Global ocean heat content over the past 3 million years
Summary
The authors reconstruct mean ocean temperature (MOT) and global ocean heat content across the past 3 million years using noble gas measurements (notably Xe/Kr ratios) from Antarctic blue-ice and deep ice cores, including Allan Hills samples. They combine these empirical snapshots with model simulations and comparisons to sea-surface-temperature (SST) based reconstructions to produce a long-term, splined record of ocean heat over multiple climate intervals (pre‑MPT, MPT and post‑MPT).
The study emphasises methods for identifying and correcting post‑depositional gas fractionation, evaluates the influence of sea‑ice coverage and ocean circulation on noble gas saturation state, and cross-checks noble gas MOT against SST reconstructions and transient climate model outputs. Data and code to reproduce the splined record are openly available (DOIs provided in the paper).
Key Points
- The reconstruction spans ~3 million years using noble gas isotope measurements from Antarctic ice (Allan Hills cores among others).
- Noble gas ratios (Xe/Kr, Kr/Ar, δ38/36Ar) are used to infer mean ocean temperature (MOT) and hence global ocean heat content.
- The record is compared with SST-based reconstructions and transient CGCM simulations to assess consistency and drivers of long‑term ocean temperature change.
- The authors explicitly examine intervals bracketing the Mid‑Pleistocene Transition (pre‑MPT, MPT, post‑MPT) to assess changes in noble gas saturation and climate state.
- Careful treatment of post‑depositional alteration and fractionation is central to trusting MOT estimates; extended data show depth diagnostics and sensitivity experiments (sea ice, circulation).
- Data are available from the US Antarctic Program Data Center (DOI) and code for the splined record is on Zenodo (DOI), enabling reproducibility.
Why should I read this
Short version: if you care about how hot the oceans really were across ice ages and warm intervals — and what that means for sea level and ice‑sheet stability — this paper gives you a cleaned, long‑run record and the data to play with. It’s the sort of study climate nerds and policy‑minded researchers will cite when arguing about past Earth energy imbalance and long‑term sea‑level benchmarks.
Author style
Punchy. The paper pairs careful lab measurements with model checks and openly shares data and code — so it reads like solid, reproducible palaeoclimate work. If you’re working on past climates, ocean heat, sea level or ice‑sheet sensitivity, dive into the methods and extended data.
Source
Article Date: 18 March 2026