A three‐dimensional ocean mesoscale simulation using data from the SEMAPHORE experiment: Mixed layer heat budget

Abstract

A primitive equation model is used to simulate the mesoscale circulation associated with a portion of the Azores Front investigated during the intensive observation period (IOP) of the Structure des Echanges Mer‐Atmosphere, Proprietes des Heterogeneites Oceaniques: Recherche Experimentale (SEMAPHORE) experiment in fall 1993. The model is a mesoscale version of the ocean general circulation model (OGCM) developed at the Laboratoire d'Océanographie Dynamique et de Climatologie (LODYC) in Paris and includes open lateral boundaries, a 1.5‐level‐order turbulence closure scheme, and fine mesh resolution (0.11° for latitude and 0.09° for longitude). The atmospheric forcing is provided by satellite data for the solar and infrared fluxes and by analyzed (or reanalyzed for the wind) atmospheric data from the European Centre for Medium‐Range Weather Forecasts (ECMWF) forecast model. The extended data set collected during the IOP of SEMAPHORE enables a detailed initialization of the model, a coupling with the rest of the basin through time dependent open boundaries, and a model/data comparison for validation. The analysis of model outputs indicates that most features are in good agreement with independent available observations. The surface front evolution is subject to an intense deformation different from that of the deep front system, which evolves only weakly. An estimate of the upper layer heat budget is performed during the 22 days of the integration of the model. Each term of this budget is analyzed according to various atmospheric events that occurred during the experiment, such as the passage of a strong storm. This facilitates extended estimates of mixed layer or relevant surface processes beyond those which are obtainable directly from observations. Surface fluxes represent 54% of the heat loss in the mixed layer and 70% in the top 100‐m layer, while vertical transport at the mixed layer bottom accounts for 31% and three‐dimensional processes account for 14%.