The Macquarie Island (LoFlo2G) high-precision continuous atmospheric carbon dioxide record

• Published in: Atmospheric Measurement Techniques
• Main Authors: A. R. Stavert, R. M. Law, M. van der Schoot, R. L. Langenfelds, D. A. Spencer, P. B. Krummel, S. D. Chambers, A. G. Williams, S. Werczynski, R. J. Francey, R. T. Howden
• Format: Article in Journal/Newspaper
• Language: English
• Published: Copernicus Publications 2019
• Subjects: Environmental engineering; TA170-171; Earthwork. Foundations; TA715-787; Southern Ocean; Macquarie Island
• Online Access: https://doi.org/10.5194/amt-12-1103-2019; https://doaj.org/article/887524a7fb8c4f40b77e2996a6a5eef7

The Southern Ocean (south of 30 ∘ S) is a key global-scale sink of carbon dioxide ( CO 2 ). However, the isolated and inhospitable nature of this environment has restricted the number of oceanic and atmospheric CO 2 measurements in this region. This has limited the scientific community's ability to investigate trends and seasonal variability of the sink. Compared to regions further north, the near-absence of terrestrial CO 2 exchange and strong large-scale zonal mixing demands unusual inter-site measurement precision to help distinguish the presence of midlatitude to high latitude ocean exchange from large CO 2 fluxes transported southwards in the atmosphere. Here we describe a continuous, in situ, ultra-high-precision Southern Ocean region CO 2 record, which ran at Macquarie Island (54 ∘ 37 ′ S, 158 ∘ 52 ′ E) from 2005 to 2016 using a LoFlo2 instrument, along with its calibration strategy, uncertainty analysis and baseline filtering procedures. Uncertainty estimates calculated for minute and hourly frequency data range from 0.01 to 0.05 µmol mol −1 depending on the averaging period and application. Higher precisions are applicable when comparing Macquarie Island LoFlo measurements to those of similar instruments on the same internal laboratory calibration scale and more uncertain values are applicable when comparing to other networks. Baseline selection is designed to remove measurements that are influenced by local Macquarie Island CO 2 sources, with effective removal achieved using a within-minute CO 2 standard deviation metric. Additionally, measurements that are influenced by CO 2 fluxes from Australia or other Southern Hemisphere land masses are effectively removed using model-simulated radon concentration. A comparison with flask records of atmospheric CO 2 at Macquarie Island highlights the limitation of the flask record (due to corrections for storage time and limited temporal coverage) when compared to the new high-precision, continuous record: the new record shows much less noisy seasonal variations ...