Climate change represents an existential threat to humanity. The IPCC’s most optimistic future emissions projection, RCP2.6, models a further 0.3-1.7oC of global temperature rise by 2100. RCP8.5, an emissions projection based on a business as usual scenario, models a further 2.6-4.8oC temperature rise by 21001.
Global mean sea level rise is often cited as one of the gravest potential impacts of climate change. The largest potential sea level rise is stored in the polar ice sheets. The Greenland ice sheet sequesters around 7m of sea level equivalent and 58m of sea level equivalent is stored in Antarctica2.
RCP2.6 projects further global mean sea level rise of 0.26m-0.55m by 2100. RCP8.5 models rates of sea level rise approaching 16mm yr-1 by 2100, and total sea level rise approaching 1m3. Low lying cities such as Miami, New Orleans, Dhaka and Mumbai would be threatened by such magnitudes of sea level rise. In conjunction with heightened tropical cyclone frequency and magnitude4 sea level rise represents a genuine threat to human life.
The largest uncertainties in the IPCC’s future sea level rise projections arise from a lack of understanding of the dynamics and subsequent responses of the Greenland and Antarctic ice sheets to a warming world5. Given this barrier, much academic rigour has been placed upon understanding the response of sea level to warming climates of the past. The rationale behind such research is that past warming events represent (approximate) analogues of the present-day Earth system – and that the response of the polar ice sheets (and hence sea levels) of the past is likely comparable to their response to contemporary warming.
What are the limitations of such studies?
The Mid-Pliocene Warm Period (3.3-3.0 million years ago)
One such example of a time period commonly portrayed as analogous to the present day is the mid-Pliocene warm period. This period has been heavily studied because it represents the most recent period in Earth history during which atmospheric CO2 concentration was comparable to today (400ppm)6. Global mean surface temperature was approximately 2oC warmer as a result7. Ice sheet modelling from the period suggests there was widespread deglaciation of Greenland8, complete deglaciation of the West Antarctic ice sheet9 and minor deglaciation of the East Antarctic ice sheet10.
Sea level reconstructions from the period are collated using coral reefs and marine oxygen isotopes records. Coral reefs grow in shallow water and thus approximate the sea surface – hence coral reefs radiometrically dated from the mid-Pliocene warm period are used to reconstruct mid-Pliocene global mean sea level11. Variations in oxygen isotopes from benthic Foraminifera in marine sediment cores approximate global ice volume, and thus the eustatic component of sea level rise12.
Coral reef reconstructions are associated with large uncertainties because corals do not grow at the absolute sea surface and because glacio-isostatic adjustment is poorly accounted for13. Oxygen isotope reconstructions are also associated with large uncertainties because oxygen isotope variations can be induced solely by temperature variations14. Because of these limitations, uncertainties in sea level reconstructions from the mid-Pliocene warm period are commonly of the same magnitude as the reconstructed sea level rise15.
Coral reefs from the mid-Pliocene warm period suggest that global mean sea level was 10-30m higher than today16 and oxygen isotope records suggest that global mean sea level was 12-30m higher than today17. Despite the uncertainties attached to these estimates, it is widely accepted that mid-Pliocene sea levels were at least several metres higher than the present day, but likely no higher than 20m18.
Because of the uncertainties associated with such sea level reconstructions, calculation of an annual or decadal rate of sea level rise (which is relevant for present day anthropogenic climate change) has not been possible19. The total contributions of different polar ice sheets have been accurately modelled but the rates and mechanisms of decay of such ice sheets is largely unresolved.
The Last Glacial Termination (~20,000 years ago)
Another period widely cited as analogous to the present-day Earth system is the last glacial termination. This led to a 130m rise in global sea level over 13,000 years, largely due to the melting of the Eurasian, Laurentide and Cordilleran ice sheets20. Sea level from the period is better resolved than in the mid-Pliocene as glacio-isostatic effects have been significantly smaller over the last 20,000 years.
Because of the high-resolution sea level records available, rates of sea level rise from the period have been accurately resolved. The average rate of sea level rise during the last glacial termination was 13mm yr-1, but was as high as 40mm yr-1 during individual meltwater episodes21.
Because of the large uncertainties associated with past sea level reconstructions, rates of sea level rise can only be calculated for very recent events – such as the last glacial termination. Older reconstructions, such as those from the mid-Pliocene warm period, have such large uncertainties attached to them that their relevance to present day anthropogenic climate change is limited. Recent events, such as the last glacial termination, have much smaller errors associated with their reconstructions, and thus are more relevant as rates of sea level rise can be resolved.
However, recent sea level reconstructions are also associated with caveats. Reconstructions of past sea level capture the equilibrium response of sea level to the climate system (the response of sea level to climate change after all aspects of the climate system – e.g. cryosphere, ocean, atmosphere, vegetation – have reached an equilibrated state). Because emissions reductions are likely to occur prior to 2050, the Earth system is highly unlikely to ever reach an equilibrated state in response to anthropogenic climate change. Such an equilibrium would likely take thousands of years to attain.
Because of this, and because the equilibrium response of sea level is irrelevant for short term mitigation and adaptation strategies, we seek to determine the transient response of components of the Earth system to climate change (e.g. if the temperature goes up by one degree in 20 years, how much will sea level rise).
Unfortunately, periods where we have calculated the transient response of sea level to warming climates, such as the last glacial termination, are poor analogues for the present-day Earth system. The last glacial termination represents a period during which the Earth was undergoing fundamental transitions between a glacial and interglacial state. Such a transition is not akin to present day anthropogenic perturbations of the Earth system22.
In short, studies of periods of past sea level change are poorly resolved, and capture the equilibrium response of sea level to past warming; periods such as the last glacial termination and the mid-Pliocene warm period are also poor analogues of present day climate change. Studies that seek to quantify the transient response of individual components of the climate system (e.g. the Greenland or Antarctic ice sheets to mid-Pliocene warming) are vital in quantifying how such components respond under warming scenarios. However, citation of snapshot periods of past sea level rise as direct analogues for the present day is likely misguided.