Temperature flux carried by individual eddies across 47°N in the Atlantic Ocean

Abstract

Surface geostrophic velocity fields derived from satellite altimetry between January 1993 and April 2014 are used to detect and investigate eddies in the North Atlantic between 40°N–55°N and 60°W–10°W. Focus is on a zonal section along 47°N, roughly at the boundary between the subpolar and the subtropical gyres. Sea surface temperature data are used to quantify the temperature anomalies associated with eddies and the respective surface temperature fluxes related to these eddies. Identified eddy pathways across 47°N are related to the mean background velocity from full‐depth ship observations carried out on 11 cruises between 2003 and 2014. The analysis is repeated in two model simulations with 1/4° and 1/12° horizontal resolution, respectively, for the period 2002–2013. The analysis reveals almost 37,000 altimeter‐derived eddies with a lifetime longer than 1 week in the area. The highest number of eddies is found along the pathway of the North Atlantic Current, roughly following the 4000 m isobath, and on the Grand Banks of Newfoundland. Time series of temperature fluxes by eddies crossing 47°N reveal that single isolated eddies with large SST signatures contribute ∼25% to the surface temperature flux. Relating the observed eddies to the observed top‐to‐bottom velocity distribution at 47°N points to the existence of eddy pathways across 47°. The highest‐temperature fluxes are linked to the fastest and most pronounced current branches in the western Newfoundland Basin. While there are fewer eddies in both model simulations, the key findings are consistent between the observations and the two model simulations.

Sensitivity of Labrador Sea Water Formation to Changes in Model Resolution, Atmospheric Forcing, and Freshwater Input

“…The meridional velocities have a similar behavior, with an increase mainly toward the south. When using SPG12 to explore the temperature flux carried by individual eddies across 47°N in the Atlantic Ocean, Müller et al () compared its (SPG12) mean circulation patterns with satellite altimetry data, finding that the LC appeared to be more pronounced in SPG12. This might be a reason for the increase eastward and southward velocities found in our study region.…”

“…Dukhovskoy et al () used the same configuration to look at the spreading of Greenland freshwater in the sub‐Arctic Seas, while Gillard et al () explored the pathways of the melt‐water from marine terminating glaciers of the GrIS. Müller et al () used the ANHA4 configuration to explore the temperature flux carried by individual eddies across 47°N in the Atlantic Ocean. They ran two different simulations using the ANHA4 configuration: one simulation with a horizontal resolution of (1/4)° over the entire domain and a second simulation with a two‐way nest over the subpolar gyre to increase the horizontal resolution to (1/12)°.…”

“…They ran two different simulations using the ANHA4 configuration: one simulation with a horizontal resolution of (1/4)° over the entire domain and a second simulation with a two‐way nest over the subpolar gyre to increase the horizontal resolution to (1/12)°. Both simulations, called ANHA4 and ANHA4‐SPG12 (in Müller et al, ), are used in the present study. Here nevertheless they will be found under the names Control and SPG12, respectively.…”

“…Müller et al () compared the mean circulation patterns from Control and SPG12 with that from satellite altimetry, with a focus on the region located between 60° to 5°W and 40° to 55°N within the subpolar gyre, finding a good correspondence between them. However, they did point out, as one of the differences, a more pronounced coastal branch of the LC represented in the simulations compared to that from altimetry data.…”

Heat and Freshwater Transport by Mesoscale Eddies in the Southern Subpolar North Atlantic

“…Before applying the algorithm, the velocity fields from the observations were linearly interpolated from 1/4°×1/4°to 1/6°×1/6°resolution. This did not change any features in the velocity field while improving the algorithm's performance (Liu et al, 2012;Müller et al, 2017). In the model simulation no further refinement was necessary since the grid had already a resolution of 1/12°×1/12°.…”