The Dynamical Proxy Potential of the OSNAP Array

The first data from the OSNAP (Overturning in the Subpolar North Atlantic Program) array, a recently installed observing system in the subpolar North Atlantic, has provided new insights into volume, heat, and freshwater transports at the latitudes of the array. An important next step is to put the OSNAP observations into a broader spatial and temporal context: Can the OSNAP array inform - or could even serve as a proxy for - unobserved hydrographic and circulation quantities remote from the array? Here, we quantify the proxy potential of the OSNAP array by means of a novel, dynamics-based technique that uses adjoint modeling and uncertainty quantification within the ECCO (Estimating the Circulation and Climate of the Ocean) state estimation framework.

Adjoint-derived sensitivities reveal that the eastern boundary of the North Atlantic and the coasts of Iceland and Greenland are important pathways for communicating wind-driven pressure anomalies around the entire subpolar North Atlantic. Consequently, the OSNAP observing array samples climate signals that also impact remote oceanic quantities and regions, for instance ocean heat content close to Greenland’s margins, and has potential to inform these remote regions. We find that heat transport inferences across the OSNAP-West transect, extending from Labrador to South Greenland, impose an overall much stronger constraint on the ECCO state estimate than heat transport inferences across the OSNAP-East transect, extending from South Greenland to Scotland. This is largely explained by the fact that signals detected by OSNAP-West are less noisy compared to signals detected by OSNAP-East. As a result, oceanic quantities as remote as in the Nordic Seas may be constrained more efficiently by OSNAP-West than OSNAP-East observations, contrary to recent findings.

Observing systems, such as the OSNAP array and other observational efforts in the Atlantic, are expensive to deploy and maintain, and therefore often rely on short-term funding periods. Our novel method can be used to support the design of an effective, long-term Atlantic observing system, owing to the following features: the method is fundamentally dynamics-based, takes into account data redundancy between observations, and can evaluate not only existing, but also future and hypothetical observing systems.