The oxygen isotopic composition of benthic foraminifera has long been used a proxy for total ice volume. As ocean water evaporates, it becomes preferentially enriched in 16O; as water vapor rains out over ice sheets, this 16O is stored in ice sheets–particularly Antarctica, Greenland, and, during the Last Glacial Maximum, the Laurentide and Fennoscandian ice sheets. Consequently, the oceans become enriched in 18O. The precise oxygen isotope composition of foraminifera is a function of total ice sheet volume. Though the oxygen isotope composition of foraminifera is also affected by temperature, this principle has been used since the advent of paleoceanography research to understand the waxing and waning of the major ice sheets.
However, the above picture neglects a critically important–and long-recognized–phenomenon. The oxygen isotopic composition of ice is known to vary with temperature, both spatially and through time. In short, in a warmer climate, ice sheets contain relatively more 18O than in a colder climate. Thus, changes in benthic foraminifera oxygen isotopes can occur due to both changes in ice volume/sea level or changes in climate. In fact, these processes act in concert. For example, as climate cools, ice sheets grow larger, sequestering more 16O in the ice sheets. In turn, as the climate cools, more and more 16O is cycled into the ice sheet. Both processes leave the ocean (and benthic foraminifera) enriched in 18O. Remarkably, this climatic phenomenon has been ignored in studies of Cenozoic ice sheet volume and sea level.
Working with Matt Winnick, we applied this principle to estimates of sea level and ice sheet volume during the mid-Pliocene Warm Period. The mid-Pliocene is a period when CO2 levels were roughly the same as the modern (400 ppm), temperatures were globally elevated by 2-3° C, and sea level is estimated to have been anywhere from 1-40 meters higher than today. Previous sea level estimates based upon oxygen isotopes suggested sea levels were 21 m higher. Such a sea level would involve not just complete melting of Greenland and West Antarctica, but also a substantial collapse of the East Antarctic ice sheet. Using well-constrained relationships between the oxygen isotopic composition of ice and temperature, we revise these estimates, and suggest that sea level was likely 9-13.5 m higher than present, and almost certainly no higher than 17 m (Winnick and Caves, 2015). As a consequence, the East Antarctic ice sheet likely only melted by a maximum of 10%, or not at all. The good news is that this estimate is largely in line with many ice sheet models, which frequently predict only a small or negligible melt of East Antarctica when CO2 levels are 400 ppm.