Hydrology and Salt Balance in a Large, Hypersaline Coastal Lagoon: Lagoa de Araruama, Brazil

Lagoa de Araruama in the state of Rio de Janeiro, Brazil, is a hypersaline coastal lagoon as a result of semi-arid climate conditions, a small drainage basin and a choked entrance channel. The lagoon has been continuously hypersaline for at least 4·5 centuries, but the mean salinity has varied substantially. It has recently decreased from 57 to 52 as indicated by density (salinity) measurements between 1965 and 1990.Analysis of more than 20 years of salinity time series data, in addition to monthly lagoon cruises to measure the spatial salinity distribution, indicate that the lagoon salinity largely fluctuates in response to the difference between evaporation and precipitation. The major factor explaining the long-term trend of decreasing salinity in the lagoon is the constant pumping of 1 m³s⁻¹of freshwater to the communities surrounding the lagoon from an adjacent watershed, and subsequent discharge of this water into Lagoa de Araruama. The net salt budget is primarily a balance between the advective import of salt from the coastal ocean and eddy diffusive export of salt to the ocean, although the extensive mining of salt from the lagoon during past decades is also a small but significant contribution to the salt budget. The flushing half-life is proposed as a useful time scale of water exchange, is calculated based on a combination of hydrological and tidal processes, and is excellent for comparison of lagoons and assessing water quality changes. The flushing half-life measures 83·5 days for Lagoa de Araruama, considerably longer than for most other coastal lagoons. The proposed dredging of a second ocean channel to Lagoa de Araruama is probably not a good idea. It is likely to accelerate the decrease of lagoon salinity and somewhat improve the lagoon water exchange. At the same time, this will eliminate the apparent buffering capacity provided by the hypersaline environment, and thus may potentially cause water quality problems.     

Modelling and observations of tidal wave propagation, circulation and residence times in Puttalam Lagoon, Sri Lanka

Tidal measurements and a depth-averaged 2D model are used to examine wave progression and circulation in a long, shallow, micro-tidal lagoon in Sri Lanka. Ranges and phase lags for different tidal constituents are used to calibrate the model. A single drag coefficient, C_d = 0.0032, gives almost perfect agreement with data. Current measurements are used for validation of the model. The lagoon tide consists of a combination of progressive and standing waves, where progressive waves dominate in the outer part and standing waves in the inner. A Lagrangian based particle-tracking method is developed to study tidally and wind induced residence times. If tides were the only factor affecting the residual circulation, the residence time inside the narrowest section would be approximately 100 days. Steady winds (of typical monsoon average) decrease the residence times to 60–90 days. Estuarine forcing due to net freshwater supply is not modelled (due to lack of reliable runoff data), but independent, long-term salinity observations and calculations based on volume and salt conservation during periods of negligible freshwater supply (the lagoon is seasonally hypersaline) indicate residence times ranging from 40 to 80 days. Model derived residence times based on tides alone represent a minimum exchange. Even weak forcing, through winds, excess evaporation or freshwater supply efficiently reduces residence times.     

The effect of submarine groundwater discharge on nutrient and salinity regimes in a coastal lagoon off Perth, Western Australia

Submarine groundwater discharge (SGD) into a coastal lagoon off Perth, Western Australia, contains nitrate and silicate in concentrations two orders of magnitude higher than those of the receiving waters. This discharge delivers enough nitrate to replace that dissolved in the lagoon water mass about every eight days and enough silicate to replace the lagoon silicate in about 48 days. The delivery rate of nitrate nitrogen by SGD is equal to about 48% of that required for observed growth rates of lagoon macrophytes. Surface salinity is lower close to the shore as a result of SGD. During calm conditions a salinity front was observed in the lagoon, with a nearshore pool of nutrient-enriched water floating above the more saline ocean water.     

Methane sources,sinks and fluxes in a temperate tidal Lagoon: The Arcachon lagoon (SW France)

The Arcachon lagoon is a 156 km² temperate mesotidal lagoon dominated by tidal flats (66% of the surface area). The methane (CH₄) sources, sinks and fluxes were estimated from water and pore water concentrations, from chamber flux measurements at the sediment–air (low tide), sediment–water and water–air (high tide) interfaces, and from potential oxidation and production rate measurements in sediments. CH₄ concentrations in waters were maximal (500–1000 nmol l⁻¹) in river waters and in tidal creeks at low tide, and minimal in the lagoon at high tide (<50 nmol l⁻¹). The major CH₄ sources are continental waters and the tidal pumping of sediment pore waters at low tide. Methanogenesis occurred in the tidal flat sediments, in which pore water concentrations were relatively high (2.5–8.0 μmol l⁻¹). Nevertheless, the sediment was a minor CH₄ source for the water column and the atmosphere because of a high degree of anaerobic and aerobic CH₄ oxidation in sediments. Atmospheric CH₄ fluxes at high and low tide were low compared to freshwater wetlands. Temperate tidal lagoons appear to be very minor contributor of CH₄ to global atmosphere and to open ocean.     

Geochemical and hydrodynamic constraints on the distribution of trace metal concentrations in the lagoon of Nouméa, New Caledonia

Seawater samples were collected in the lagoon of Nouméa (southwest New Caledonia) along two transects from eutrophic coastal bays to the oligotrophic barrier reef. Land-based emissions to the lagoon were measured with dissolved and particulate concentrations of chromium (Cr) and nickel (Ni), used as tracers of both terrigenous and industrial (Ni ore treatment) activities, as well as dissolved and particulate concentrations of zinc (Zn), used as a tracer of urban effluents. The spatial variability of metal concentrations was related to geochemical and hydrodynamic conditions, i.e., respectively: (1) natural and anthropogenic emission sources, and chemical processes occurring in the water column; and (2) water residence times. The parameter used to describe the residence time of water masses was the local e-flushing time, i.e. the time required for a tracer mass contained within a given station to be reduced by a factor 1/e. High metal concentrations were found in coastal areas (up to 9000 ng dissolved Ni L⁻¹), and steeply decreased with distance from the coast (down to 101 ng dissolved Ni L⁻¹ near the barrier reef) to reach levels similar to those found in remote Pacific waters, suggesting a rapid renewal of waters close to the barrier. Distributions of metals in the lagoon are controlled upstream by land-based emission sources and later chemical processes. Then hydrodynamics constrain metal distributions, as shown by the observed relationship between local e-flushing times and the spatial variability of metal concentrations. In addition, a change in the direction of prevailing winds yielded a decrease of dissolved metal concentrations at the same site by a factor of 2.5 (Cr and Ni) and 2.9 (Zn). It is suggested that the residence time is a key parameter in the control of elemental concentrations in the lagoon waters, as much as land-based emission sources.     

Partial Pressure of Carbon Dioxide in Coral Reef Lagoon Waters: Comparative Study of Atolls and Barrier Reefs in the Indo-Pacific Oceans

Factors controlling the CO₂ system parameters, including the partial pressure of CO₂ (PCO₂) in coral reef waters, were investigated in three mid-oceanic reefs of the Indo-Pacific region. Surface water PCO₂ in the lagoons of Majuro Atoll and Palau barrier reef in the Pacific were 25 µatm and 48 µatm higher than those of the offshore waters, respectively, while South Male Atoll lagoon of the Maldives in the Indian Ocean exhibited relatively small difference in PCO₂ compared to the offshore water. Observations from Majuro Atoll and Palau barrier reef are consistent with the view that calcium carbonate production predominates in coral reefs. On the other hand, results from South Male Atoll can be attributed to the thorough flushing of the lagoon, which is connected to the open ocean by numerous deep channels. The offshore-lagoon PCO₂ difference depends on system-level net organic-to-inorganic carbon production ratio while reef topography, especially residence time of the lagoon, has a secondary effect on the magnitude of the offshore-lagoon difference. A potential for releasing CO₂ might be more evident in an enclosed atoll where the reef water has a longer residence time. Oceanic atoll and barrier reef lagoons, which are in the terminal stage of evolutionary history of oceanic volcanic islands, have the potential to release CO₂ to the atmosphere.     

Hydrological characterisation of Ria de Aveiro,Portugal, in early summer

The hydrological features of Ria de Aveiro, a coastal lagoon on the northwest Atlantic coast of Portugal, were investigated in two sampling surveys carried out between 3/6 and 24/6/97 and between 29/6 and 7/7/97, respectively. There was a significant freshwater inflow into the lagoon during the first survey, especially in the first days, due to the recent rainfalls. Records concerning water level, salinity, temperature and current velocity were performed at several stations located along the four main channels of the lagoon. The type of tide at the mouth was determined and was observed that astronomical tide is the main forcing agent driving water circulation in Ria de Aveiro. The tide at the mouth is semidiurnal and the tidal wave propagation in the lagoon has the characteristics of a damped progressive wave. Typical estuarine longitudinal salinity and temperature gradients connected with the distance to the mouth were identified, whereas vertical and transverse gradients were found unimportant, except in the frontal zone between oceanic and fresh water masses. According to the results, even though Ria de Aveiro should be considered as vertically homogeneous, some channels may reveal characteristics of a partially mixed estuary, depending on the freshwater input.     

Physical oceanography of Rangaunu Harbour,Northland, New Zealand

Current meter, current drogue, salinity, temperature, and tidal elevation observations from Rangaunu Harbour are presented. The flow is dominated by the tides, the ebb tide in general being stronger than the flood. The time of high tide is increasingly delayed with distance from the open ocean. High tide at the head of the harbour lags about an hour behind that at the mouth. The phase of the flow relative to that of the elevation is less than that for a frictionless system. This difference from a quarter of a tidal period results from tidal energy dissipation and probably varies through the spring‐neap tidal cycle. The outer harbour has essentially coastal water which is exchanged each tide. Residence time of the inner harbour waters and the inner harbour shallows (where evaporation is sufficient to raise the salinity) is several days.     

Field studies of velocity,salinity and suspended solids concentration in a shallow tidal channel near tidal flap gates

The design and operation of mathematical models of solute mixing and sediment transport in estuaries rely heavily on the provision of good-quality field data. We present some observations of salinity, suspended sediment concentration and velocity at one of the tidal limits of a semi-enclosed tidal lagoon in Southern England (Pagham Harbour, West Sussex, UK) where the natural processes of tidal incursion and solute mixing have been heavily modified as a result of the construction of sea walls dating back to the 18th Century. These observations, made immediately downstream of two parallel tidal flap gates by conductivity-temperature-depth (CTD) profiler, and also using bed-mounted sensor frames to measure velocity at 2 fixed depths, have yielded a set of results covering 11 tidal cycles over the period 2002–04. It is clear from the results obtained that over a typical tidal cycle, the greatest vertical salinity gradients occur in the 1–2 h immediately after the onset of the flood tide, and that subsequently, energetic mixing acts to rapidly break down this stratification. Under moderate-to-high fresh water flows (>0.5 m³/s), the break-down in vertical salinity gradient is more gradual, while under low fresh water flows (<0.2 m³/s), the vertical salinity gradient is generally less pronounced. Estimates of Richardson number during the early flood-tide period reveal values that vary rapidly between <1 and about 20, with lower values occurring after around 1.5–2 h after low water. Observations of suspended sediment concentration vary widely even for similar tidal and fresh water flow conditions, revealing the possible influence of wind speed, the storage effects of the water in the lagoon downstream of the observation site, and the complexity of the hydrodynamics downstream of tidal flap gates. The data also show that most of the sediment transport is landward, and occurs during flood tides, with estimated total tidal landward flood tide flux of fine sediment of the order of 50–120 kg under low fresh water flow conditions. These observations, which reinforce the results presented in Warner et al. (2004) and elsewhere, can help to provide information about the appropriate techniques for managing sediments and pollutants, including nutrients from sewage effluent waters, in estuaries where hydraulic flap gates are used to control the entry of fresh water over the tidal cycle.     

Atoll lagoon flushing forced by waves

Water level and current measurements from two virtually enclosed South Pacific atolls, Manihiki and Rakahanga, support a new lagoon flushing mechanism which is driven by waves and modulated by the ocean tide for virtually enclosed atolls. This is evident because the lagoon water level remains above the ocean at all tidal phases (i.e., ruling out tidal flushing) and because the average lagoon water level rises significantly during periods with large waves. Hence, we develop a model by which the lagoons are flushed by waves pumping of ocean water into the lagoon and gravity draining water from the lagoon over the reef rim. That is, the waves on the exposed side push water into the lagoon during most of the tidal cycle while water leaves the lagoon on the protected side for most of the tidal cycle. This wave-driven through flow flushing is shown to be more efficient than alternating tidal flushing with respect to water renewal. Improved water quality should therefore be sought through enhancement of the natural wave pumping rather than by blasting deep channels which would change the system to an alternating tide-driven one.     

Correspondence between the distribution of hydrodynamic time parameters and the distribution of biological and chemical variables in a semi-enclosed coral reef lagoon

Hydrodynamic modeling can be used to spatially characterize water renewal rates in coastal ecosystems. Using a hydrodynamic model implemented over the semi-enclosed Southwest coral lagoon of New Caledonia, a recent study computed the flushing lag as the minimum time required for a particle coming from outside the lagoon (open ocean) to reach a specific station [Jouon, A., Douillet, P., Ouillon, S., Fraunié, P., 2006. Calculations of hydrodynamic time parameters in a semi-opened coastal zone using a 3D hydrodynamic model. Continental Shelf Research 26, 1395–1415]. Local e-flushing time was calculated as the time requested to reach a local grid mesh concentration of 1/e from the precedent step. Here we present an attempt to connect physical forcing to biogeochemical functioning of this coastal ecosystem. An array of stations, located in the lagoonal channel as well as in several bays under anthropogenic influence, was sampled during three cruises. We then tested the statistical relationships between the distribution of flushing indices and those of biological and chemical variables. Among the variables tested, silicate, chlorophyll a and bacterial biomass production present the highest correlations with flushing indices. Correlations are higher with local e-flushing times than with flushing lags or the sum of these two indices. In the bays, these variables often deviate from the relationships determined in the main lagoon channel. In the three bays receiving significant riverine inputs, silicate is well above the regression line, whereas data from the bay receiving almost insignificant freshwater inputs generally fit the lagoon channel regressions. Moreover, in the three bays receiving important urban and industrial effluents, chlorophyll a and bacterial production of biomass generally display values exceeding the lagoon channel regression trends whereas in the bay under moderate anthropogenic influence values follow the regressions obtained in the lagoon channel. The South West lagoon of New Caledonia can hence be viewed as a coastal mesotrophic ecosystem that is flushed by oligotrophic oceanic waters which subsequently replace the lagoonal waters with water considerably impoverished in resources for microbial growth. This flushing was high enough during the periods of study to influence the distribution of phytoplankton biomass, bacterial production of biomass and silicate concentrations in the lagoon channel as well as in some of the bay areas.     

Determination of residence time and mixing processes of the Ubatuba,Brazil, inner shelf waters using natural Ra isotopes

Coastal waters contain elevated dissolved activities of short-lived radium isotopes, ²²³Ra and ²²⁴Ra, having half-lives of 11.4 and 3.66 days, respectively. The input of these isotopes near the coast must be balanced by decay and mixing into the open ocean, where excess activities are zero. Since the decay rate is known, in the ideal case the mixing rate may be determined from the offshore distribution of these isotopes. This study found that samples collected in June 2000 followed the expected exponential decrease with distance offshore. We assign a dispersion coefficient of 28–39 m² s⁻¹ for this study. During January 2002 and November 2003, there was not a consistent decrease of activity with distance offshore. This is likely due to the ruggedness of the coastline, where many bays and small islands interrupt simple mixing patterns. To estimate exchange rates during 2002 and 2003, we used a model based on the decrease in the ²²⁴Ra/²²³Ra activity ratio (AR) with time for samples isolated from fresh inputs of Ra. This model yielded residence times of 1–2 weeks for samples collected within 20 km of the coast. We used this residence time to calculate the flux of ²²⁸Ra (half-life = 5.7 years) to the study area necessary to maintain the enrichment relative to ocean water. This enrichment is a factor of ten greater than the flux of ²²⁸Ra expected from submarine groundwater discharge (SGD) occurring within 50 m of shore.     

近60年来长江河口河势变化及其对水动力和盐水入侵的影响Ⅲ.盐水入侵

本文应用本系列论文Ⅱ中建立的长江河口水动力和盐水入侵三维数值模式,模拟长江河口20世纪50年代、70年代和2012年盐水入侵,定量分析不同年代河势下盐水入侵状况和变化程度及其原因。在北支,不同年代盐水入侵的变化是由分流比和潮差共同作用造成的。50年代北支盐水入侵较强,70年代大幅下降,中上段出现淡水,2012年盐水入侵极为严重,整个北支被高盐水占据,上段出现强烈的盐度锋面。50年代和2012年,北支盐水倒灌南支,大潮期间远大于小潮期间,2012年远强于50年代,70年代没有北支盐水倒灌南支现象。在南支,50年代、70年代南支大部分为盐度都小于0.45的淡水,在2012年大潮期间由于出现了强烈的北支盐水倒灌,南支上段出现盐度大于0.45的盐水。在南北港,在50年代盐水入侵最严重;大潮期间,北港净分流比南港大21.6%,北港盐度小于南港盐度,外海盐水主要通过南港入侵,出现南港盐水倒灌进入北港的现象。至70年代,南支主流转向南港,南港净分流比增大,比北港大10.4%,南港盐度明显小于北港盐度;南北港盐水入侵较弱。在2012年,南支主流再次转向北港,北港分流比比南港大10.4%,南港的盐水入侵再次强于北港。小潮期间,50年代由于南港分流比相比于大潮时更小,南港盐水上溯距离更远,上段盐度比更大;至70年代,北港分流比减少,盐水入侵减弱;至2012年,由于大潮时期北支倒灌的盐水在小潮期间到达北港,北港净盐通量比大潮时期大。由于潮动力减弱,小潮期间各年代垂向盐度分层更明显,盐水入侵变化与大潮期间一致。     

Sedimentation in mangroves and coral reefs in a wet tropical island, Pohnpei, Micronesia

A six-month-long study was conducted of the fate of turbid river plumes from the Enipein watershed in Pohnpei, Federated States of Micronesia. Pohnpei is one of the wettest places on earth, with a mean annual rainfall exceeding 4 m in the lowlands and 8 m in the highlands. The river waters were clear of sediment except after major storms with rainfall exceeding 5 cm day⁻¹. Following a storm, the river plume spread in the mangrove fringed estuary and in the coral reef lagoon. The waters were highly stratified in temperature, salinity, and suspended sediment concentration. The brackish water was flushed out in four days, while the suspended sediment all settled out in the estuary, in the mangroves, and in the lagoon including on the coral reefs, in less than one day. The mean rate of sedimentation exceeded 35 mg cm⁻² d⁻¹ both over the mangroves and on the adjacent coral reefs. While this leads to no detrimental effects on the mangroves, sediment smothers corals and leads to substantial coral mortality in the lagoon. The mud is not flushed out from the lagoon because there are no strong currents from waves or tides. This high sedimentation rate is attributable to poor farming and land-use practices on the upland areas.     

Understanding spatial and temporal dynamics of key environmental characteristics in a mesotidal Atlantic estuary (Douro,NW Portugal)

Ecosystem management and decision making process are dependent of a good knowledge of ecosystem functioning. Conceptual models allow knowledge organization through representation of relationships between variables and processes, facilitating management decisions. In this study, spatial and temporal characterization of key environmental variables, as well as relationships between them, was studied aiming the design of a conceptual model of the Douro estuary. This temperate mesotidal estuary is limited upstream by a hydroelectric power dam that controls freshwater inflow and prevents the propagation of the tide upstream, which, in turn, influences water circulation and biogeochemical dynamics of the system. During one year, from December 2002 to December 2003, water column data were collected monthly at 10 stations along the estuary, during ebb and flood tides. Spatial and temporal variability of water column salinity, temperature, nutrients, phytoplankton biomass, total particulate matter (TPM), phytoplankton primary production (PP), faecal coliform bacteria (FC) and community respiration (CR) were analysed. Salinity stratification was assessed by means of the Estuary Number (Ne) and variation of this index as well as other key characteristics with river flow was analysed. Freshwater discharge controlled salinity stratification and freshwater residence time. Ne indicated that the Douro was stratified for river flows <300 m³ s⁻¹, and freshwater residence time was >1 day for the same conditions. A decaying exponential relationship between PP and river flow was found, whereas nitrate and TPM increased logarithmically and linearly, respectively, with river flow. Regarding spatial distribution, nitrate and PP decreased downstream, showing that the river was a source of nutrients and phytoplankton, while the opposite trend was found for TPM, FC, ammonium and CR. The latter increase was probably due to untreated sewage discharge in the urbanized middle and lower estuarine stretches. Reduction of nitrate coming from the watershed and of bacterial contamination in the urban stretches of this highly modified water body, according to the European Water Framework Directive, emerges as the main water quality issues for this estuarine ecosystem.     

Numerical investigation of deep water circulation in the Faroese Channels

The overflow of dense water from the Nordic Seas through the Faroese Channels is investigated numerically using the Massachusetts Institute of Technology General Circulation Model. The model is forced by the removal of a barrier that separates different water masses in the bottom layer of the Faroe-Shetland Channel at the north-eastern boundary. An analysis of the output reveals that during its adjustment in the rotating channel the propagating flow is unstable and forms cyclonic and anti-cyclonic eddies in the Faroese Channels. The life-time of the cyclonic eddy is about 10 days, but an anti-cyclonic eddy that is formed upstream of the sill crest of the Faroe Bank Channel has a longer life-time. However, after 50 days it eventually loses its structure below 400 m due to the decay of a counter-rotating current. In the upper 400 m layer this anti-cyclonic eddy remains persistent for longer. Observational evidence of the eddy is confirmed by the tracks of experimental drifters released in the area and by the temperature and salinity fields observed in the Faroese Channels.The pinching of isotherms along the Wyville Thomson Ridge results in the concentration of cold water on the southern side of the Faroese Channels that overflows into the Rockall Trough. The model results demonstrate that the main part of the cold water outflows through the Faroe Bank Channel, rather than across the Wyville Thompson Ridge, due to Earth rotation. The apparent similarity of modelled temperature, salinity and velocity sections to recent measurements in this area adds confidence to these results.     

Estimating submarine groundwater discharge around Isola La Cura, northern Venice Lagoon (Italy), by using the radium quartet

The four naturally-occurring radium isotopes (²²³Ra, ²²⁴Ra, ²²⁶Ra and ²²⁸Ra) were used to estimate the submarine groundwater discharge (SGD) in the Isola La Cura marsh area in the northern Venice Lagoon (Italy). By determining the radium contributors to the study area (river, coastal ocean and sediments) the radium excess in the lagoon water was quantified through a mass balance model. This radium excess is attributed to a submarine groundwater discharge source and represents the most important input of radium. Possible endmembers were considered from analysis of groundwater samples (subtidal and marsh piezometers, marsh wells and seepage meters) that were enriched in Ra by one to two orders of magnitude relative to surface waters. In particular, a permeable layer at 80 cm depth in the surrounding marsh is considered to be representative of the most likely SGD source, although similar radium activities were measured in other subtidal porewater samples collected in the Isola La Cura area. The estimated SGD flux to the study area ranged from 1 · 10⁹ to 6 · 10⁹ L·d^− 1, the same order of magnitude as the overall riverine input to the lagoon (3 · 10⁹ L·d^− 1). A major fraction of this SGD flux is likely recirculated seawater, as evidenced by the endmember salinity. The water residence time of 2 days was estimated by both using the shortest-lived radium isotope and estimating the volume of water exchanged between the lagoon and the open sea during a tidal cycle (tidal prism approach). This SGD flux could be used to estimate the input of other chemical species (metals, nutrients, etc.) via SGD which might affect the Venice Lagoon ecosystem.     

Residence times of Manukau Harbour,New Zealand

Residence times of the three main channel systems of Manukau Harbour were calculated by methods based on the salinity measured in 10 tidal excursion segments: the freshwater replacement technique and a volume exchange model. The amount of freshwater present in the segments at high water was low (< 11% throughout the year), so salinities were high, with minimum values (winter) in each segment being in the range 30.1–34.0 (compared with a median value of 34.5 in the nearby coastal waters). The proportion of each segments’ volume which was exchanged with an adjacent segment during a tidal cycle varied markedly across the harbour, from ≤ 5 % in several shallow inner harbour segments to 65% in a deep segment adjacent to the entrance channel. For annual average conditions the freshwater replacement times of the channels were 11–18 days. Slightly longer residence times—12–26 days—were obtained from the volume exchange model. Calculated replacement times for two non‐conservative tracers— inorganic phosphorus and nitrogen—were approximately half those for freshwater, as 5–12% of these species disappeared from the water column each day.     

Freshwater transport estimates and the global overturning circulation: Shallow, deep and throughflow components

Meridional ocean freshwater transports and convergences are calculated from absolute geostrophic velocities and Ekman transports. The freshwater transports are analyzed in terms of mass-balanced contributions from the shallow, ventilated circulation of the subtropical gyres, intermediate and deep water overturns, and Indonesian Throughflow and Bering Strait components. The following are the major conclusions:

1.

Excess freshwater in high latitudes must be transported to the evaporative lower latitudes, as is well known. The calculations here show that the northern hemisphere transports most of its high latitude freshwater equatorward through North Atlantic Deep Water (NADW) formation (as in [Rahmstorf, S., 1996. On the freshwater forcing and transport of the Atlantic thermohaline circulation. Climate Dynamics 12, 799-811]), in which saline subtropical surface waters absorb the freshened Arctic and subpolar North Atlantic surface waters (0.45 ± 0.15 Sv for a 15 Sv overturn), plus a small contribution from the high latitude North Pacific through Bering Strait (0.06 ± 0.02 Sv). In the North Pacific, formation of 2.4 Sv of North Pacific Intermediate Water (NPIW) transports 0.07 ± 0.02 Sv of freshwater equatorward.In complete contrast, almost all of the 0.61 ± 0.13 Sv of freshwater gained in the Southern Ocean is transported equatorward in the upper ocean, in roughly equal magnitudes of about 0.2 Sv each in the three subtropical gyres, with a smaller contribution of <0.1 Sv from the Indonesian Throughflow loop through the Southern Ocean. The large Southern Ocean deep water formation (27 Sv) exports almost no freshwater (0.01 ± 0.03 Sv) or actually imports freshwater if deep overturns in each ocean are considered separately (−0.06 ± 0.04 Sv).This northern-southern hemisphere asymmetry is likely a consequence of the “Drake Passage” effect, which limits the southward transport of warm, saline surface waters into the Antarctic [Toggweiler, J.R., Samuels, B., 1995a. Effect of Drake Passage on the global thermohaline circulation. Deep-Sea Research I 42(4), 477-500]. The salinity contrast between the deep Atlantic, Pacific and Indian source waters and the denser new Antarctic waters is limited by their small temperature contrast, resulting in small freshwater transports. No such constraint applies to NADW formation, which draws on warm, saline subtropical surface waters .

2.

The Atlantic/Arctic and Indian Oceans are net evaporative basins, hence import freshwater via ocean circulation. For the Atlantic/Arctic north of 32°S, freshwater import (0.28 ± 0.04 Sv) comes from the Pacific through Bering Strait (0.06 ± 0.02 Sv), from the Southern Ocean via the shallow gyre circulation (0.20 ± 0.02 Sv), and from three nearly canceling conversions to the NADW layer (0.02 ± 0.02 Sv): from saline Benguela Current surface water (−0.05 ± 0.01 Sv), fresh AAIW (0.06 ± 0.01 Sv) and fresh AABW/LCDW (0.01 ± 0.01 Sv). Thus, the NADW freshwater balance is nearly closed within the Atlantic/Arctic Ocean and the freshwater transport associated with export of NADW to the Southern Ocean is only a small component of the Atlantic freshwater budget.For the Indian Ocean north of 32°S, import of the required 0.37 ± 0.10 Sv of freshwater comes from the Pacific through the Indonesian Throughflow (0.23 ± 0.05 Sv) and the Southern Ocean via the shallow gyre circulation (0.18 ± 0.02 Sv), with a small export southward due to freshening of bottom waters as they upwell into deep and intermediate waters (−0.04 ± 0.03 Sv).The Pacific north of 28°S is essentially neutral with respect to freshwater, −0.04 ± 0.09 Sv. This is the nearly balancing sum of export to the Atlantic through Bering Strait (−0.07 ± 0.02 Sv), export to the Indian through the Indonesian Throughflow (−0.17 ± 0.05 Sv), a negligible export due to freshening of upwelled bottom waters (−0.03 ± 0.03 Sv), and import of 0.23 ± 0.04 Sv from the Southern Ocean via the shallow gyre circulation.

3.

Bering Strait’ssmall freshwater transport of <0.1 Sv helps maintains the Atlantic-Pacific salinity difference. However, proportionally large variations in the small Bering Strait transport would only marginally impact NADW salinity, whose freshening relative to saline surface water is mainly due to air-sea/runoff fluxes in the subpolar North Atlantic and Arctic. In contrast, in the Pacific, because the total overturning rate is much smaller than in the Atlantic, Bering Strait freshwater export has proportionally much greater impact on North Pacific salinity balances, including NPIW salinity.

     

Line P ocean temperature and salinity, 1956-2005

Vertical profiles of temperature and salinity have been measured for 50 years along Line P between the North American west coast and mid Gulf of Alaska. These measurements extend 1425 km into the gulf at 13 or more sampling stations. The 10-50-m deep layer of Line P increased in temperature by 0.9 °C from 1958 to 2005, but is significant only at the 90% level due to large interannual variability. Most of this increase in temperature accompanies the 1977 shift in wind patterns. Temperature changes at 100-150 m and salinity changes in both layers are not statistically significant. Much of the variance in temperature is in the upper 50 m of Line P, and temperature changes tend to be uniform along Line P except for waters on the continental margin. Salinity changes are dominated by variability in the halocline between 100 and 150 m depth and are less uniform along Line P. Largest oscillations in temperature and salinity are between 1993 and 2003. These events can be understood by considering changes in eastward wind speed and wind patterns that are revealed in the first two modes of the Pacific Decadal Oscillation. Changes in these patterns are indicators for both Ekman surface forcing (Surface ocean currents flow to the right of the wind direction) and Ekman pumping (Surface waters diverge away from regions of positive wind stress curl, leading to upwelling of colder saltier water). Changes in temperature along the nearshore part of Line P suggest Ekman surface forcing is the stronger of the two processes in the upper layer. The change in salinity anomalies in the halocline along the seaward end of Line P, following the wind shift in 1977, is in agreement with enhanced upwelling caused by stronger Ekman pumping in this region.