinity effects reported in previous core-top studies. Our data reveal insignificant pH effects on Mg/Ca in both species. Overall, our core-top calibrations for G. **ruber (w) and *T. sacculifer *(w/o sac) lend strong support to previous calibrations, strengthening our confidence in the use of planktonic Mg/Ca as a reliable proxy for sea surface temperature reconstructions.