The Gulf of California, a narrow, semienclosed sea, is the only evaporative basin of the Pacific Ocean. As a result of evaporative forcing, salinities in the gulf are 1 to 2 %0 higher than in the adjacent Pacific at the same latitude. This paper examines the thermohaline structure of the gulf and the means by which thermohaline exchange between the Pacific and the gulf occurs, over time scales of months to years. In addition to evaporative forcing, air-sea heat fluxes and momentum fluxes are important to thermohaline circulation in the gulf. From observations presented here, it appears that the gulf gains heat from the atmosphere on an annual average, unlike the Mediterranean and Red seas, which have comparable evaporative forcing. As a result, outflow from the gulf tends to be less dense than inflow from the Pacific. Winds over the gulf change direction with season, blowing northward in summer and southward in winter. This same seasonal pattern appears in near-surface transports averaged across the gulf. The thermohaline circulation, then, consists of outflow mostly between about 50 m and 250 m, inflow mostly between 250 m and 500 m, and a surface layer in which the direction of transport changes with seasonal changes in the large-scale winds. Using hydrographic observations from a section across the central gulf, total transport in or out of the northern gulf is estimated to be 0.9 Sv, heat gain from the atmosphere is estimated to be 20 to 50 W m -2, and evaporation is estimated to be 0.95 m yr -l. These estimates are annual averages, based on cruises from several years. Seasonal variations in thermohaline structure in the gulf are also examined and found to dominate the variance in temperature and density in the top 500 m of the water column. Salinity has little seasonal variability but does exhibit more horizontal variablility than temperature or density. Major year-to-year variations in thermohaline structure may be attributable to E1 Nifio-Southem Oscillation events.

Abstract. An array of shallow pressure gauge pairs is used to determine shallow geostrophic flow relative to an unknown mean velocity in the five principal straits that separate the eastern Indian Ocean from the interior Indonesian seas (Lombok Strait, Sumba Strait, Ombai Strait, Savu/Dao Straits, and Timor Passage). Repeat transects across the straits over several tidal cycles with a 150-kHz acoustic Doppler current profiler were made during three separate years, and provide a first look at the lateral and vertical structure of the upper throughflow in these straits as well as a means of "leveling" the pressure gauge data to determine the mean shallow velocity and provide transport estimates. We estimate a total 2-year average transport for 1996-1997 through Lombok,

Abstract.A

Abstract.Recent IntroductionThe California Current system is traditionally described in terms of three major components: first, the California Current itself, historically held to be a broad, relatively sluggish equa- In the analysis presented here, several aspects of this traditional picture are challenged. First, we find that the synoptic CC just outside the bight (we define the bight as the region east of the Santa Rosa Ridge and including the SB Channel ( Figure 2b)) is neither broad nor sluggish but, rather, tends to be jet-like, as was found during the coastal transition zone (CTZ) experiment off northern California [Brink and Cowles, 1991].Because the CC moves onshore and offshore seasonally, the average picture is a broad, slow feature, but this is not representative of the synoptic current or even of the seasonal average maps. Second, we find that the average geostrophically balanced flow off southern California is divided by the Santa Rosa Ridge, with poleward flow in the bight and equatorward flow offshore, except in spring when an equatorward anomaly weakens or reverses the poleward flow in the bight. Third, we find that wind stress curl over the SC Bight is strong enough and of the right sign to explain the observed poleward flow inshore of the SR Ridge in terms of a Sverdrup balance. In contrast, the observed CC just outside the bight cannot be similarly explained: wind stress curl offshore of the SR Ridge 7695

Five cross‐strait hydrographic sections repeated several times during the Gibraltar Experiment in 1985–1986 are used to examine the structure of the interface layer between the inflowing Atlantic waters and outflowing Mediterranean waters in the Strait of Gibraltar. The interface is 60–100 m thick, with a strong vertical salinity gradient identified by fitting individual salinity profiles to a piecewise‐linear, three‐layer model. The interface is deeper, thicker, fresher, and colder on the west end of the strait than in the Narrows, where there is a minimum in thickness and a maximum in salinity gradient. Farther east, the interface thickens again and continues to get saltier, warmer, and shallower. Property variations in all three layers are also cast in terms of the three principal water types involved in the exchange. The traditional Knudsen model of exchange is extended to three layers, assuming that the interface is a transport‐carrying third layer with uniform vertical shear. As much as half of the inflowing or outflowing transport occurs in the interface layer. Transport converges in both the upper and lower layers, implying, over the length of the strait, vertical exchange between layers that is comparable to about half the horizontal exchange. The richness of structure and complexity of interaction between the interface and the upper and lower layers argues against the use of two‐layer models to characterize exchange through the Strait of Gibraltar.

The salinity and dissolved oxygen of water masses in the Indonesian Seas, from historical hydrographic data, are examined on isopycnal surfaces. We focus primarily on the Banda Sea, from which the bulk of the throughflow transport flows into the Indian Ocean. Dissolved oxygen proves to be a problematic conservative tracer in this region due to biological consumption in the upwelling regime of the Arafura Sea and the subsequent spreading of relatively low oxygen water over a broad area. The remaining analysis is thus restricted to salinity. We first consider a hypothesis of simple isopycnal advection and mixing between North and South Pacific low-latitude western boundary current sources. Three regimes are apparent. The surface and upper thermocline layers, down to 25.8rr0, are too fresh to fit the hypothesis. Vertical mixing of surface precipitation and runoff excesses down into the water column must be invoked. Vertical mixing is also apparent on deeper isopycnals, below 27.0rr0, where a contribution from the deep Indian Ocean can be discerned. In between, in the lower thermocline, the 0-S data are consistent with the hypothesis of purely isopycnal spreading, given a simple variation in the ratio of sources toward increasing South Pacific contribution with depth. In this regime the juxtaposition of sources leads to relatively strong gradients in water mass properties across the Banda Sea. This gradient translates into a difference in outflow characteristics between Timor and Ombai Straits, which appear to draw their waters, respectively, from the eastern and western Banda Sea. We then consider how the presence of vertical mixing modifies our inferred water mass ratio under a variety of boundary conditions. Vertical mixing effects are particularly important in the upper thermocline. In the Banda Sea, on all isopycnals down to rr 0 = 26.5, an increase in the degree of vertical mixing tends to decrease the relative importance of the North Pacific source. The sensitivity of the water mass ratio to degree of vertical mixing decreases with depth, and the tendency is reversed below 26.5rr 0. In the Banda Sea, for •lausible values of throughflow residence time and verticaldiffusivity (h = (Kt) •/ < 100 m), the North Pacific contributes 80-90% of the water mass at 26rr0, 50-60% at 26.5rr0, and only 10-30% at 27rr 0. This large South Pacific contribution in the lower thermocline is supported by other studies.

No abstract

The Gulf of California is a long, narrow marginal sea lying between the Baja California peninsula and the mainland of Mexico. Air‐sea fluxes of heat and moisture in the gulf are enhanced because of geographical isolation from the Pacific provided by the mountainous Baja California peninsula. In the northern gulf, annual evaporation rates are about 1 m y−1. Unlike most evaporative basins, however, the gulf gains heat from the atmosphere at an annual average rate of 20 to 80 W m−2 (Bray, 1988). Given the unusual air‐sea forcing of the gulf, what form or forms should water mass formation take? The annual moisture loss and heat gain require that high‐salinity surface water be transported downward to an intermediate depth and that cold, fresh inflow be transported upward to an intermediate depth. This is accomplished through several mechanisms. (1) Winter convection: this occurs only in a limited geographical region, the Wagner Basin of the far northern gulf, except in El Niño–Southern Oscillation years, when convection appears to be more widespread. (2) Dispersion of convected water in small eddylike features: this occurs within a large‐scale southward transport possibly driven by the large‐scale density gradient associated with atmospheric fluxes. (3) An anticyclonic circulation in the northern gulf: this is found south of the convection region and transports high‐salinity water off the shallow shelves and to substantial depths, where it may mix with water of central gulf origin. (4) Tidal mixing: most of the energy available for mixing in the northern gulf derives from the tides. In particular, tidal mixing over the sill in the island region is responsible for the substantial reduction in salinity of northern gulf waters as they enter the central gulf.