Carlos J. Lozano

General Circulation of the Eastern Mediterranean

Abstract A novel description of the phenomenology of the Eastern Mediterranean is presented based upon a comprehensive pooled hydrographic data base collected during 1985–1987 and analyzed by cooperating scientists from several institutions and nations (the POEM project). Related dynamical process and modeling studies are also overviewed. The circulation and its variabilities consist of three predominant and interacting scales: basin scale, subbasin scale, and mesoscale. Highly resolved and unbiased maps of the basin wide circulation in the thermocline layer are presented which provide a new depiction of the main thermocline general circulation, composed of subbasin scale gyres interconnected by intense jets and meandering currents. Semipermanent features exist but important subbasin scale variabilities also occur on many time scales. Mesoscale variabilities modulate the subbasin scale and small mesoscale eddies populate the open sea, especially the south-eastern Levantine basin. Clear evidence indicates Levantine Intermediate Water (LIW) to be present over most of the Levantine Basin, implying that formation of LIW is not localized but rather is ubiquitous. The Ionian and Levantine basins are confirmed to form one deep thermohaline cell with deep water of Adriatic origin and to have a turnover time of one and a quarter centuries. Prognostic, inverse, box and data assimilative modeling results are presented based on both climatological and POEM data. The subbasin scale elements of the general circulation are stable and robust to the dynamical adjustment process. These findings bear importantly on a broad range of problems in ocean science and marine technology that depend upon knowledge of the general circulation and water mass structure, including biogeochemical fluxes, regional climate, coastal interactions, pollution and environmental management. Of global ocean scientific significance are the fundamental processes of water mass formations, transformations and dispersion which occur in the basin.

The Atlantic Ionian Stream

Abstract This paper describes some preliminary results of the cooperative effort between SACLANT Undersea Research Centre and Harvard University in the development of a regional descriptive and predictive capability for the Strait of Sicily. The aims of the work have been to: (1) determine and describe the underlying dynamics of the region; and, (2) rapidly assess synoptic oceanographic conditions through measurements and modeling. Based on the 1994–1996 surveys, a picture of some semi-permanent features which occur in the Strait of Sicily is beginning to emerge. Dynamical circulation studies, with assimilated data from the surveys, indicate the presence of an Adventure Bank Vortex (ABV), Maltese Channel Crest (MCC), and Ionian Shelf Break Vortex (IBV). A schematic water mass model has been developed for the region. Results from the Rapid Response 96 real-time numerical modeling experiments are presented and evaluated. A newly developed data assimilation methodology, Error Subspace Statistical Estimation (ESSE) is introduced. The ideal Error Subspace spans and tracks the scales and processes where the dominant, most energetic, errors occur, making this methodology especially useful in real-time adaptive sampling.

Physical and biological modeling in the Gulf Stream Region : I. Data assimilation methodology

Abstract Physical and biological data are assimilated into a time-evolving, mesoscale-resolution three-dimensional (3-D) ocean model using optimal interpolation. Simulations are conducted in the Gulf Stream region during the BIOSYNOP/Anatomy of a Meander Experiment in September–October of 1988. Physical data assimilation only or biological data assimilation only resulted in misalignment of the physical and biological fronts, causing spurious cross-frontal fluxes of biological quantities. Assimilation of both physical and compatible biological fields was necessary for adequate equilibration of the simulated fields. The resulting combined 4-D fields substantially extend the value of the observations alone. A technique is presented for deriving the necessary, dynamically consistent 3-D physical and biological field estimates from data for initialization and assimilation into time-evolving model simulations.

Feature-oriented regional modeling and simulations in the Gulf of Maine and Georges Bank

The multiscale synoptic circulation system in the Gulf of Maine and Georges Bank (GOMGB) region is presented using a feature-oriented approach. Prevalent synoptic circulation structures, or ‘features’, are identified from previous observational studies. These features include the buoyancy-driven Maine Coastal Current, the Georges Bank anticyclonic frontal circulation system, the basin-scale cyclonic gyres (Jordan, Georges and Wilkinson), the deep inflow through the Northeast Channel (NEC), the shallow outflow via the Great South Channel (GSC), and the shelf–slope front (SSF). Their synoptic water–mass (T2S) structures are characterized and parameterized in a generalized formulation to develop temperature–salinity feature models. A synoptic initialization scheme for feature-oriented regional modeling and simulation (FORMS) of the circulation in the coastal-to-deep region of the GOMGB system is then developed. First, the temperature and salinity featuremodel profiles are placed on a regional circulation template and then objectively analyzed with appropriate background climatology in the coastal region. Furthermore, these fields are melded with adjacent deep-ocean regional circulation (Gulf Stream Meander and Ring region) along and across the SSF. These initialization fields are then used for dynamical simulations via the primitive equation model. Simulation results are analyzed to calibrate the multiparameter feature-oriented modeling system. Experimental short-term synoptic simulations are presented for multiple resolutions in different regions with and without atmospheric forcing. The presented ‘generic and portable’ methodology demonstrates the potential of applying similar FORMS in many other regions of the Global Coastal Ocean. r 2003 Elsevier Science Ltd. All rights reserved.

Real-time regional forecasting

Abstract An observational network, dynamical models and data assimilation schemes are the three components of an ocean prediction system. Its configuration for a regional real-time forecasting system proceeds in three phases, based on previous knowledge and experience of the area. In the initial (exploratory) phase, identification of dominant scales (synoptic, mesoscale and submesoscale), processes and interactions is obtained. In the intermediate (dynamical) phase, a clear resolution of the important dynamics and events must be reflected in the nowcasts and forecasts. This is carried out via energy and vorticity analysis (EVA). The third phase is designed to validate the predictive capability of the forecasts. Both qualitative verification and quantitative skill are utilized. At each stage, high quality data sets are required. Observing System Simulation Experiments are essential to the development of the regional ocean prediction system. Initializations and updates are obtained by the fusion of multiple data streams, i.e., the melding of feature models, previous data driven simulations and observations. Nowcasts and forecasts are generated via sequential assimilation combining ship-acquired and sensed remote data. Nested models and nested observations are employed for adequate resolution. The approach is illustrated with recent real-time experiences at sea in the Iceland-Faeroe frontal region, the Straits of Sicily and the Eastern Mediterranean basin.

An interdisciplinary ocean prediction system: Assimilation strategies ana structured data models

Abstract An overview of ongoing research efforts for regional nowcasts, forecasts and hindcasts of physical, biogeochemical and acoustical fields is given. The estimation of oceanic fields is carried out using a modular and flexible system approach, intended to optimize data and model contact, facilitating a unified approach to interdisciplinary research. Recent developments in structured data models are presented, and generic and robust methods to combine structured data models, data streams and model fields, based upon suboptimal interpolation schemes are examined in the framework of data assimilation and dynamical interpolation.

Strait of Sicily water masses

We have derived a water mass model for the Strait of Sicily, based on 1994 and 1995 cruise data. The model consists of seven water masses, suggested by the measured shapes of the vertical temperature and salinity distributions. The core of the Atlantic water is distributed below the surface as a shallow layer, in a depth range of 40 to 100 m, with a salinity minimum. It is capped by upper and surface layers above and a mixed region below. At the bottom, Levantine water is present with a transition region above. Between the mixed and transition region there is, on occasion, a fresher water layer. The structure and statistics of water masses is analyzed over the Strait of Sicily region in terms of their temperature, salinity, and depth values. Objective analysis of the temperature, salinity, and depth parameters is performed in latitude and longitude. The water masses are tracked in terms of their parameter signatures. Changes in temperature and salinity distributions are interpreted. 2-D ellipses that represent the water masses, in terms of means and standard deviation, are derived in a space of temperature, salinity, and depth. Their axes are the standard deviations of parameter space ranges. The areas of the ellipses are compared against the temperature and salinity data distribution. The water mass composition ratios are computed and analyzed. Hypotheses and mechanisms for the origin and mixing of water masses are suggested.

A consistent baroclinic quasigeostrophic ocean model in multiply connected ocean domains

The Harvard ocean baroclinic quasigeostrophic model is further extended to enable the treatment of multiply connected domains with arbitrary coastal boundary geometry. A set of sufficient quasigeostrophic boundary conditions at physical boundaries are determined by requiring consistency with a regular asymptotic expansion in the Rossby number of the primitive equations. To take advantage of fast Helmholtz solvers in regular domains, the physical multiply connected domain is embedded in a regular grid, and boundary conditions are imposed by using a variation of the capacitance matrix method. The accuracy of the method is exhibited by comparison with exact solutions.

An asymptotic theory of combined wave refraction and diffraction

Abstract A uniformly valid asymptotic theory for water waves is presented, which accounts for the combined effects of refraction due to slowly varying water depth and diffraction by a semi-infinite thin break water. The present theory is more rigorous and generical than the approximate solution developed by Liu and Mei, which becomes invalid near the edge and the tip of the breakwater. The effects of wave breaking are ignored.

Dynamical adjustment of quasi-synoptic data sets in the Eastern Mediterranean: testing a full basin coastal methodology

The availability of nearly synoptic basin-wide data sets, has provided the basis for the extension of a regional coastal methodology to full basin. The kinematic boundary conditions for a quasi-geostrophic model in an arbitrary domain, (partially) open or closed and simply- or multiply-connected, are briefly reviewed. A method for the implementation of these conditions with real ocean data is presented for the full Eastern Mediterranean and subregions, and the sensitivities are discussed. Objectively analyzed maps are produced for model initialization. Simulations of two-week duration are carried out in the Levantine and the full Eastern Mediterranean basins to determine the best domain-average amplitude for the barotropic mode and the best treatment of coastal constraint.