Impact of Different Atmospheric Forcing Sets on Modeling Labrador Sea Water Production

Abstract

AbstractA numerical modeling sensitivity study is carried out within the Labrador Sea by varying the atmospheric conditions. From forcing NEMO simulations with five atmospheric products commonly used in ocean modeling (DFS5.2, ERA‐Interim, CGRF, ERA5, and JRA55‐do), we calculate the air–sea heat fluxes that occur over the Labrador Sea (2002–2015 annual‐average net heat flux: −53.4, −51.0, −46.6, −58.5, and−47.9 W m−2). With differences up to 12 W m−2 in net surface heat flux averaged over a central region of the Labrador Sea, each product supplied different atmospheric conditions. While the salinity‐dependent surface buoyancy fluxes were similar across all simulation, differences between each simulation's solar and nonsolar heat flux led to significant changes in the level of stratification (up to 400 J m−3), depth of the mixed layer (up to 300 m), and thickness of Labrador Sea Water (LSW; up to 300 m). Greater buoyancy loss from the Labrador Sea produced LSW with greater density. However, the production rate of LSW was not clearly affected by small changes in the surface buoyancy flux.