«A global boundary current circulation observing network U. Send, R. Davis, J. Fischer, S. Imawaki, W. Kessler, C. Meinen, B. Owens, D. Roemmich, T. ...»
A global boundary current circulation observing network
U. Send, R. Davis, J. Fischer, S. Imawaki, W. Kessler, C. Meinen, B. Owens, D. Roemmich, T.
Rossby, D. Rudnick, J. Toole, S. Wijffels, L. Beal
The global system of oceanic boundary currents is a critical element of the ocean’s role in
climate, yet the current sustained observing infrastructure is poorly suited for capturing the
relevant processes and variability. Broadly speaking, western boundary currents (WBCs) represent a key climate mechanism while eastern boundary currents (EBCs) are strategic locations for societally relevant impacts of climate processes.
WBCs, by balancing the entire interior circulation of the oceans in a narrow, swift, warm current, play a dominant role in the poleward heat transport by the oceans, which in turn is comparable to heat transport by the atmosphere at the tropical/subtropical boundary.
Equatorward low-latitude WBCs in the Pacific are thought to contribute much of the meridional mass flux variability associated with ENSO (Meinen et al. 2001, Kug et al. 2003), and subpolar WBCs are critical due to their buoyancy transport in latitudes with low water column stability (e.g. Katsman, Spall, and Pickart 2004, Talley 2008). Deep boundary currents are the essential contributor to the thermohaline circulation. Beyond their direct transport of heat, mass and buoyancy, theory suggests that western boundary currents play a still poorly-understood role in the destruction and redistribution of potential vorticity (PV) that has been created by the basinscale winds, that in turn is key to the homogenized PV regions of the interior mid-latitude gyres.
Equatorward WBCs in the tropics are much less studied, but should serve an analogous function in allowing the PV modification that permits flow to and across the equator, as is required in all the tropical oceans.
EBCs are relevant as they display the local effects of climate variability and are of vital economic and ecological importance for a variety of reasons. Some of the world’s most important fisheries are along eastern boundaries, because they are sites of active upwelling, and strong eddy activity, but the ecosystems are often under stress and subject to important management activities. Also, variability in sea surface conditions along the eastern boundary affect atmospheric conditions along the western side of continents. Finally, the eastern boundary is tied robustly to the equator through coastally trapped waves, so that equatorial anomalies (as due to an El Niño) propagate poleward, causing fundamental changes in circulation and water properties. The mechanisms are still poorly understood by which the effects of equatorial disturbances reach poleward. The relative roles of atmospheric teleconnections, coastally trapped waves, and direct advection must be elucidated. Observations are also required of the pathways by which water reaches the upwelling zone, and how water properties get modified along this path. Such data are essential to better understanding and management of fisheries. A challenge for an observing system is the role of eddies in the transport of physical and biological properties, and in the regulation of primary productivity. Eddies are strong relative to the mean flow along eastern boundaries, so their quantification is important.
A number of boundary current properties should be observed in a sustained fashion in order to complete a global ocean observing system. Foremost, the transport of mass, heat, and fresh water by these currents must be monitored to complement broad-scale observations and define basin-wide transports. Time series of both the integral of transports through trans-basin sections
and the structure of temperature, salinity and velocity fields are needed to:
(1) establish the mean and seasonal cycles of mass, heat, and freshwater transports in order to explain the global circulation, heat engine, and water cycle and to validate the representation of these processes in dynamical models used to simulate global climate and its long-term variations;
(2) define interannual climate variability in order to (a) identify the key processes linking the ocean and atmosphere and connecting different regions, (b) characterize the physical environment affecting ecosystem management, and (c) define the temporal sequences of change to validate models used to forecasting climate variability; and (3) describe, in real-time, the structure of non-seasonal variability to assist, through data assimilation, initialization of models used to forecast weather and climate for operational purposes.
The program’s objectives also serve to define the user community for the observations.
Boundary current and basin-wide transports and fluxes are essential for basic oceanographic/climate research and for societally-mandated documentation and attribution of climate variability and change. As ocean and coupled modelling skill levels grow, applications grow in ocean data assimilation modeling and seasonal-to-decadal prediction. Indeed, the global ocean observing system, for all of its users, is seriously incomplete without boundary current observations.
Beyond mass, heat, and freshwater transports important to climate, a number of internal boundary-current properties need to be observed. These include the impact of the strong eddy activity, changes in potential vorticity, air-sea interaction, and ecosystem dynamics.
Having relatively small scales, boundary currents usually cannot be adequately sampled with the principal broadscale networks that we rely on in the interior: Argo floats and satellite altimetry.
The inherent properties of boundary currents pose particular observational challenges. The scales must be resolved to measure heat and freshwater transport even where geostrophy allows volume transport to be computed without such resolution. Boundary currents produce turbulence on multiple scales in response to coastline, bathymetric irregularities, and flow instabilities. The speed of WBC flows means they are inherently nonlinear, producing internally-generated variability that can be the dominant term in the momentum and vorticity balances, and which demands sustained sampling. High flow speeds through thick layers can make operation of various platforms untenable in some regions. WBCs often have large vertical extent, which can require full-depth profiling and cause assumptions about reference levels to be especially unrealistic. Finally intense fishing activity close to coasts poses problems of vandalism, while the need to work within EEZ zones can lead to political complications.
These diverse challenges will not be met be any particular tool, and a global western boundary current strategy will demand multiple observational techniques tuned to the particular conditions of each current system. This document summarizes the challenges and technologies, existing and newly-emerging, that are proposed to meet them.
Mooring Arrays Arguably, bottom-anchored moorings supporting instruments that span the water column are the “gold standard” for observing the time-varying ocean currents and water properties over sustained periods of a year or longer. The state of the art for mooring deployment duration is around 1 year for surface moorings and 2-2.5 years for subsurface systems, with 4-5-year subsurface installations undergoing proof-of-concept demonstrations now. Sensor durations however lag behind. While discrete conductivity-temperature-pressure sensors and some other devices can easily accommodate a 15-minute sample rate for 2+ years, many of the other instruments now available exceed their battery or memory capacities after about a year at this sample rate. The extended durations cited above require sampling at reduced frequency, risking aliasing error. Remarkably, some 30 years after the invention of the Vector Averaging Current Meter, there is no modern current meter presently available that is able to acquire “true” vectoraveraged velocity data (or the practical equivalent) from the deep ocean at ½ hour interval for 2.5 years (the effective limit of the VACM). The sample-rate/endurance issue is even more problematic for the upper ocean where the buoyancy period is just a few 10s of minutes or shorter.
While often deployed in 2-dimensional arrays to explore flow dynamics, moorings have also been widely used to estimate transport by arranging them in a “picket-fence” configuration across a current. Perhaps the most widely-cited open-ocean transport value derived from a mooring line is the International Southern Ocean Study Drake Passage estimate for the Antarctic Circumpolar Current at 134 +/- 13 Sv (Whitworth, 1983; Whitworth and Peterson, 1985). As the ISOS program discovered, a key design requirement for a transport array is that the mooring spacing be less than the decorrelation distance of the flow variations. Some ISOS mooring losses seriously degraded the array resolution in places, leaving open the possibility that at times, the Drake Passage array missed some of the ACC transport. Complementary horizontallyaveraged velocity estimates derived from bottom pressure and dynamic height difference estimates spanning these gaps and the geostrophic relation were helpful in quantifying such errors. Another design requirement is that a transport array fully span the current being studied.
Apart from those regions where a current is confined bathymetrically and it is feasible for the array to extend that full distance, the point where a particular current ends can be quite nebulous.
This can make difficult transport estimate intercomparisons between arrays at different locations along a flow or between observations and model results. One aid to the former is the ability to partition transport estimates by density since bounds on the diapycnal velocity can be estimated.
Such partitioning requires knowledge of the ocean temperature, salinity and velocity profiles.
These array design considerations, endurance limits and the costs associated with building, deploying and recovering oceanographic moorings dictate that the community can afford to instrument only a limited number of sites with transport-resolving arrays. Less-costly alternative approaches are detailed in adjoining paragraphs, but there is no capability presently able to return high temporal- and spatial-resolution volume and water-property transport estimates with accuracy comparable to those from a dense oceanographic moored array.
Example: Labrador Current Downstream of the Greenland – Scotland overflows and the subsequent entrainment of ambient water, and also downstream of the main subpolar convection sites (the Labrador and Irminger Seas), North Atlantic Deep Water is first combined in the Deep Western Boundary Current (DWBC) at the exit of the Labrador Sea. From there, all constituents of NADW are exported to both, the basin-wide subpolar recirculations and the deep limp of the MOC (e.g. Bower et al.
2009). A moored ‘Labrador Sea Export Array’ was installed at this location to monitor the strength and variability of the DWBC on time scales from intraseasonal to decadal. The initial array, deployed from 6/1997 to 6/1999, consisted of 5 full ocean depth moorings with 21 current meters and ADCP’s covering all deep water levels from upper LSW down to the DSOW. In subsequent deployment periods, the array was reduced to 1 to 3 moorings, but the most recent 1year installment (5/2009) has now 5 moorings across the core of the outflow.
Figure 1: Two-year mean alongshore flow of the Deep Labrador Current (DLC) at the exit of the Labrador Sea (Fischer et al., 2004), with transports in water mass classes defined by isopycnals; rhs: a decade of current meter records from the centre of the DLC (blue circles) for upper, LSW, and DSOW levels; annual means in red.
The current meter data, supplemented by shipboard LADCP and profiling floats, are used to establish a ‘reference’ state and corresponding transports of the DLC (Fischer et al., 2004; Fig 1;
lhs). It is also used for additional comparisons (Dengler et al., 2006) and monitoring of possible long term changes. A pronounced, about 100 km wide boundary current – the Deep Labrador Current (DLC) -- hugs the continental slope off the Labrador shelf break. The flow is baroclinic at the shelf edge, and also further offshore where the DSOW is focused in a bottom-intensified current core. The intermediate layers – mainly LSW - show only weak vertical shears.
During the last decade, weak convection activity (Avsic et al., 2006) and the lateral advection of heat by eddies led to substantial warming of the upper 2000 m of the Labrador Sea. This warming exists in the central Labrador Sea at a rate of 0.05 °C/y and is also present at the corresponding depth range in the DWBC along its path from the exit of the Labrador Sea to the Tail of the Grand Banks.
In contrast, the intensity of the DLC does not show any long term variations during that period, although interannual transport / flow variability of 10 – 20% has been observed.
Ships of Opportunity (SOOP) The unique advantages of SOOP (Goni et al., 2009) sampling in the boundary currents are both logistical and scientific, including (i) regular and low-cost access along repeating transects and (ii) the ability to integrate boundary currents with interior circulations or into basin-wide transports. The core SOOP measurement is eXpendable BathyThermograph (XBT) profiling to 800 m depth, with the High Resolution mode of XBT sampling (HRX) providing profiles at spatial intervals 10 to 50 km in boundary current and interior regions respectively (e.g.
Roemmich et al., 2001). Additional SOOP measurements may consist of expendable conductivity-temperature-depth (XCTD) profiling, acoustic Doppler current profiling (ADCP, Rossby et al., 2005), marine meteorological observations, and surface water properties.
Regular SOOP/HRX sampling is being carried out in all five mid-latitude western boundary currents, with multiple crossings of the Kuroshio, the Gulf Stream, and the East Australian Current/East Auckland Current system. HRX boundary current sampling began as early as 1986 (East Auckland Current) and most of the time-series are longer than 15 years. Combinations of HRX and satellite altimetry (Figure 1, Ridgway et al., 2008) have demonstrated the ability to estimate boundary current transport variability accurately on interannual and decadal timescales.