Low-salinity Tongue Research Articles

A Modified U-Net Model for Predicting the Sea Surface Salinity over the Western Pacific Ocean

The prediction of oceanic features is always an important issue in oceanography, where deep learning has been proven to be a useful tool. In this study, we applied the improved U-net model to predict the monthly sea surface salinity (SSS) over the western Pacific (WP) Ocean, and the model was designed to use the SSSs from six consecutive months to predict the SSS in the next month. The monthly satellite-based SSSs in 2015–2020 were used for model training, and the data collected after January 2021 were used to evaluate the model’s predictive abilities. The results showed that the predicted SSSs represented the general patterns of SSSs over the WP region. However, the small-scale features were smoothed out in the model, and the temporal variations were also not well captured, especially over the East China Sea and Yellow Sea (ECS&YS) region. To further evaluate the potential of the U-net model, a more specific model was conducted for the ECS&YS region (Domain 2), which successfully predicted both spatial and temporal variations in the SSSs, including the spreading and retreating of the low-salinity tongue. Based on the comparison between the two domains and sensitivity experiments, we found that the prediction biases were contributed by the spatial distributions of the SSSs, the domain size, and the filter numbers. In addition, further multi-step prediction experiments suggested that our U-net model could also be used for long-time prediction, and we have examined up to five months. Overall, this study demonstrated the great ability and potential of the U-net model for predicting SSS, even though only a few trainable data are available.

Retrieval and Analysis of Sea Surface Salinity in the Adjacent Waters of the Yangtze River Estuary Based on Multisource Satellite Data

Cao, K.; Sun, W.; Chen, L.; Meng, J., and Zhang, J., 2020. Retrieval and analysis of sea surface salinity in the adjacent waters of the Yangtze River Estuary based on multisource satellite data. Journal of Coastal Research, 36(3), 590–599. Coconut Creek (Florida), ISSN 0749-0208.Moderate Resolution Imaging Spectroradiometer, Sea-Viewing Wide Field-of-View Sensor, and Visible Infrared Imaging Radiometer Suite multisource satellite remote sensing reflectance data are used with in situ sea surface salinity (SSS) measurements in the Yellow Sea and East China Sea. A linear algorithm is developed for the retrieval of summer SSS in adjacent waters of the Yangtze Estuary in the Yellow Sea and East China Sea. The accuracy of SSS retrieval results is validated using independent in situ measurements. Results show that the mean bias of the model is –0.34 psu, the root-mean-square error is 1.55 psu, and the correlation coefficient is 0.80. The linear model is extended to the retrieval of multisource satellite monthly remote sensing reflectance data to obtain monthly SSS products of the Yangtze Estuary. The spatial coverage of fused satellite SSS data is about 10% greater than that of single-satellite monthly retrieved SSS data. The SSS in the Yangtze Estuary research area has lower values nearshore and higher values offshore. The SSS gradually rises from nearshore areas to open water, and an obvious low-salinity water tongue forms in the Yangtze Estuary. The eastward expansion of freshwater flows of the Yangtze River into the sea is mainly affected by the runoff of the Yangtze River, while a southerly wind and northeastward flow in the research area are the main factors responsible for the eastward expansion of the Yangtze River freshwater flows to the NE. The effect of freshwater on the Yellow Sea and East China Sea extends to 125.5° E, and the low-salinity freshwater tongue distributes northeastward. The expansion types of the Yangtze River freshwater include NE expansion, N-NE expansion, E-NE expansion, bidirectional expansion, and nondirectional expansion.

Dynamics of an Equatorial River Plume: Theory and Numerical Experiments Applied to the Congo Plume Case

Abstract The Congo River has the second largest rate of flow in the world and is mainly responsible for the broad tongue of low-salinity water that is observed in the Gulf of Guinea. Despite their importance, near-equatorial river plumes have not been studied as thoroughly as midlatitude plumes and their dynamics remain unclear. Using both theory and idealized numerical experiments that reproduce the major characteristics of the region, the authors have investigated the dynamics of the Congo River plume and examine its sensitivity to different forcing mechanisms. It is found that near-equatorial plumes are more likely to be surface trapped than midlatitude plumes, and the importance of the β effect in describing the strong offshore extent of the low-salinity tongue during most of the year is demonstrated. It is shown that the buoyant plume constrained by the geomorphology is subject to the β pulling of nonlinear structures and wavelike equatorial dynamics. The wind is found to strengthen the intrinsic buoyancy-driven dynamics and impede the development of the coastal southward current, in coherence with observations.

Change in the intensity of low-salinity water inflow from the Bay of Bengal into the Eastern Arabian Sea from the Last Glacial Maximum to the Holocene: Implications for monsoon variations

A 100–400km wide region of the coastal Eastern Arabian Sea (EAS), off the west-coast of India, is characterized by a low-salinity tongue formed by the inflow of low-salinity surface water from the Bay of Bengal (BoB). This low-salinity tongue is largely driven by the sea level higher in BoB than in the Arabian Sea and by alongshore pressure gradient between southern- and northern-EAS, and is expected to respond to summer monsoon freshwater flux to the bay. Here, we report past variation in the relative intensity of summer- and winter-monsoons based on changes in the north–south salinity gradient within this low-salinity tongue. The salinity gradient is estimated from paired measurement of δ18O and Mg/Ca in Globigerinoides sacculifer extracted from sediment cores collected at northern high-salinity and southern low-salinity locations within this tongue.The Last Glacial Maximum (LGM) to peak-Holocene δ18O and sea surface temperature gradients at both locations are ~−2‰ and +2°C respectively, while the sea surface salinity gradient at northern-EAS is 0.5psu higher than in the southern-EAS, suggesting distinctly different SSS structure in the LGM-EAS. The north–south surface salinity gradient was also larger by ~0.5psu during the LGM (1.2psu) as compared to the gradient during the Holocene (0.7psu). Increased north–south surface salinity gradient during the LGM suggests diminished flow of low-salinity water into the coastal EAS caused by combined effect of decreased freshening of the BoB and reduced seasonal mountain-river discharge into the EAS. Such surface hydrographic conditions in the coastal EAS clearly indicate significantly weakened summer monsoons and strengthened winter monsoons during the LGM.

Injection of Antarctic Intermediate Water into the Atlantic subtropical gyre in an eddy resolving ocean model

Antarctic Intermediate Water (AAIW) injection mechanism into the Atlantic subtropical gyre at the Brazil‐Malvinas (B/M) Current confluence is investigated by using an eddy resolving ocean model. The temporal residual‐mean (TRM) velocity framework is used to evaluate transport due to eddies. It is found that the AAIW injection is caused by eddy‐induced transport. A low potential vorticity (PV) water in Malvinas Current generated in the southeast Pacific and near Burdwood Bank encounters the Brazil Current at the B/M confluence, which has relatively high PV value. Since stirring of eddies transports low PV water into high PV water region, the low PV water in Malvinas Current is injected into Brazil Current. The downward TRM velocity exceeds 3 × 10−5 m s−1 at the confluence, which represents strength of the injection. This strong downward transport forms the low salinity tongue of the Atlantic AAIW in this eddy‐resolving ocean model.

The effects of river discharge and seasonal winds on the shelf off southeastern South America

The Río de la Plata waters form a low salinity tongue that affects the circulation, stratification and the distributions of nutrients and biological species over a wide extent of the adjacent continental shelf. The plume of coastal waters presents a seasonal meridional displacement reaching lower latitudes (28°S) during austral winter and 32°S during summer. Historical data suggests that the wind causes the alongshore shift, with southwesterly (SW) winds forcing the plume to lower latitudes in winter while summer dominant northeasterly (NE) winds force its southward retreat. To establish the connection between wind and outflow variations on the distribution of the coastal waters, we conducted two quasi-synoptic surveys in the region of Plata influence on the continental shelf and slope of southeastern South America, between Mar del Plata, Argentina and the northern coast of Santa Catarina, Brazil. We observed that: (A) SW winds dominating in winter force the northward spreading of the plume to low latitudes even during low river discharge periods; (B) NE winds displace the plume southward and spread the low salinity waters offshore over the entire width of the continental shelf east of the Plata estuary. The southward retreat of the plume in summer leads to a volume decrease of low salinity waters over the shelf. This volume is compensated by an increase of Tropical waters, which dominate the northern shelf. The subsurface transition between Subantarctic and Subtropical Shelf Waters, the Subtropical Shelf Front, and the subsurface water mass distribution, however, present minor seasonal variations. Along shore winds also influence the dynamics and water mass variations along the continental shelf area. In areas under the influence of river discharge, Subtropical Shelf Waters are kept away from the coastal region. When low salinity waters retreat southward, NE winds induce a coastal upwelling system near Santa Marta Cape. In summer, solar radiation promotes the establishment of a strong thermocline that increases buoyancy and further enhances the offshore displacement of low salinity waters under the action of NE winds.

Recent changes of the thermohaline circulation in the subpolar North Atlantic

Time series of hydrographic sections in the northern North Atlantic from the period 1990 to 2004 are analyzed for changes in the characteristics and distribution of water masses that are involved in the thermohaline circulation (THC). During the 1990s, the North Atlantic Oscillation (NAO) alternates from a positive phase (strong westerlies) to a negative phase (weak westerlies). The reduced ocean heat loss confined the convection in the Labrador Sea to the upper 1,200 m, generating a new salinity minimum layer characterizing the Upper Labrador Sea Water (ULSW), and led to a warming and salinization of the older LSW below due to lateral mixing. The Lower LSW, formed in the first half of the 1990s, spread in the subpolar gyre and reached the Newfoundland and Irminger basins after about 1 to 2 years, where the associated isopycnal doming contributed to eastward frontal shifts in the upper layer. After 5 and 6 years, it arrived in the Iceland and West European basins, respectively. The collapse of the isopycnal dome in the Labrador Sea, associated with the drainage of the Lower LSW, resulted in a slowing of the cyclonic circulation of the subpolar gyre. This was accompanied in the upper layer by a westward shift of the southeastern extension of the gyre and a northward advection of warm and saline subtropical water in its eastern part, which finally reached the Labrador Sea after about 7 years. In the upper layer of the Labrador Sea, the advection of warm and saline water dominated over the heat loss to the atmosphere and the freshwater gain from melting ice and precipitation in the NAO-low period, so that no accumulation of freshwater but an increase of the heat and salt contents were observed, as in the whole eastern part of the subpolar gyre. Within 1 to 2 years after the drop of the NAO in the winter of 1995/1996, the Subarctic (Subpolar) Front shifted northward and westward north of about 50°N, favored by the retreat of the low-salinity tongue extending eastward from the southern Labrador Sea, and it shifted southward and eastward in the Newfoundland Basin. Therefore, the enhanced northward advection of subtropical waters in the northeastern North Atlantic is balanced by the enhanced southward advection of subarctic waters, including Lower LSW in the Newfoundland Basin, indicating a strong response of the gyre component of the THC.

Impact of Satellite-Derived Forcings on Numerical Ocean Model Simulations and Study of Sea Surface Salinity Variations in the Indian Ocean

AbstractThis study focuses on two major aspects: the impact of satellite forcings (winds and precipitation) on the simulations of a multilayer Indian Ocean (IO) model (IOM) and the analysis of the processes responsible for salinity variations in the Indian Ocean during dipole years (1994 and 1997). It is observed that the European Remote Sensing Satellite-2 (ERS-2) scatterometer wind-driven solutions describe the interannual variabilities of sea surface temperature (SST) more realistically than the National Centers for Environmental Prediction (NCEP) wind-driven solutions. The equatorial westward current jet [hereafter referred to as reverse Wyrtki jet (RWJ)] originating near the Sumatra coast in response to anomalous easterlies during fall of 1994 and 1997 is quite strong in the scatterometer-forced solutions. This RWJ is found to be weak in the NCEP solution. Two more experiments differing by their precipitation forcings [climatological and interannually varying Global Precipitation Climatology Project (GPCP) rainfall] are carried out. Model-simulated variables like SST, sea surface salinity (SSS), and mixed layer depth (MLD) have been compared with in situ observations to verify the performance of the model. The model suggests a dipolelike structure in surface salinity during late 1994 and 1997, with low salinity in the central equatorial Indian Ocean (EIO) and high salinity near the Sumatra coast. The low-salinity tongue is caused by the transport of fresh surface waters via RWJ, which is further strengthened by a southward branch (which is absent in normal years) coming from the Bay of Bengal. A major inference of the study is that the low-salinity tongue is caused mainly by advection, not by a direct effect of precipitation. On the contrary, the high salinity near the Sumatra coast is due to the strong upwelling caused by anomalous easterlies. Another inference made out of this study is that there is apparently a definite signature of the evolution of the dipole event in the MLD approximately 2 months prior to the peak occurring in SSS in the south EIO.

Diatoms in Southeast Pacific surface sediments reflect environmental properties

Diatom abundance and assemblage composition determined for 47 surface sediment samples from the Southeast Pacific (50°S–15°N), combined with existing data for the Peru and Chile margins, demonstrate responses to regional temperature, upwelling, and productivity. High diatom abundances (# valves/g) mark the eastern equatorial Pacific upwelling and the coastal upwelling areas, in particular the upwelling centers off Peru. Freshwater diatoms reflect the low-salinity tongue off the Chilean fjords. Diatom species composition distinguishes between coastal and eastern equatorial Pacific upwelling conditions, and records sea-surface temperatures. Q-mode factor analysis defines five floral assemblages. Factors 1 and 4 determined by the genus Chaetoceros (F1) and Thalassionema (F4) reflect coastal and equatorial upwelling conditions, respectively. Factors 2 and 3 characterized by the genus Thalassiosira and Azpetia nodulifera can be associated with El Niño conditions. A 5th factor, described by Paralia sulcata, records a near-shore upwelling center off Point Concepción, central Chile. Statistical transfer functions relate diatom species percentages to sea-surface temperature and productivity with error estimates of ±0.9 °C and ±23 gC/m 2 yr, respectively, and provide new tools for estimating past temperature and productivity along the west coast of South America.

Pathways and variability of the Antarctic Intermediate Water in the western equatorial Pacific Ocean

In the western equatorial Pacific the low-salinity core of Antarctic Intermediate Water (AAIW) is found at about 800 m depth between potential density levels σ θ = 27.2 and 27.3. The pathways of AAIW and the degradation of its core are studied, from the Bismarck Sea to the Caroline Basins and into the zonal equatorial current system. Both historical and new observational data, and results from numerical circulation model runs are used. The observations include hydrographic stations from German and Japanese research vessels, and Eulerian and Lagrangian current measurements. The model is the JAMSTEC high-resolution numerical model based on the Modular Ocean Model (MOM 2). The general agreement between results from the observations and from the model enables us to diagnose properties and to provide new information on the AAIW. The analysis confirms the paramount influence of topography on the spreading of the AAIW tongue north of New Guinea. Two cores of AAIW are found in the eastern Bismarck Sea. One core originates from Vitiaz Strait and one from St. George’s Channel, probably arriving on a cyclonic pathway. They merge in the western Bismarck Sea without much change in their total salt content, and the uniform core then increases considerably in salt content when subjected to mixing in the Caroline Basins. Hydrographic and moored current observations as well as model results show a distinct annual signal in salinity and velocity in the AAIW core off New Guinea. It appears to be related to the monsoonal change that is typically found in the near-surface waters in the region. Lagrangian data are used to investigate the structure of the deep New Guinea Coastal Undercurrent, the related cross-equatorial flow and eddy-structure, and the embedment in the zonal equatorial current system. Results from 17 neutrally buoyant RAFOS floats, ballasted to drift in the AAIW core layer, are compared with a numerical tracking experiment. In the model 73 particles are released at five-day intervals from Station J (2.5°N, 142°E), simulating currents at a moored time series station north of New Guinea. Observed and model track patterns are fairly consistent in space and season. Floats cross the equator preferably north of Cenderawasih Bay, with a maximum range in eddy-motion in this region north of New Guinea. The northward route at 135°E is also reflected in a low-salinity tongue reaching up to 3°N. At that longitude the floats seem to ignore the zonally aligned equatorial undercurrents. Farther to the east (139–145°E), however, the float observations are consistent with low-latitude bands of intermediate currents.

Interannual and Decadal Variabilities of NPIW Salinity Minimum Core Observed along JMA’s Hydrographic Repeat Sections

We investigated the variation of the North Pacific Intermediate Water (NPIW) distribution in the western North Pacific, focusing on the intermediate salinity minimum (S < 34.2) core observed along the meridional hydrographic sections including the 137°E repeat section by the Japan Meteorological Agency. This core is a cross-section of a low salinity tongue extending westward along the recirculation in the subtropical gyre. The core size shows remarkable variabilities in interannual and decadal time scales. The salinity change in the density layer during the period of core expansion (shrinking) represents the spatial salinity change in the tongue toward the west (east). Thus, we conclude that the core size variation is associated with the zonal wobble of the tongue having thicker distribution to the east, rather than temporal changes of the water mass itself. The core size at 137°E is well correlated with the meridional gradient of the depth in the isopycnal surface at the salinity minimum representing the recirculation intensity, suggesting a relation with the intensity of the subtropical gyre. A significant lag-correlation between the gradient and the wind forcing over the North Pacific suggests that the first mode baroclinic Rossby waves excited in the central North Pacific propagated westward to change the intensity of the recirculation in interannual time scales. In decadal time scales, it is found that the wind stress curl and heat flux fields in the North Pacific precede the recirculation by about 11 years.

Past 100 ky surface salinity-gradient response in the Eastern Arabian Sea to the summer monsoon variation recorded by δ 18O of G. sacculifer

Northward flowing coastal currents along the western margin of India during winter–spring advect low-salinity Bay of Bengal water in to the Eastern Arabian Sea producing a distinct low-salinity tongue, the strength of which is largely governed by the freshwater flux to the bay during summer monsoons. Utilizing the sedimentary records of δ 18O G. sacculifer , we reconstructed the past salinity-gradient within that low-salinity tongue, which serves as a proxy for the variation in freshwater flux to the Bay of Bengal and hence summer monsoon intensity. The north–south contrast in the sea level corrected (residual)-δ 18O G. sacculifer can be interpreted as a measure of surface salinity-contrast between those two locations because the modern sea surface temperature and its past variation in the study region is nearly uniform. The core-top residual-δ 18O G. sacculifer contrast of 0.45‰ between the two cores is assumed to reflect the modern surface salinity difference of 1 psu and serves as a calibration for past variations. The residual-δ 18O G. sacculifer contrast varies between ∼0.2‰ at ∼75 ky B.P. (i.e., late-Marine Isotope Stage 5) and ∼0.7‰ at ∼20 ky B.P. (i.e., Last Glacial Maximum), suggesting that the overall salinity difference between the northern- and southern-end of the low-salinity tongue has varied between ∼0.6 and ∼1.6 psu. Considerably reduced difference during the former period than the modern suggests substantially intensified and northward-extended low-salinity tongue due to intense summer monsoons than today. On the other hand, larger difference (∼1.6 psu) during the latter period indicates that the low-salinity tongue was significantly weakened or withdrawn due to weaker summer monsoons. Thus, the salinity-gradient in the eastern Arabian Sea low-salinity tongue can be used to understand the past variations in the Indian summer monsoons.

SAFARI cruise: Mapping river discharge effects on Southern Brazilian shelf

We report results of a hydrographic cruise organized within the framework of the Russian‐Brazilian joint project SAFARI (Southwestern Atlantic Freshwater Assimilation and River Inputs) and aimed at mapping special features of salinity and temperature distributions connected with the freshwater discharges from the Plata river and Patos‐Mirim lagoon estuaries on the Southern Brazilian shelf. Apart from the low salinity tongue in the inner shelf that has not been mapped in detail before, the intense river runoff results in strong temperature inversions (up to 4°C per 20–30 m) seen in the upper layer in winter. We argue that these inversions cannot result from the advective mechanism alone and should also be attributed to the freshwater related increase of vertical stability that leads to reduced mixing and thus enhanced winter cooling of the sea surface. Spatial distributions of salinity, inversion layer thickness and inversion magnitude over the shelf and continental slope are presented. In a belt extending for hundreds kilometers northward from the Plata estuary in the inner part of the shelf, salinity is up to 10 psu lower than that in the adjacent ocean beyond the river plumes. The spatial extent of the Patos‐Mirim lagoon plume is determined and the fine structure of the salinity distribution near the lagoon mouth is described.

History of the South Java Current over the past 80 ka

A sediment core located below the present South Java Current (SJC) was used to reconstruct paleoclimate and paleoceanography on the basis of biogenic and terrigenic proxy-records. The core spans the past 80 ka of environmental change and shows considerable contrasts from the glacial to the Holocene. Presently, the core site is situated beneath a seasonally varying low-salinity tongue which is advected from the Java Sea via the Sunda Strait. It carries terrigenous matter of a characteristic signature. During the last glacial period (stage 4-2), when sea level was lower than during the Holocene, the Sunda Strait was closed and the terrigenous supply from that source ceased. As the core site is close to the equator, our results indicate that atmospheric and oceanographic circulation was alternatively dominated by the Northern Hemisphere East Asian Monsoon system and the Southern Hemisphere Australian Monsoon system. Between 20 and 12 ka, the (Australian) SE Winter Monsoon reached its maximum and intensified the westward flowing SJC. Increased mixing of the surface waters led to a slight rise in paleoproductivity. A similar but much weaker situation prevailed from 74 to 70 ka. During most of the glacial period, from 70 to 20 ka, strong northeasterly winds associated with the East Asian Winter Monsoon intensified the Indian Monsoon Current and the eastward flowing SJC, and may have also carried dust across the equator to our core site. During this glacial phase, populations of the giant diatom Ethmodiscus rex were thriving and may indicate a reduced deep and intermediate thermohaline circulation at the site. The monsoonal system as we know it today, with distinct dry and wet seasons, may not have been active before ∼12 ka.

Observations of the Western Boundary Current of the Bay of Bengal from a hydrographic survey during March 1993

The characteristics of the Western Boundary Current of the Bay of Bengal (BBWBC) are studied utilising MICOM STD profiles collected during 9–23 March 1993 aboard F.O.R.V. Sagar Sampada from an area between 11°N and 18°N and from the coastline to ∼400 km offshore. North of 15°N, the waters nearer the coast were warmer and more saline, with a cross-shore gradient associated with an alongshore current regime. A cold, low-salinity tongue was found offshore north of 15°N. The isotherms in the vertical sections from 11°N to 18°N showed an offshore deepening, indicating a northerly flow. The nearshore isotherms and isohalines along the 13° and 14°N sections, however, indicated the presence of a cyclonic eddy. The geostrophic flow regime was dominated by the presence of eddies and poleward flows. The poleward flow, with a cross-stream of dimension 30 km, was stronger (> 90 cm/s), and extended deeper (500 m) in the northwestern Bay. This flow was regarded as the BBWBC during March 1993. South of 13°N, the poleward flow appeared as part of an anticylconic eddy with speeds up to about 40 cm/s. The net volume transports were directed northward across all the sections (except the 13°N section) with magnitudes varying between 3 and 17 Sv (1 Sv =10 6 m 3/s). Southward transport of 3 Sv was observed nearer the coast along 14°N due to the cyclonic eddy. The hydrographic features along a section from off Madras to Andamans during May 1987 exhibited the characteristics of a well-developed WBC of a subtropical gyre. The volume transport in the BBWBC region was found to be considerably higher in May 1987 (16 Sv) than in March 1993 (5 Sv). It is concluded that the flow regime inferred in the western Bay during March 1993 represented the developing phase of the western boundary current system, which appeared first at the northwestern Bay.

The Mechanism for Antarctic Intermediate Water Renewal in a World Ocean Model

Abstract Realistic representation of the low-salinity tongue of Antarctic Intermediate Water (AAIW) has been achieved in a coarse-resolution ocean general circulation model. The authors find that this water mass is not generated by direct subduction of surface water near the polar front. Instead, the renewal process is concentrated in the southeast Pacific Ocean off southern Chile. The outflow of the East Australian Current progressively cools (by heat loss to the atmosphere) and freshens (by assimilation of polar water, carried north by the surface Ekman drift) during its slow movement across the South Pacific toward the AAIW formation zone. Further, deep, warm advection near Chile enables more convective overturn, resulting in very deep mixed layers from which AAIW is fed into the South Pacific and also into the Malvinas Current. Along with this isolated region of AAIW renewal, the model relies on alongisopycnal mixing of fresh surface water from the polar front to capture a realistic circumpolar tongue...

Yangtze Brackish water plume—circulation and diffusion

A system of 3-D linearized momentum equations, the continuity equation and a simplified equation of state for sea water are analytically solved. The solution comprises three parts. The first part is a wind-driven current, the second a jet-like current caused by Yangtze River outflow and the last a density flow. The computed current velocity is obtained by solving an advection-diffusion equation of salinity by a finite-difference method. The results show that the Yangtze brackish water plume spreads as a large low salinity tongue. At the surface, the turning of the tongue axis is mainly controlled by wind stress, Yangtze run-off and bottom topography. Near the bottom, the axis changes little all the year around. At about 10–15 m depth, there is a sharp halocline in summer, but at deeper layers the vertical distribution is almost always uniform. The computed salinity distribution agrees fairly well with the observed one.

Phytoplankton community composition across the West Wind Drift off South America

The biogeography of phytoplankton sampled along 92°W and between 45 and 36°S in conjunction with a detailed XBT section corroborates the physical evidence of a counterflow to the West Wind Drift. Distinct communities exist to either side of an east-west tending front that separates the low-salinity surface water tongue off Chile from subantarctic waters to the south. A plankton group dominated by Ceratium trichoceros and other dinoflagellates exists south of the front in the subantarctic waters, while a distinct group dominated by Rhizosolenia bergonii and other diatoms occupies the low-salinity tongue; the northern limit of the lower salinity surface water is the Subtropical Convergence where a third, distinct plankton community is also dominated by dinoflagellates, mainly C. furca and C. gibberum. Factor analyses support the existence of the distinct groups; it also provides some evidence for the origin of the low-salinity surface water as a mixture of subantarctic surface and coastal runoff by identifying a unique factor axis for Nitzshcia longissima, a pennate diatom generally associated with shallow coastal waters and which comprises a major fraction of the phytoplankton assemblage near the frontal zone but within the low-salinity tongue.

Equatorial undercurrent and deeper flows in the central Pacific

In April, 1971, absolute current velocity was measured down to 1000 m at 1°N, 0° and 1°S, along 150°W. Two velocity time series were obtained: a 25-day record from the Equatorial Undercurrent (102 m) at 1°N and a 152-day record from the deep water (3100 m) at 1°S. The profiles show eastward flow at the surface, a strong Equatorial Undercurrent, a secondary maximum of eastward flow at the bottom of the thermostad, and a westward flow below 500 m. The surface current was counter to the trade winds. Although there was a high-salinity core south and a low-salinity tongue north of the equator, usually interpreted as indirect evidence of equatorward flow at the level of the Undercurrent, the flow at 1°N (at 102 m) was poleward for all 25 days. There was an accelerating eastward and northward current at 1°N; the data are consistent with a northward translation of the Undercurrent at a speed of 7 cm s −1. The winds at Christmas Island (2°N, 157°W) and the Undercurrent velocity at 1°N, 150°W, showed evidence of a 3- to 4-day period; it appears that at this time the currents in the thermocline may have been responding directly to the winds. The meter moored in deep water at 1°S showed a relatively steady eastward velocity component with a mean of 4 cm s −1. Below the Undercurrent in the central Pacific at least two reversals are indicated in the east-west current profile.