A century and half of reconstructed ocean warming offers clues for the future

Patrick Heimbach

Due to a scarcity of direct observational data, most global estimates of ocean warming start only in the 1950s. However, a team of international scientists including ICES Professor Patrick Heimbach has now succeeded in reconstructing ocean temperature change from 1871 to 2017.

Over the past century, increased greenhouse gas emissions have given rise to an excess of energy in the Earth system. More than 90% of this excess energy has been absorbed by the ocean, leading to increased ocean temperatures and associated sea level rise, while moderating surface warming.

The multi-disciplinary team of scientists has published estimates in the Proceedings of the National Academy of Sciences (PNAS), that global warming of the oceans of 436 x 10^21 Joules has occurred from 1871 to present (roughly 1000 times annual worldwide human primary energy consumption) and that comparable warming happened over the periods 1920-1945 and 1990-2015.

The estimates support evidence that the oceans are absorbing most of the excess energy in the climate system arising from greenhouse gases emitted by human activities.

Professor Laure Zanna of the University of Oxford, who led the international team of researchers said: “Our reconstruction is in line with other direct estimates and provides evidence for ocean warming before the 1950s.”

The new estimate relied on a sequence of mathematical and numerical simulation methods. As part of an initial step, Heimbach, a professor in the Jackson School of Geosciences and leader of the ICES Computational Research in Ice and Ocean Systems Group, oversaw the production of a modern-day (1992 onward) state estimate of the global ocean circulation that optimally combines satellite and in-situ measurements with a global ocean circulation model. The effort is part of a NASA-funded consortium for Estimating the Circulation and Climate of the Ocean (ECCO). From this estimate, team members at Oxford and the University of Reading derived the extended reconstruction by a technique known as Green’s function method, originally developed by Prof Samar Khatiwala.

“Simulation-based science and engineering plays an important role in ocean and climate research,” said Heimbach. “Advanced computational algorithms help us to tackle the reconstruction problem, to further narrow the range in reconstructions of past changes, and to improve our confidence in simulations of future climate change. Adding another angle of view on the problem provides us with a consistent quantitative picture of a complex system that is the global ocean circulation and its role in climate.”

Much work remains to be done to validate the method. Computational science has powerful tools in store for characterizing and reducing the range of plausible reconstructions. This is important, particularly in the earlier part of the reconstruction, where direct observations are increasingly sparse. Nevertheless, the consistency of the new estimate with direct temperature measurements gives the team confidence in their approach.

This work offers an answer to an important gap in knowledge of ocean warming, but is only a first step. It is important to understand the cause of the ocean circulation changes to help predict future patterns of warming and sea level rise. Bringing advanced methods of computational science and engineering to bear on this problem remains an active field of convergence research across the mathematical, computational and geoscience communities.

The research was cited in the New York Times' discussion of the speed of climate change and the ocean's role to buffer the change.

Posted: Jan. 8, 2019