IOCCP-JAMSTEC 2018 Inter-laboratory calibration exercise of a certified reference material for nutrients in seawater

Abstract

In 2017, the International Ocean Carbon Coordination Project (IOCCP) and the Japan Agency for Marine-Earth Science and Technology (JAMSTEC) decided to conduct the 6th International Inter-Calibration Exercise, the “2017/18 inter-laboratory comparison exercise of Certified Reference Materials of Nutrients in Seawater, CRM”. As was the case with the previous four inter-comparison (IC) studies organized by MRI-JMA and the previous IOCCCP-JAMSTEC IC exercise in 2015, the aim of this IC exercise was to improve the level of comparability by exchanging knowledge among participating laboratories. The scale of the study was expanded. One hundred seven laboratories in 33 countries across five continents received a letter of invitation to the IC exercise, and 71 laboratories in 30 counties agreed to participate in this IC exercise. Results were returned from 69 laboratories in 30 countries. The data were statistically analyzed, and the results are described in this report. The agreement between consensus median/mean and certified values was within uncertainty for all five samples used in this IC exercise. Only small discrepancies existed among core laboratories that reported close-to-consensus values and certified values. Normalized cumulative distributions for nitrate and phosphate obtained in 2018 were better (i.e., flatter) than the normalized cumulative distributions obtained from previous IC exercises conducted in 2008, 2012, and 2015. The indication is that comparability of nitrate and phosphate analysis among the laboratories has gradually improved from 2008 to 2018. In contrast with the nitrate and phosphate results, the normalized cumulative distribution for silicate obtained in 2018 was similar to the distributions in previous years. The indication is that comparability of silicate analysis among the laboratories has not changed during these years. The results of this IC exercise also showed that nonlinearity of the calibration curves for the nutrient analyses in each laboratory is still a significant source of reduced comparability of nutrient data. Analysis of ranked scatter plots led to this conclusion. The implication is that we need to use a set of nutrient CRMs, the concentrations of which cover the whole range of nutrient concentrations in the world’s oceans to maintain comparability of results. Thirty-eight laboratories among 69 laboratories used a CRM/RM. The remaining 31 laboratories did not use a CRM/RM or did not reply to the questionnaire. It is obvious that the number of the laboratories that use CRMs has been increasing since 2008. In general, the reported results from laboratories that used CRMs were located in the central part of the ranked plots and showed good Z-scores, as expected. The results of this IC exercise imply that the majority of the participating laboratories are very capable of measuring nutrient concentrations in seawater, and using CRMs will further increase the comparability of results. The results may be SI traceable in the near future.