While trawling for high-resolution ocean sediment cores, I noticed an interesting core offshore Morocco, reported by McGregor et al (Science 2007) earlier this year. This study was released just prior to IPCC complete with press release. Can’t we all think back to a quieter time when a scientist having labored to produce a detailed analysis of ocean sediments or tree cores would merely publish his study in an academic journal? When I first became familiar with the climate field, I was astonished at the idea of scientists issuing press releases that would make mining promoters cringe. This study is an interesting example of the counterplay between the fine print of the academic article and the promotion in the newspaper.
The heading to this thread is from an article by the Australian Broadcasting Corp (ABC) that went on to say:
Global warming puts fish stocks at risk Their analysis found that climate-induced changes in the ocean have never been more dramatic than in the past three or so decades.
The core in question was from 30.845N 10.0983W in 1999 at 355 m depth. GeoB6008-1 was 5.4 m long, sampled at 2 cm intervals, with resolution decreasing downcore from 6 to 25 years; the top value dated to 1971, the bottom to 518 BC, GeoB6008-2 was a short core sampled at 1 cm intervals and dated from 1998 at the top to 1912 at the bottom, with resolution decreasing from 2 to 4 years.
Alkenone values were obtained for use in SST calculations (available here) . The raw alkenone values are shown in the graph below against depth, and,as you see, they show a remarkable change through the 20th century. On the right is shown the HadSST gridcell value for the relevant gridcell, showing an increase of about 1 deg C. This core (31N) is just to the north of the “Main Development Region” for hurricanes (up to 20N). The vertical dashed line marks the depth allocated to the start of the 20th century in GeoB-1.
Left: Alkenone versus depth. The points show the short GeoB-2 core and the lines the long GeoB-1 core. The vertical dashed line marks the depth allocated to the start of the 20th century in GeoB-1. Right HadSST for gridcell 32.5N, 12.5 W containing the core location.
So this looks like evidence that Al Gore would love. Something unprecedented in the core. Fish stocks at stake. But there’s a catch (which Richard T may have spotted already.) The catch is nothing to do with bioturbation – although, as an aside, having seen similar sorts of changes through the mixed layer in connection with coarse fraction, I do wonder whether there might be some unconsidered interaction going on in the mixed layer that might account for this change. (But that’s not the issue today.)
Take one more look at the graph and see if you can figure out the sleight of hand that I’ve done here.
Here’s the trick. I’ve plotted the alkenone values upside down (relative to their use as temperature proxies.) The 20th century SST estimates based on alkenone show an unprecedented decrease in SST values so that there is a divergence of about 1.5 deg C between the alkenone SST values and HadSST values during the 20th century.
And despite the headline from ABC that “Global warming puts fish stocks at risk”, the follow up paragraph in that article and a contemporary Daily Telegraph article both reported unprecedented cooling:
Dr Helen McGregor and her research team from MARUM Research Centre Ocean Margins at the University of Bremen in Germany report in the journal Science that climate-induced changes in the ocean have never been more dramatic than in the past three or so decades. Sea surface temperatures in this part of the Atlantic Ocean declined by 1.2′C during the 20th century, say the researchers.
The seeming conundrum was explained as follows:
“The stronger the greenhouse effect the stronger the cold water pump works – and the cooler the coastal waters off Morocco,” says McGregor.
The Daily Telegraph article mentions that these upwelling systems are excellent for fishermen.
Upwelling systems like that off northwestern Africa are of great economic importance. Although they cover less than one per cent of the global sea surface, about 20% of global fishing takes place there and they are potentially important for drawing down carbon dioxide from the atmosphere.
However the scientists were able to find a cloud in the silver lining. McGregor warned that the fish might have trouble swimming in the nutrient-rich upwelling system:
While an increase in nutrients has the potential to increase fish stocks, the strength of the offshore currents may be too strong for fish to swim against, says McGregor.
I had innocently thought that fish were nicely adapted to the water. Will swimming lessons for fish be part of IPCC mitigation strategy? Inquiring minds want to know. McGregor also warned darkly of further bad consequences if fish were unable to swim in the nutrient-rich waters:
“Our research suggests that upwelling will intensify with future greenhouse warming, however it is unclear whether fish will be able to counter the stronger upwelling currents and benefit from the likely increase in phytoplankton (phytoplankton increase due to higher nutrient levels),” said Dr McGregor.
If phytoplankton populations are left unchecked then the higher deposition of organic matter on the sea floor could lead to the build-up of toxic hydrogen sulphide gas (rotten-egg gas), poisoning marine life and further degrading marine ecosystems.”
The news report ended by saying:
The new study is released just hours before the official release of a major IPCC update about global warming due to be unveiled in Paris this evening.
The Divergence Problem
The authors discuss the “Divergence problem” in their supplementary material as follows:
Note on instrumental versus alkenone SSTs: Direct comparison of the alkenone SSTs with instrumental SSTs is problematic, as ships-of-opportunity pass too far from the Cape Ghir coast to measure SST in the upwelling area (S35, S36). A compilation of sporadic SST measurements between 1947 and 1979AD collected by different research vessels operating in the Cape Ghir region, gave an estimated mean annual SST of 17.1′C (S37). In contrast, Hadley Centre SSTs (HadISST_1.1_SST -blended ship and, when available, satellite data (S38)), averaged over the 1′x1” grid square that includes site GeoB6008, and for the same period, gives an annual SST of 19.9′C. The absolute UK’37 SST for the uppermost sample for GeoB6008-1 and GeoB6008-2 are 18.7′C (1971AD) and 17.7′C (1998AD), respectively.
S35. W. S. Wooster, A. Bakun, D. R. McLain, Journal of Marine Research (1976).; S36. L. Nykj’r, L. Van Camp, Journal of Geophysical Research 99, 14197 (1994).S37. E. Hagen, C. Z’licke, R. Feistel, Oceanologica Acta 19, 577 (1996).
Also: this SST series with its sharp 20th century decline is sufficiently resolved to meet inclusion standards for multiproxy studies. Using Loehle’s methods, the series would be averaged. Under more “advanced” methods, the series would be flipped over and the increased cold interpreted as evidence of global warming, as ended up happening in Moberg and Juckes.
McGregor, Helen V; Dima, M; Fischer, Helmut W; Mulitza, Stefan (2007): Rapid 20th century increase in coastal upwelling off northwest Africa, Science, 315, 637-639, doi:10.1126/science.1134839. Also see PAGES newsletter here