Implementing FORMS for the Monterey Bay Forecasting System Using HOPS and ROMS

Abstract

LONG-TERM GOALSOur long-term goal is to implement the feature-oriented regional modeling methodology for the Monterey Bay (MB) forecasting system in retrospective and real-time operations using (i) the Regional Ocean Modeling System (ROMS) and (ii) the Harvard Ocean prediction System (HOPS). OBJECTIVESThree specific objectives are pursued to achieve our long-term goal: (i) to develop a synoptic highresolution regional climatology for Monterey Bay and the California Current System (CCS), (ii) to implement the feature-oriented regional modeling system (FORMS) capabilities in the West Coast ROMS and HOPS modeling efforts for synoptic nowcast, forecast and 3D-VARS and ESSE-based assimilation in the Monterey Bay region, and (iii) to apply this methodology in real time within the nested ROMS and HOPS modeling efforts. APPROACHThe western coast of the U.S. includes both an offshore region and a very dynamic coastal region. The offshore region is primarily dominated by the large-scale California Current, the California Undercurrent and parts of subtropical and sub-polar gyre circulations in the eastern Pacific. The coastal region includes features such as upwelling fronts, cold pools inshore of these fronts, filaments, squirts, mushroom-head vortices, mesoscale and sub-mesoscale eddies, and meanders.In the FORMS approach (Gangopadhyay and Robinson, 2002;Gangopadhyay et al., 2003), such structures and their parameterized forms can be implemented to characterize the relevant circulation entities. Such implementation requires careful and detailed scientific analyses to identify the spatial and temporal scales and variability that define and distinguish these features from one another while preserving their individual characteristics. After the major features are identified and their representations chosen, the latter are used in the initialization of a basic dynamical model (e.g., HOPS or ROMS). Dynamical adjustment accomplishes two important tasks: i) a consistent dynamical interaction of the features, and, ii) the generation of smaller scales, such as squirts and sub-mesoscale eddies. We will carry out a series of numerical experiments to implement the feature models in the CCS within the nowcast-forecast-assimilation framework of both ROMS and HOPS.