| 18 Sep 2026
MioMG: Reconstructing robust ocean temperatures from Miocene foraminiferal Mg/Ca records
Abstract. Paleoclimate records from the Miocene epoch (~5.3 to ~23.0 Million years ago) present an opportunity to assess the Earth’s climate system under a warmer-than-modern mean state, when atmospheric CO2 levels were within the range of those projected for the next century under intermediate climate change scenarios. Ocean temperature is a key climate system parameter because latitudinal temperature gradients provide critical information on the climate system and potential carbon cycle feedbacks in the geological past. Over the last several decades, magnesium-to-calcium (Mg/Ca) ratios in both planktic and benthic foraminiferal shells have been applied as a temperature proxy providing insight into spatial and temporal change in upper ocean and bottom water temperature across the Miocene. However, Miocene climate models struggle to reconstruct the reduced latitudinal temperature gradient inferred from available proxy records, including those based on Mg/Ca. Further, recent advances in planktic and benthic foraminiferal Mg/Ca-temperature paleothermometry show that non-thermal factors and seawater chemistry need to be considered when converting Mg/Ca to temperature. Here, we present the MioMG temperature compilation, led by the Mg/Ca working group within the international MioOcean community effort. Fifty-two Miocene foraminiferal Mg/Ca records from both planktic and benthic foraminifera were reevaluated and temperatures were recalculated to provide robust, consistent temperature estimates including uncertainties. Conversion of planktic and benthic Mg/Ca to temperature requires consideration of carbonate chemistry sensitivity for individual species, sensitivity of foraminiferal Mg/Ca to the Mg/Ca seawater composition (Mg/Casw), and appropriate calibration. We compiled records of the seawater Mg/Ca composition (Mg/Casw), based on a diverse set of proxies, and present this record for use in future application of Mg/Ca-derived Miocene temperature reconstructions. Miocene Mg/Ca records are derived from extant and extinct foraminiferal species. Further, preservation information is essential to determine the quality of reconstructions. Our recalculated upper ocean temperature (UOT) estimates showcase warm Middle Miocene temperatures and a cooling trend towards the Late Miocene. Revised bottom water temperature (BWT) estimates also depict warm Middle Miocene values followed by a cooling trend, which is consistent among sites located at different paleodepths. Overall, absolute temperature estimates are higher than previously published estimates due to application of corrections based on the updated MioOcean Mg/Casw curve. The MioMG compilation is presented on an interactive Streamlit-based web platform allowing for exploration of the data for specialists and non-specialists to aid in application of dataset for Miocene researchers. The dataset is publicly available at Zenodo (https://doi.org/10.5281/zenodo.22258386, Sosdian et al. 2026).
This work provides a compilation of planktonic and benthic foraminiferal Mg/Ca values from the Miocene, with new calibrations that consider non-thermal variables. One of the most significant sources of uncertainty in this thermometer is the variation in seawater Mg/Ca (Mg/Casw) at geological timescales. This work proposes compiled Mg/Casw data derived from various archives. The new calibration equations enable uncertainties to be estimated. The work provides a numerical tool for calculating upper ocean temperature (UOT) and bottom water temperature (BWT), and makes recommendations for obtaining consistent data in the future.
This is the result of significant collaborative efforts to obtain reliable and consistent seawater temperature data across the Miocene — a pivotal period for studying climate mechanisms in a warm mean state. The availability of temperature data based on clumped isotope thermometers that are free from changes in Mg/Casw, as well as the compilation of foraminiferal Nd/Ca, makes this work timely. I highly appreciate the well-coordinated collaborative efforts of the Mg/Ca working group as part of the international MioOcean community. This work will be helpful not only to researchers focusing on the Miocene, but also to those focusing on other periods, as it provides a way to work in a consistent manner. I would like to see this work published in Earth System Science Data.
The manuscript is well written, with clear figures and tables. There are some issues with the numbering of the figures and tables, which I will demonstrate below. I have a limited number of minor and specific comments.
Line 189: Remove “trace metal”. Indeed, Mg in foraminiferal tests is not a trace element. It is a minor element.
Lines 203–205: Where can the summary of the preservation state of the foraminiferal tests be found? In Table S6?
Line 242: “Table A1–A2” should be replaced by “Table S1–S2”.
Lines 266–267: The definition of “key sites” is unclear. Please add a short explanation.
Lines 204 and 526: “Table A1” should be replaced by “Table S1”.
Lines 317 and 387: A foraminiferal species should be spelled out at the beginning of a sentence.
Lines 361 and 621: “Table A2” should be replaced with “Table S2”.
Line 413: Add a closing bracket after “ODP U1337”.
Line 543: “Where” should start with a lowercase letter.
Line 602: “0.09 from Anand et al. (2003)”. This exponential constant is for planktonic foraminifera. Please provide a brief justification for selecting this value for benthic foraminifera.
Line 649: “the offset might be species and site specific”. While this may be beyond the scope of this study, is it possible to extract the trend for species and regions from the new compilation?
Line 711: “NaCa” should be “Na/Ca”.
Lines 712, 714, 724 and 746: “Fig. C1” should be “Fig. B1”.
Line 789: “utilising additional proxies (e.g. Sr/Ca ratios etc.)”. This is a very interesting idea. Did the authors identify any trends between Mg/Ca and Sr/Ca in the compiled data?
Line 817: “Fig. B1”. I could not find the corresponding figure.
Lines 826 and 827: “Fig. B1” should be replaced by “Fig. A1”.
Lines 829–830: “light microscopy” should be replaced by “LM”.
Line 846: Before “MMCT”, insert “Middle Miocene Climate Transition”.
Line 872: Replace “Middle Miocene Climate Transition” with “MMCT”.
Line 931: “ wuellerstorfi” should be in italics.
Line 969: Insert “(Figure 6b)” after “5.3 Ma”.
Line 970: Delete “(Figure 6b)”.
Lines 978, 985 and 987: Replace “Fig. D” with “Fig. C”.
Line 998: “their water depths”. Do the indicated water depths correspond to those on the past or the present-day values?
Fig. 3: Add a geological timescale to the x-axis.
Fig. 5: Explain the grey shaded area. I suppose that it represents the 95% confidence level for GE19 UOT.