Intercomparison of observation-based anthropogenic carbon estimates across the Atlantic Ocean using hydrographic cruise data

Since the industrial revolution, the ocean has absorbed nearly one third of anthropogenic carbon dioxide (C_ant), playing a key role in climate regulation while marine systems simultaneously undergo substantial stress. The storage of C_ant in the ocean is spatially heterogeneous, making its quantification both challenging and essential. Because C_ant cannot be measured directly, its estimation relies on indirect approaches that can be broadly classified into model-based and observation-based methods. Observation-based approaches are particularly valuable, as they commonly serve as benchmarks for evaluating model performance. These methods include those based on transient tracer data such as CFCs (e.g., Transit Time Distributions, TTD) as well as approaches relying on marine carbonate system observations, including back-calculation techniques or repeated measurements used to infer decadal carbon changes. In this study, we estimate C_ant concentrations using carbon and CFC data from the GO-SHIP A25-OVIDE and A05-RAPID sections in the North Atlantic and the GO-SHIP A17-FICARAM section in the South Atlantic. C_ant is derived using several observation-based methodologies, including the φ-method, Cant-TMI, TrOCA, TRACE, and a TTD-based approach. This intercomparison provides an up-to-date assessment of widely used observation-based techniques and offers a detailed analysis of their underlying assumptions, methodological differences, areas of convergence, and inherent limitations.