Tillys Petit, David Smeed, Adam T. Blaker, Shane Elipot, William E. Johns, Jules B. Kajtar, D. Rayner, Bablu Sinha, Ryan H. Smith, Denis L. Volkov, Ben Moat
Abstract Observation‐based estimates of the Atlantic Meridional Overturning Circulation (AMOC) and meridional heat transport (MHT) are necessary to better understand their evolution in the coming years. The RAPID‐MOCHA‐WBTS array at 26°N is the only trans‐Atlantic observing system to provide 20+ years of continuous measurements of the AMOC and MHT. While the design of the array has continuously evolved as our understanding of the AMOC has advanced, and as new technologies have become available, a new goal is to design a lower‐cost and more sustainable observing system to continue AMOC estimations with high accuracy. Using the RAPID array data and ocean reanalyzes, we evaluate the error in the AMOC estimate due to the choice of data included in its calculation. We find that the trend and variability of the volume transport in the upper 3,000‐m of the water column are not captured with sufficient accuracy by synoptic hydrographic data or ocean reanalyzes. However, moorings in the deep ocean interior along the eastern boundary and the Mid‐Atlantic ridge can be replaced by hydrographic data from repeat trans‐Atlantic hydrographic sections to reliably estimate the AMOC trend and variability. Experiments simulating the observing system in a high‐resolution ocean model further show that the additional error in the long‐term AMOC estimate induced by the substitution of mooring measurements below 3,000‐m depth at these locations is small (0.30 Sv) as compared to the AMOC uncertainty.