Anthropogenic Carbon in the Arctic Ocean: Perspectives from different TTD Approaches and Tracer Pairs

At present, it is well-known that the fast increase in atmospheric carbon dioxide (CO₂) concentrations resulting from human activities (C_ant), drives the dramatic changes observed in our environment such as global warming and ocean acidification. The Arctic Ocean has been identified as one of the fastest-changing regions of the world ocean and can therefore be considered as a sentinel for future global scenarios.

Here, C_ant-rich waters coming from the Atlantic Ocean become isolated from the atmospheric input of CO₂ as they flow at an intermediate depth below the mixed layer, making the Arctic Ocean a key region for intermediate-to-long-term storage of C_ant. Despite having such an important role, the magnitude of the C_ant inventory and its change over time in the region is yet not fully understood, particularly if we are to consider future changes in ice coverage and therefore ocean circulation.

A way of estimating oceanic C_ant inventories is by applying the so-called Transit Time Distribution (TTD) method, which implies the use of transient tracers such as the anthropogenically produced CFC-12 and SF₆.

In this work we present a new estimate of C_ant inventory for the Arctic Ocean in 2015 assessed with the TTD method using both well-established tracers (CFC-12 and SF₆, both having a global source) as well as novel ones (anthropogenic radionuclides ¹²⁹I and ²³⁶U, both having primarily a point-like source represented by European nuclear reprocessing plants, as well as a global one represented by the global fallout from nuclear bomb testing).

The TTD was here applied following a relatively novel approach to infer the statistical parameters that describe the age distribution within a water sample, the mean (G) and the width (D). Unlike the “classical TTD” approach, the one used in this study allows the statistical parameters of the TTD to be constrained for each individual sample rather than finding values that are most representative of the region and time studied. We first show a comparison of the two TTD approaches by comparing mean and mode ages as well D/G ratios of this study (new TTD method) to those presented in Rajasakaren et al. 2019 (classical TTD method), using CFC-12 and SF₆ as our tracers’ pair. We then compare TTD results obtained from the two tracers’ pairs, CFC-12/SF₆ and ¹²⁹I-/²³⁶U, using the new TTD method.

Finally, we estimate and compare C_ant concentrations and inventories obtained with the two pairs of transient tracers to one-another as well as to previous estimates of C_ant in the region by Rajasakaren et al (2019) obtained with the “classical TTD”. This study demonstrates for the first time the feasibility of using anthropogenically produced radionuclides with input functions and chemical properties different than CO₂ as proxies for C_ant estimates.