Interlaboratory study for coral Sr/Ca and other element/Ca ratio measurements. (23rd September 2013)

• Record Type: Journal Article
• Authors: Hathorne, Ed C.; Gagnon, Alex; Felis, Thomas; Adkins, Jess; Asami, Ryuji; Boer, Wim; Caillon, Nicolas; Case, David; Cobb, Kim M.; Douville, Eric; deMenocal, Peter; Eisenhauer, Anton; Garbe‐Schönberg, Dieter; Geibert, Walter; Goldstein, Steven; Hughen, Konrad; Inoue, Mayuri; Kawahata, Hodaka; Kölling, Martin; Cornec, Florence L.; Linsley, Braddock K.; McGregor, Helen V.; Montagna, Paolo; Nurhati, Intan S.; Quinn, Terrence M.; Raddatz, Jacek; Rebaubier, Hélène; Robinson, Laura; Sadekov, Aleksey; Sherrell, Rob; Sinclair, Dan; Tudhope, Alexander W.; Wei, Gangjian; Wong, Henri; Wu, Henry C.; You, Chen‐Feng;
• Is Part Of: Geochemistry, geophysics, geosystems. Volume 14:Number 9(2013)
• Page Start: 3730
• Page End: 3750
• Publication Date: 2013-09-23
• Subjects: Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 550.5
• DOI: 10.1002/ggge.20230
• Languages: English
• ISSNs: 1525-2027

Abstract

<abstract abstract-type="main"> <title> <x xml:space="preserve">Abstract</x> </title> <p>[1] The Sr/Ca ratio of coral aragonite is used to reconstruct past sea surface temperature (SST). Twenty‐one laboratories took part in an interlaboratory study of coral Sr/Ca measurements. Results show interlaboratory bias can be significant, and in the extreme case could result in a range in SST estimates of 7°C. However, most of the data fall within a narrower range and the <italic>Porites</italic> coral reference material JCp‐1 is now characterized well enough to have a certified Sr/Ca value of 8.838 mmol/mol with an expanded uncertainty of 0.089 mmol/mol following International Association of Geoanalysts (IAG) guidelines. This uncertainty, at the 95% confidence level, equates to 1.5°C for SST estimates using <italic>Porites</italic>, so is approaching fitness for purpose. The comparable median within laboratory error is <0.5°C. This difference in uncertainties illustrates the interlaboratory bias component that should be reduced through the use of reference materials like the JCp‐1. There are many potential sources contributing to biases in comparative methods but traces of Sr in Ca standards and uncertainties in reference solution composition can account for half of the combined uncertainty. Consensus values that fulfil the requirements to be certified values were also obtained for Mg/Ca in JCp‐1 and for Sr/Ca and Mg/Ca ratios in the JCt‐1 giant clam reference material. Reference values with variable fitness for purpose have also been obtained for Li/Ca, B/Ca, Ba/Ca, and U/Ca in both reference materials. In future, studies reporting coral element/Ca data should also report the average value obtained for a reference material such as the JCp‐1.</p> </abstract>