Satellite images [ Kelly, 1985; Abbott and
Zion, 1987], drifter trajectories [ Davis, 1985], and
shipboard surveys [ Flament et al., 1985;
Kosro and Huyer, 1986] taken during the 1981--1982 Coastal Ocean
Dynamics Experiment (CODE) focused attention on filaments of cold water
which extend 100s of km offshore along the west coast of the United States
(Figure 1) 
and in other coastal upwelling
regions [e.g., Haynes and Barton, 1991]. While such
filaments had been identified in earlier satellite images of sea surface
temperature, the CODE observations showed that associated with the cold
filaments are strong current jets extending 100 m or more below the
surface, with core speeds in excess of 50 cm/s. The large volume of cold,
upwelled coastal water carried offshore in these filaments was intriguing
and puzzling, since it was equivalent to the expected wind-driven offshore
transport over a 1000 km section of coast [ Kosro and
Huyer, 1986]. Here was an unexpected, and potentially very effective
mechanism for exchange between the shelf and the deep ocean.
These exciting results motivated the subsequent
Coastal Transition
Zone (CTZ) study [ CTZ Group, 1988] which included field
programs in 1987 and 1988 focusing on filaments off northern California.
A wide variety of physical, chemical and biological observations were
acquired during the field programs to understand the cause of the cold
filaments and their impact on the shelf-slope environment. As emphasized
by Brink and Cowles [1991], the successes of the CTZ
study can be attributed in large part to the interdisciplinary approach
which provided different perspectives for addressing the key scientific
questions. A brief description is given below of a few of the results from
this study. For more details the reader is referred to the August 1991
Journal of Geophysical Research special issue on CTZ.
Two related objectives of the CTZ study were determining the source of the water in the cold filaments and what caused the filaments to form. A number of hypotheses had been suggested regarding the formation and maintenance of the filaments [ Strub et al., 1991]. One hypothesis was that filaments were a consequence of convergence in the wind-driven shelf circulation, possibly associated with prominent topographic features such as Point Arena. Another hypothesis was that filaments were due to the offshore, mesoscale (horizontal scales of 10--100 km) eddy field drawing cold shelf water offshore. The former hypothesis identified the shelf as the source and cause of the filaments, while the latter assumed the cold upwelled shelf water only provided a ``dye'' for exposing the offshore eddy field. Strub et al. [1991] provided a convincing argument that neither of these hypotheses was correct. Based on synthesis of the CTZ observations, they argued that the cold filaments are part of a continuous, meandering, southward jet (the California Current) that separates warm offshore water from colder, upwelled coastal water [ Kosro et al., 1991; Brink et al., 1991]. As the jet meanders onshore it entrains cold upwelled water which is carried with the jet as it then meanders offshore. Geochemical, hydrographic, and biological observations indicate that the cold water in the filaments subducts below warmer, lighter water as it flows offshore, resulting for example in deep phytoplankton layers [ Kadko et al., 1991; Washburn et al., 1991]. This important result helped explain the absence of onshore meanders in satellite images and clarified the pathway taken by water in the cold filaments. The details of how and where this meandering current originally entrains cold upwelled water remain unclear. Huyer et al. [1991] found that the temperature and salinity characteristics of the filament water were distinct from water inshore and offshore of the filament. Based on this evidence and similar findings for phytoplankton species [ Mackas et al., 1991; Hood et al., 1991], Strub et al. [1991] argued that the filament water must have an upstream source.
What causes the California Current to meander? Numerical modeling studies suggested that the large-scale dynamics of the meandering jet are complicated [e.g., Walstad et al., 1991; Haidvogel et al., 1991]. While the meandering is consistent with an unstable jet [ Pierce et al., 1991; Allen et al., 1991], the relative roles of coastal topographic features such as capes [ Haidvogel et al., 1991], bottom bathymetry [ Narimousa and Maxworthy, 1989], and wind stress [ McCreary et al., 1991] in generating the instabilities remain unclear. Analysis of an extensive hydrographic data set acquired during the Northern California Coastal Circulation Study (NCCCS) suggested that the structure and variability of the California Current may be influenced by both the coastal topography and the associated shelf and slope circulation [ Bray and Greengrove, 1993].
Though flow convergence on the shelf and offshore eddies do not appear to be the cause of cold filaments, there is growing evidence that these processes can be important in exchange between the shelf and deep ocean. Magnell et al. [1990] gave a graphic example of a persistent shelf flow convergence in the vicinity of Cape Mendocino that appeared to eject water from the shelf, based on a combination of moored current meter, hydrographic, drifter and satellite observations acquired during NCCCS. Both the CTZ and NCCCS observations also provided considerable evidence that mesoscale eddies impinge on the shelf (Figure 1). Largier et al. [1993] used one and a half year long current meter records acquired as part of the NCCCS study to argue that the dominant source of current variability over the Northern California shelf on time scales of weeks to months is not the wind stress but mesoscale oceanic eddies impinging on the shelf. This key result provided a new perspective on the shelf circulation of the west coast of the U. S. which had predominantly been viewed in the context of wind-driven circulation on time scales of days to weeks. The question of how much of the exchange between the deep ocean and the shelf is caused by mesoscales eddies remains unresolved. Do these eddies move passively up onto the shelf with little interaction with the surrounding water, or do they mix with the surrounding shelf water and deposit offshore water on the shelf or entrain shelf water and carry it offshore? Schumacher et al. [1993] monitored an anticyclonic eddy over the shelf in Shelikof Strait, Alaska and found very little exchange between the eddy and the surrounding water. Washburn et al. [1993] tracked an anticyclonic eddy that moved from the deep ocean onto the shelf near Point Arena. They showed that this eddy entrained sediment from the shelf and carried it offshore, providing an interesting mechanism for the export of sediment from the shelf (Figure 1).