The upper layers of the world’s oceans have been warming much faster than oceanographers realised over the past few decades, according to a new study.
Sparse sampling of the Southern Hemisphere’s oceans have led to conservative estimates of warming. But the new research, published in Nature Climate Change, suggests that the amount of heat energy entering the upper 700 metres of these oceans has been underestimated by anything from 48% to 152%.
Because the Southern Hemisphere contains 60% of the world’s ocean, this means that globally, the oceans have been heating between 24% and 58% faster than previously thought.
Meanwhile, a separate study has found that very little of the warming since 2005 has penetrated the “bottom half” of the oceans, below 2 kilometres, although shallower waters have continued to warm.
Models and boats
More than 90% of the excess heat generated by global warming goes into the oceans, causing sea levels to rise as the water expands.
But while climate models were good at replicating the precise satellite observations of sea-level rise since 1993, they were inconsistent with global estimates of the oceans’ changing heat content, as calculated using the relatively few direct measurements of temperature taken in southern regions.
Paul Durack, a researcher at Lawrence Livermore National Laboratory, who led the new study, said temperatures in the Southern Ocean had only been sparsely sampled, which had led previous observations to underestimate the actual warming.
“Historical observations are much more numerous in the Northern Hemisphere because the developed countries (Europe, US, Japan) had boats which were travelling across the Northern Hemisphere oceans. As there’s less land in the Southern Hemisphere, this led to far fewer observations,” he said.
That situation was not remedied until 2004, when the Argo network of 3600 submersible floats began to produce global data on the world’s oceans.
Dr Durack and his colleagues calculated that southern ocean warming has been underestimated between 1970 and 2004, before researchers had access to the Argo network.
Heat not reaching the deepest seas
The Argo floats were used in the second study, to profile the movement of heat down to depths of 2 km.
The results suggest that sea-level rise is due to changes in the shallower oceans and to the addition of meltwater from land, rather than to thermal expansion in the deeper parts of the ocean, report researchers led by William Llovel of NASA’s Jet Propulsion Laboratory, also in Nature Climate Change.
“One thing that is clear from both these studies is the ocean is continuing to warm,” Dr Durack said.
Richard Allan, a climatologist at the University of Reading, said the results were consistent with his own previous research showing continued ocean heating from 1985 to 2012.
Professor Allan said the new research “helps to reconcile observed ocean heating with sea level rise”. Over the past two decades, seas have risen by an average of 3 mm a year.
Previously, the lack of data on Southern Hemisphere ocean temperatures led oceanographers to assume, for the sake of conservatism, that unsampled areas were not warming. The new data paint a much more realistic picture, Prof Allan said.
The new research also gives a clearer explanation of where the excess heat of global warming is going, Prof Allan said.
“The ocean heating rate is consistent with rising greenhouse gas concentrations. The slowing in global surface warming reflects a change in the vertical distribution of the heating to deeper layers below around 300 metres,” he said.
He added that the lack of heating expansion in the deepest oceans is “unfortunately not reassuring”, given that this situation may change in the future.
“Heating of the deep ocean below is not yet detectable, nor is any contribution to current rates of sea-level rise. Since the layer below 2000 metres contains half the ocean volume, as this begins to warm over many hundreds of years in response to the surface warming there is the potential for accelerating sea level rise.”
This article was amended on October 8, 2014, to clarify that the previous problem was with the sparse pre-2004 temperature measurements, not the climate models.