Intercalibration in chemical oceanography—Getting the right number

Abstract

If you're a new biological oceanographer who is interested in the role of phosphorus in phytoplankton productivity and make up 10 L of a Redfield-like growth media with 3.2 mg nitrate-nitrogen and 0.3 mg phosphate-phosphorus, but the phytoplankton won't grow, you've made a systematic error. The Redfield ratio (Redfield et al. 1963) is atomic, not weightbased, and therefore your solution has an atomic N:P ratio of 16:0.7, not the 16:1.5 that you had planned for a phosphorusenriched culture. Similarly, if you were analyzing a seawater sample for cadmium using ICP-MS and hadn't accounted for isobaric molybdenum interference, you'd overestimate the actual concentration of cadmium in your sample. In both cases, you're not getting the "right" number, either because of miscalculations or from an analytical error. Getting the "right" number in chemical oceanography may seem like an obvious goal, but what is right? Is it getting the same value every time you analyze your sample, or the real value-something that you can trace to an absolute standard? The former is precision, the measurement of random errors, whereas the latter is accuracy, the measurement of random and systematic errors. A precise value is not necessarily accurate, but good accuracy requires good precision. The problem with determining the right or correct concentration of a chemical constituent in seawater is finding the best means to evaluate accuracy. Systematic errors can occur at each stage in the process of acquiring a sample, then storing it, and finally to analyzing it. Oceanic trace metals are in the nano-to picomolar concentration range, and thus working on metal ships assures that contamination is probable. In fact, sampling contamination was shown to be a factor that affected the accuracy of most trace metal data until the late 1970s (e.g., Bruland et al. 1979). From the analytical perspective, seawater is not a simple matrix, its high ionic strength makes analysis a difficult task, and combined with very low concentrations for constituents like trace metals, the quest for accuracy is made even harder.Most chemical oceanographers have used certified reference materials to establish accuracy at the analytical stage of their studies. Nevertheless, many of these are not actual marine materials, but close substitutes. In this respect, the need for appropriate certified marine reference materials (water and particles) was thoroughly addressed by the US Committee on Reference Materials for Ocean Science (NRC 2002), although very few of these materials have been produced to date. Interestingly, numerous laboratories are now determining trace element isotopes in seawater (e.g., Lacan et al. 2006, John andAtkins 2012), and it is likely that these Intercalibration in chemical oceanography-Getting the right number Gregory A. CutterDept. of Ocean, Earth, and Atmospheric Sciences, Old Dominion University, Norfolk, VA 23529, USA AbstractIntercalibration has a strict metrological definition, but in brief, it's an open sharing of methods and results between...

Coupling of the distribution of silicon isotopes to the meridional overturning circulation of the North Atlantic Ocean

“…Comparison of data from the one station that was sampled during both years (station USGT10-12 during 2010, station USGT11-24 during 2011) allows an assessment of combined sampling, analytical and environmental variability. The 2011 section also sampled the Bermuda Atlantic Time Series (BATS) station (USGT11-10) which was the site on the earlier US GEOTRACES intercalibration exercise in the Atlantic (Cutter, 2013).…”

“…The second approach employed the use of reference sea water collected as part of the prior US GEOTRACES intercalibration study at the BATS station (Cutter, 2013) to check data accuracy. Prior analysis of the reference waters collected on the intercalibration cruise from 4000 m in our laboratory gave an average δ 30 Si(OH) 4 value of þ1.41 70.08‰ compared to the overall average from analyses by four participating laboratories of þ1.45 70.05‰ (Brzezinski, unpublished).…”

Diagnosing oceanic nutrient deficiency

“…Public release of these data in the form of the consistent quality-controlled [61] GEOTRACES IDP2014 [20] has provided a hugely valuable community resource which promises to revolutionize our understanding of the sources, sinks and cycling of trace elements, alongside their interactions with oceanic microbial systems and wider biogeochemical cycles. Here, the IDP2014 is used to derive large-scale patterns in relative nutrient deficiency of multiple nutrient elements.…”

Insights into Particle Cycling from Thorium and Particle Data

“…Moreover, a significant intercalibration effort has already been undertaken to ensure the comparability of measurements from different laboratories (Cutter 2013). It is thus timely to revisit previous studies on particle cycling in the ocean, as the concurrent measurement of many particulate components as part of the GEOTRACES program could potentially add even more observational constraints.…”