Wind-induced ocean circulation along California and Baja California coasts in June 1999

• Published in: Atmospheric research 2009-09, Vol.94 (1), p.106-133
• Main Authors: Beg Paklar, Gordana, Koračin, Darko, Dorman, Clive
• Format: Article
• Language: English
• Subjects: Atmospheric model MM5; Earth, ocean, space; Exact sciences and technology; External geophysics; Marine; Meteorology; Ocean model POM; U.S. West Coast; Upwelling; Wind stress curl
• ISSN: 0169-8095; 1873-2895
• DOI: 10.1016/j.atmosres.2009.01.019

The circulation along the California Coast in June 1999 was influenced by strong upwelling-favourable winds. To study the ocean circulation and upwelling variability, the Princeton Ocean Model (POM) was setup for the California and Baja California coasts and it was forced with interactively computed wind stress and surface heat fluxes. Atmospheric fields used in the surface flux estimates during the POM runs were obtained from Mesoscale Model 5 (MM5) outputs, while the sea surface temperatures (SST) were prescribed by the ocean model. One-way coupled air–sea modelling was used in the process-oriented approach to asses the influence of the wind stress, wind stress curl, initial density gradients, and nesting procedure with global model for the predicted circulation and accompanied upwelling. Hourly SSTs measured at twelve buoys operated by the National Data Buoy Center (NDBC) were used to evaluate sensitivity modelling studies via root-mean-squared-errors (RMSE). Low-pass filtered currents measured at 15 stations along central and southern Californian coast were also used in the ocean model evaluation procedures. In order to resolve the particular roles and importance of the wind stress vs. wind stress curl in driving coastal upwelling, besides baseline experiment we performed a numerical experiment with curl-free wind forcing. Modelled surface currents obtained in the baseline experiment indicate southward and southwestward flow in the deep ocean areas and alternating equatorward and poleward current near the shore. Wind stress curl appeared to be crucial both for the current field structure and SST distribution. The inshore poleward current obtained in the baseline experiment and evidenced by direct measurements was replaced with an equatorward current in the curl-free experiment. Furthermore, a significant temperature drop occurred along the entire coast in the curl-free experiment, whereas in the baseline the areas of maximum upwelling were confined to the regions with the wind maxima. Climatological temperature profile integrated with the satellite SST appeared to be superior for initialization of the ocean model in comparison with Levitus climatology. Due to too high temperatures predicted by HYbrid Coordinate Ocean Model (HYCOM) along the West Coast, expected improvement of the baseline experiment was not achieved by nesting POM with the Pacific-scale HYCOM. Additionally, a selection of a drag coefficient parameterization appears to have a significan