Comparing a Multivariate Global Ocean State Estimate With High-Resolution in Situ Data: An Anticyclonic Intrathermocline Eddy Near the Canary Islands

The provision of high-resolution in situ oceanographic data is key for the ongoing verification, validation and assessment of operational products, such as those provided by the Copernicus Marine Core Service (CMEMS). Here we analyze the ability of ARMOR3D -a multivariate global ocean state estimate that is available from CMEMS-to reconstruct a mesoscale anticyclonic intrathermocline eddy that was previously sampled with high-resolution independent in situ observations. ARMOR3D is constructed by merging remote sensing observations with in situ vertical profiles of temperature and salinity obtained primarily from the Argo network. In situ data from CTDs and an Acoustic Doppler Current Profiler were obtained during an oceanographic cruise near the Canary Islands (Atlantic ocean). The analysis of the ARMOR3D product using the in situ data is done over (i) a high-resolution meridional transect crossing the eddy center and (ii) a three-dimensional grid centered on the eddy center. An evaluation of the hydrographic eddy signature and derived dynamical variables, namely geostrophic velocity, vertical vorticity and quasi-geostrophic (QG) vertical velocity, demonstrates that the ARMOR3D product is able to reproduce the vertical hydrographic structure of the independently sampled eddy below the seasonal pycnocline, with the caveat that the flow is surface intensified and the seasonal pycnocline remains flat. Maps of ARMOR3D density show the signature of the eddy, and agreement with the elliptical eddy shape seen in the in situ data. The major eddy axes are oriented NW-SE in both data sets. The estimated radius for the in situ eddy is similar to 46 km; the ARMOR3D radius is significantly larger at similar to 92 km and is considered an overestimation that is inherited from an across-track altimetry sampling issue. The ARMOR3D geostrophic flow is underestimated by a factor of 2, with maxima of 0.11 (-0.19) m s(-1) at the surface, which implies an underestimation of the local Rossby number by a factor of 3. Both the in situ and ARMOR3D eddies have decelerating flows at their northern edges. The ARMOR3D QG vertical velocity distribution has upwelling/downwelling cells located along the eddy periphery and similar magnitudes to the in situ-derived QG vertical velocity.