Detection of Oceanographic Fronts on Variable Water Depths Using Empirical Orthogonal Functions

Abstract

Oceanographic baroclinic fronts are boundaries between water masses with differences in hydrography. Often the presence of baroclinic fronts coincides with oceanographic currents. The detection of fronts in oceanographic data, e.g., ocean model data, is normally carried out by looking for sudden horizontal changes in the sea surface values such as height, temperature, or salinity. Applications involving a single parameter may, therefore, overlook significant oceanic fronts in other parameters. Also some fronts may be more exaggerated below the sea surface and, therefore, missed by approaches that rely on sea surface values only. Here, we suggest a parametrical approach that combines both salinity and temperature profiles from an ocean model data set simultaneously using empirical orthogonal functions. This ensures that both salinity- and temperature-based fronts may be detected even if they are present below the sea surface. The suggested parameterization allows the user to customize the emphasis on salinity or temperature, and even put an emphasis on particular depth regions. Unlike prior methods, this method can handle profiles of different lengths, and thereby detect fronts in both shallow and deep ocean. The method is demonstrated on data from two different numerical ocean models. The same parameterization is used for both data sets. No calibration or tuning was required and the resulting fronts are comparable. The method is applied on a full year of data without changing the parameters of the method. The detected fronts exhibit expected behavior, which builds confidence in the method's robustness.