Accelerating our knowledge of carbon cycling in the ocean: Accelerator mass spectrometry and the WOCE/CLIVAR/GO-SHIP Programs

Increasing the density of radiocarbon measurements in the oceans was recognized as important to achieving the goals of the World Ocean Circulation Experiment. The only way to obtain the required number of measurements was to use accelerator mass spectrometry (AMS), a relatively new technology in 1987, for the measurements. The use of AMS reduced the sample size as well as the analysis time by orders of magnitude. Realizing this goal in the United States meant building an AMS, a facility to house it, and the laboratory required to prepare the samples, as well as developing all the necessary protocols. The result was the National Ocean Sciences Accelerator Mass Spectrometry Facility (NOSAMS)

To date, NOSAMS has analyzed over 25,000 individual seawater samples and reported their DI¹³C and DI¹⁴C analyses to the community. Other groups that have contributed significant numbers of analyses are based in Japan, Germany, Australia, New Zealand, and Norway. These records are beginning to show the significance of these measurements as well as the value of continuing these surveys in the future. Comparison of measurements made at similar locations decades apart requires that the measurement protocols be robust and the results reproducible. We will provide evidence of the robustness of the dataset at NOSAMS as well as between laboratories.

One of the initial uses of the WOCE data was to evaluate ocean general circulation models, none of which were able to predict the observed distribution. More recently, researchers are using the time series measurements to estimate the transport time of water parcels in the ocean, reexamine the age of deeper ocean waters, and to demonstrate that the oceanic radiocarbon distribution is now controlled by internal ocean processes, not air-sea exchange. Advances in accelerator technology coupled with changing sample processing methods suggest that radiocarbon will continue to provide critical information for understanding the past and present as well as predicting future changes. The NOSAMS dataset is sufficient to demonstrate marked improvement in ocean models over the past decades, but initialization of and calibration to the measured decadal changes is still extremely valuable to the modeling community.