Oxygen isotope analyses and deep-sea temperature changes: implications for rates of oceanic mixing

Abstract

Since the first measurements were made more than thirty years ago¹, analysis of oxygen isotopes in foraminifera from deep-sea sediments has gained widespread use as a palaeoclimatic and stratigraphic indicator. Isotopes stages are commonly used to define the ice ages² but controversy still remains regarding the relative contributions of temperature and ice-volume change to the down-core δ¹⁸O record. For the past decade, temperature has commonly been considered a trivial portion of the signal³. Recent reports, however, suggest that as much as one-third of the measured δ¹⁸O variations can be ascribed to cooler glacial-maximum temperatures^4,5. This interpretation requires mean deep-ocean temperatures <0 °C through much of late Quaternary time. Here we explore some of the consequences of this hypothesis for the heat budget of the deep sea, and conclude that either thermohaline overturn must have been faster than at present, or that cooling of the deep ocean at the glacial maximum may have been overestimated.