『Abstract
The global surface seawater dimethylsulphide (DMS) database (http://saga.pmel.noaa.gov/dms/)
contains ＞50,000 data points and is the second largest trace gas
database after carbon dioxide. However, there has been relatively
little quality control on the data that have been collated to
data. Furthermore, the recent development of technologies capable
of high frequency (＞2 Hz) DMS measurements will have a disproportionate
effect on the database in future years. At this juncture, the
comparability of analytical techniques, sample handling methodologies
and standards are pressing issues that the DMS community needs
to address. In October 2010, during the Fifth International Symposium
on Biological and Environmental Chemistry of DMS(O) and Related
Compounds held in Goa, India, attendees participated in a discussion
concerning the current DMS database and its future development.
We develop some of the ideas from that session and combine them
with available data. From the few inter-comparison exercises that
have been conducted we show that variability between existing
measurements within the DMS database is likely to be ≦ 25%. Tests
comparing different DMSP・HCl standards demonstrate that a reference
calibration standard would be beneficial for the DMS community.
Confidence in future data collation would be substantially improved
with a comprehensive inter-comparison experiment between new analytical
techniques and sampling methodologies (e.g., mass spectrometers
with equilibrators attached to a continuous flow of seawater)
and more established methods (i.e., filtered samples analysed
with purge and trap gas chromatography). We conclude with recommendations
for the future expansion of the DMS database and its data quality
control.
Keywords: DMS; Dimethylsulphide; DMSP; Dimethylsulphoniopropionate;
Data comparability; Quality control; Reference standard』
Introduction
Comparability of standards
Analytical methodologies and sample handling
Filtration
Equilibration systems
Previous field inter-comparison exercises
Recommendations for future data collation
Conclusions and recommendations
Acknowledgments
References