A Numerical Investigation of the Circulation In the Greenland and Norwegian Seas

Abstract

An ocean general circulation model is used to investigate the circulation in the Greenland and Norwegian seas. The primitive equations are solved on 12 levels with a horizontal grid-point distance of ?20 km. This resolution is sufficient to take into account mesoscale topographic structures which, according to observations and the model results presented here, have a large influence on the distribution of hydrographic variables. However, the model is not eddy-resolving; the baroclinic deformation radius in this region is about 5 km only. At first a 2-year experiment is run with the forcing functions and the initial stratification derived from climatological annual-mean data. Compared with quasi-synoptic data, the calculated temperature and salinity distributions show a much better adjustment to the topography than does the climatology. The velocities and transports lie within the range of observed long-term mean values. The model fields obtained by this experiment are used to initialize another experiment which is forced with historical 6-hourly wind-stress data of a 2.5-year period. A direct comparison of simulated and observed transport time series shows reasonable agreement. The calculated fluctuations are analyzed using standard statistical methods. The simulated fluctuations and their correlations with the wind data are interpreted with regard to existing simple analytical models. Despite the complex physics retained in the numerical model and the complicated bottom relief and coastlines, many of the analytical results are reproduced. In particular, shelf-wave?like motions on the shelves and trapping of the fluctuations by closed f/H-contours seem to be important. As a consequence, the predominantly barotropic circulation on the shelves is highly correlated with the wind-stress component parallel to the coast for the frequencies considered. In an intermediate frequency range (?0.2 to 0.01 cpd) the deep along-isobath circulation is coherent with the contour integral of the wind stress along the f/H-contours enclosing the deep basins. There is some evidence from observations to support these model results.