One of the great strengths of science is the acceptance that new data may require a revision of past conclusions.
This is particularly true when the new data – or a different interpretation of existing and published data – reveals mistakes. There has just been a good example of this.
Measuring ocean temperatures
On 1st November this year a joint team at the Scripps Institution of Oceanography at the University of California San Diego, and Princeton University published a paper in Nature entitled Quantification of ocean heat uptake from changes in atmospheric O2 an CO2 composition.
The authors of the paper were Resplandy (Princeton University), Keeling, Eddebbar, Brooks (Scripps Institution of Oceanography, USA), Dunne (Geophysical Fluid Dynamics Laboratory, USA), Long (National Center for Atmospheric Research, USA), Koeve. and Oschlies (Helmholtz Centre for Ocean Research, Germany), Bopp (Sorbonne University, France), and Wang (Fudan University, China).
Basically, their report stated that the Earth’s oceans are warming up faster than previously thought. This was really big news in the context of concern about climate change and received much publicity.
The group carried out their research by measuring atmospheric oxygen and carbon dioxide. The levels of these gases increase as the ocean warms and the gases are released. This concept was said to be “a whole-ocean thermometer”.
The results showed that the oceans gained 1.33 ± 0.20 x 1022 joules (13.3 zettajoules ) of heat per year between 1991 and 2016. This is equivalent to a planetary energy imbalance of 0.83 ± 0.11 watts per m2 of the Earth’s surface. This figure was 60% greater than the most recent assessment of the Intergovernmental Panel on Climate Change (IPCC).
These results were interpreted by the authors to mean that ocean warming is at the “high end of previous estimates” and that the Earth was more sensitive to fossil-fuel emissions than previously thought. This would mean that emissions of greenhouse gases produced by human activities would have to be reduced by 25% more than previously estimated to avoid temperatures increasing by 2ºC (3.6ºF) above pre-industrial levels.
If all this was true, the there would be significant implications for policy-relevant measurements of the Earth’s response to climate change “such as climate sensitivity to greenhouse gases and the thermal component of sea-level rise”.
Problems interpreting ocean temperature data
However, on 6th November, a UK-based scientist, Nicholas Lewis, published a blog entitled Major problem with the Resplandy et al. ocean heat uptake paper. On examining the data presented in the paper Lewis concluded that the ocean heat uptake was 10.1 zettajoules and not the 13.3 zettajoules claimed. This meant that the trend in ocean heat content was below the average for 1993-2016. Lewis makes a number of other critical comments on the data and its interpretation.
As Lewis says in his blog “Because of the wide dissemination of the paper’s results, it is extremely important that these errors are acknowledged by the authors without delay and then corrected”.
On 9th November, a note from one of the authors (Ralph Keeling) appeared on the Scripps website. This note recognised two of the problems brought up by Lewis (a) incorrectly treating systematic errors in the oxygen measurements, and (b) the use of a constant land oxygen:carbon change ratio of 1:1.
The note continued to say that the paper’s authors were recalculating the results but felt that these problems “do not invalidate the study’s methodology or the new insights into ocean biogeochemistry on which it is based”. It was expected that the corrections would have a small impact on the calculations of overall heat uptake “but with larger margins of error”.
I think the response by Ralph Keeling is laudable. He is quoted as saying in the Washington Post “Unfortunately, we made mistake here. I think the main lesson is that you work as fast as you can to fix mistakes when you find them”. Keeling issued a more detailed comment on the mistakes and thanked Nic Lewis for bring the problem to the authors’ attention.
Keeling concluded that “The revised uncertainties preclude drawing any strong conclusions with respect to climate sensitivity or carbon budgets based on the APO (atmospheric potential oxygen) method alone, but they still lend support for the implications of the recent upwards revisions in OHC (ocean heat content) relative to the IPCC AR5 (recent report by the IPCC) based on hydrographic and Argomeasurements (Argo consists of a global array of floats in the oceans that measure temperatures).
Nature also made a statement to the Washington Post, emphasising the important to the publication of ensuring accuracy and accepting their responsibility to correct errors in published papers.
I think that Gavin Schmidt, head of the NASA Goddard Institute for Space Studies, in a statement to the Washington Post, summed up this situation perfectly: “The key is not whether mistakes are made, but how they are dealt with — and the response from Laure (Resplandy) and Ralph (Keeling) here is exemplary. No panic, but a careful re-examination of their working”.