Production of monthly mean climatological archives for the Nordic Seas

Abstract

Abstract Considered is the possible use of numerical models to enhance the information contained in oceanographic observations. For this purpose a three-dimensional baroclinic, primitive equation ocean model was used to produce climatological dynamically consistent data archives containing monthly mean fields of sea surface elevation, currents, salinity and temperature. The archives cover the North Sea, the Norwegian and Greenland Seas, the Barents Sea, the Kara Sea and parts of the Arctic Ocean with a horizontal grid size of 20 km. In the vertical the fields are stored at 31 standard oceanographic levels from 0 to 4500 m. First, a Hydrographic Archive was produced based on climatological data from Sidney Levitus and Peter Damm, and refined with observed hydrographic data collected by the Institute of Marine Research. Then, the model was run in diagnostic mode with the fields of salinity and temperature from the Hydrographic Archive held fixed, until a stationary circulation was obtained. The produced dynamically adjusted fields of sea level and current constitute the Diagnostic Archive. To obtain dynamically consistent fields of salinity and temperature together with sea level and current, the model was further run prognostically for seven years, forced by monthly mean wind stress, fresh water runoff, M 2 tide, and boundary values taken from the Diagnostic Archive. The archive produced was denoted the Prognostic Archive. The Hydrographic and Diagnostic Archives provide a satisfactory description of the monthly mean oceanic circulation in the areas of interest. In spite of some weaknesses in the circulation produced by the model, both the Diagnostic and Prognostic Archives should be valuable for regional modelling purposes. The transports through various sections, including the Atlantic inflow through the Shetland-Faeroes channel, are all reasonable, and can be used as boundary conditions for higher resolution ocean models. Finally, this work shows the advantages of combining hydrographie observations and numerical models.