1
Submitted to DSR II, Special Issue on the Black Sea
Currents at intermediate and deep layers of the Black Sea
inferred from autonomous profiling floats
G. Korotaev1, T. Oguz2, S. Riser3
1
Marine Hydrophysical Institute, National Academy of Sciences of Ukraine,
Sevastopol, Ukraine
2
Institute of Marine Sciences, Middle East Technical University, Erdemli, Turkey
3
School of Oceanography, University of Washington, Seattle, USA
Abstract
1.5 year observations of float trajectories at three depths of the Black Sea (200,
750, 1550m) offered a new perspective for the circulation structure at depths below
the permanent pycnocline. The observations provided first direct and quantitative
evidence of strong currents and organized flow structure at intermediate and deep
layers. Their magnitude is typically about 5-10% of the surface currents, and may be
as high as 5 cm s-1 at 1550 m along the steep topographic slope around the periphery.
A well-defined cyclonic circulation was shown to persist to the bottom without any
reversal in the direction of deep circulation with respect to that in the surface. The
deep currents were found to be strongly controlled by steep topographic slope, and
were correlated well with the surface currents at seasonal and longer time scales.
These observations thus provide the first evidence for discarding the traditional
concept of weak currents and associated very slow deep thermohaline circulation
system in the Black Sea.
1. Introduction
Albeit physical structure of the Black Sea have been observed rather intensely
during the last century, overwhelming majority of the hydrographic measurements
have been conducted within the uppermost 300, which occasionally extended up to
750 m. Because of its very weak stratification, the deep layer has been usually
2
assumed to be almost motionless and possess a temporally invariant thermohaline
structure at seasonal-to-interannual time scales. The currents and flow fields above the
permanent pycnocline (located around 100-200 m depth) have then been estimated
geostrophically by locating the depth of no-motion around 300-500m.
These
estimates, together with the satellite data, suggested a highly dynamic upper layer
circulation structure prevailing throughout the year (Oguz et al., 1992; 1993, 1994;
1998; Sur et al., 1994, 1996; Sur and Ilyin, 1997; Oguz and Besiktepe, 1999;
Gawarkiewicz et al., 1999; Ginsburg et al., 2000, 2002a,b; Sokolova et al., 2001;
Afanasyev et al., 2002; Korotaev et al., 2001 and 2003; Zatsepin et al., 2003). The
building block of the circulation system involve a basinwide cyclonic circulation cell,
encircled by the 40-80 km wide Rim Current zone and a wealth of anticyclonic and
cyclonic eddies located on the coastal and interior sides of the Rim Current,
respectively. The entire system evolves continuously at different time scales subject
to external forcing and internal dynamics. The schematic pattern representing the
most robust and quasi-persistent features of the upper layer circulation is shown in
Fig. 1. Direct observations from the surface buoys (Eremeev et al., 2003) and the
ADCP measurements (Oguz and Besiktepe, 1999) provided magnitude of the Rim
Current jet about 40-50 cms-1, increasing occasionally up to 100 cms-1 along the main
axis of the jet within the upper 100m. The cyclonic circulation is driven by the quasipermanent positive wind stress curl field intensifying in winter and weakening in
summer months, and is further supported by the buoyancy contrast between the fresh
water inflow from rivers around the basin and the salt water supply through the
Bosphorus Strait (Whitehead et al., 2001; Korotaev et al., 2001).
The multi-ship hydrographic surveys carried out within the framework of the
HydroBlack and ComsBlack Programs constituted during the early 1990s the first
reliable data sets for exploring deep water physical characteristics of the Black Sea
(Oguz et al., 1993; 1994; 1998). One of the most important findings of these studies
was an indication of horizontally as well as vertically highly-structured flow systems
formed by a series of large mesoscale eddies (with typical size of about 100 km) and
sub-basin scale gyres (with typical size of several hundreds of km’s) within the
intermediate layer between 300 and 1000 m. These observations, providing only the
baroclinic component of the circulation, however underestimated intensity of the flow
3
field particularly along the strong topographic slope zone around the periphery of the
basin. As they were carried out at best with 20nm horizontal grid resolution, they were
not able to capture well the Rim Current jet structure along the narrow topographic
slope zone around the basin. The strength of the intermediate layer flow system has
partly been elucidated by the ADCP absolute current velocity measurements carried
out during April 1993 in the western Black Sea. These measurements reported for the
first time the sub-pycnocline currents of 10-20 cm s-1 within 200-350 m layer (Oguz
and Besiktepe, 1999), the latter depth being the approximate limit of the ADCP
measurements.
The recent US-Turkish-Ukranian collaboration between the Office of Naval
Research, School of Oceanography-University of Washington, Institute of Marine
Sciences-Middle East Technical University, and Marine Hydrophysical InstituteUkranian Academy of Sciences has led to a new opportunity to explore further the
sub-pycnocline characteristics of the flow field by means of three autonomous
profiling floats. As a part of our ongoing efforts towards the development of an
operational oceanographic system within the framework of the Black Sea GOOS
Program, the primary goals of this project are to enhance the near-real-time
measurement capability of temperature and salinity within the water column, and to
promote a better understanding of the intermediate and deep layer circulations and
water mass characteristics at seasonal and longer time scales. The present paper
describes some results from these Lagrangian subsurface current measurements
performed during the September 2002-April 2004 observation period.
2. Methodology
Three profiling floats, manufactured by Scool of Oceanography-University of
Washington, were deployed from R.V Bilim during 2 September 2002 within the
southwestern part of the sea at approximately 100 nm offshore of the Bosphorus
Strait. Two of the floats were put into the parking depths of 1550m and 200m at the
site of 42.25oN and 30.34oE. This location, denoted by A in September 2002 mean sea
level anomaly map (Fig. 2), roughly lies inside the cyclonic gyre of the western basin.
The third one was deployed into the parking depth of 750 m at the site 41.83oN and
29.83oE which is slightly closer to the coast. Its location, denoted by B in Fig. 2, lies
4
along the main axis of the Rim Current system. They were programmed to drift with
currents at their parking depths for a week and then to measure temperature and
salinity vertically at 58 levels from 1550 m to the sea surface. The floats parked at
shallower depths than 1550m first come down to the sampling depth of 1550m before
they rise up to the surface. When they reach the sea surface they transmit the data to
orbiting satellites that are equipped with the ARGOS system within approximately 8to-12 hours.
The floats then sink to their parking depths and begin another
observation cycle. The accuracy of the temperature, salinity and pressure sensors is
typically better than 0.005oC, 0.01, 5m (The ARGO Science Team, 2000). The
temperature and salinity profiles together with float trajectories are documented in
near-real time mode at the URL site http://flux.ocean.washington.edu/metu. The
currents at their parking depths were estimated by the displacements from their last
positions before diving to their previous position after surfacing. Their trajectories
(Fig. 3a-c) are subject to some errors as they are periodically modified by the surface
transports during their 8-12 hour presence at the surface every week.
The interpretation of float trajectories is further complemented by the
horizontal circulation maps derived from the analysis of altimeter data provided at the
URL site http://las.aviso.oceanobs.com/las/servlets/dataset. The original data set
comprises weekly global merged products of several satellites such as GFO,
Topex/Poseidon (or Jason-1) and ERS-2 (or Envisat) at 1/3o resolution. The data was
first extracted for the Black Sea region and then optimally interpolated to 0.1o square
grid in order to produce a series of sea level anomaly maps at selected times of the
observation period (Fig. 4a-h).
3. Float trajectories and variations of current velocity
Soon after their deployments, all three floats drifted eastward almost parallel to
the coast by the peripheral current system (Fig. 3a-c). The float BS634 at 200m first
moved eastward very slowly (~1 cm s-1) suggesting its entrapment within the western
cyclonic gyre during its first two months (Fig. 2). By mid-November, as leaving the
eddy and approaching towards the Anatolian coast during its eastward drift, it was
encountered with the Rim Current system. Its speed of translation was then increased
to about 5 cms-1 (Fig. 5a). Thereafter, the float started drifting at a gradually
5
increasing rate of 10 cms-1 during the early winter (December-January) and 15 cms-1
during next February-March period as it moved along the southeastern and eastern
coasts. Stronger current speeds and more organized Rim Current system, known to be
a robust feature of the Black Sea winter upper layer circulation, are therefore also
evident at 200m depth. The absence of the Batumi gyre during winter months (Fig.
4b) is supported by the drift of the float BS634 along the southeastern corner. In April
2003, the float moving along the northeastern coast is trapped within the Caucasian
coastal anticyclonic eddy (Fig. 4c), which is one of the quasi-persistent features of the
circulation system usually persisting during spring and summer months (Korotaev et
al., 2002; 2003). The anticyclonic rotation of the float within the eddy is further
supported by the changes in both magnitude and direction of the current components
in Fig. 5a.
As the Caucasian eddy was disintegrated in August 2003 (Fig. 4d), the float
started drifting along the Rim Current jet in September, and then was diverted towards
south in October to move along the western flank of a cyclonic eddy of the eastern
interior basin (Fig. 4e).
The float continued to move first southward along the
periphery of the gyre in November and December 2003 (Fig. 4f, g), and then turned
northward along its eastern flank during January-February 2004 (Fig. 4h). Upon
reaching the northern coast, it continued to travel along the periphery of the cyclonic
gyre towards the Crimean coast on the west (Fig. 4h). Excluding approximately 5
month of entrapment both in the Caucasian anticyclone and the eastern basin interior
cyclone, the float accomplished its approximately 2000 km route along the periphery
of the basin from its deployment site near the Bosphorus Strait to the tip of the
Crimean Peninsula within about 300 days. This implies an average speed of the Rim
Current about 7 cm s-1 at 200 m.
The float BS631 drifting at 750m followed initially very similar route along
the southern coast (Fig. 3b), and reached the Cape Sinop region in January 2003 with
one month delay with respect to the float BS634. The speed of its translation was
around 2-4 cms-1, which attained its maximum value of about 4.5 cms-1 during
December 2002 (Fig. 5b). Afterwards, it experienced a different route, and instead of
continuing to move along the southern coast of the eastern basin it was deflected
towards north into the basin’s interior during February 2003. As suggested by the
6
altimeter data (Fig. 4a), the deflection seems to be caused by a temporary diagonally
tilted, elongated anticyclonic gyral motion extending in the northeast-southwest
direction within the eastern basin. This system apparently blocked eastward
penetration of the current system along the southern coast. Once it entered into the
interior, it moved slowly to the northeast within the next three months. The slow
movement on the order of 1 cms-1 (Fig. 5b) was caused primarily due to frequent
diversion of its excursion by eddy activity in the region, rather than weakness of the
currents. Towards the end of May 2003, it reaches near the northeastern coast, and
joins into the anticyclonic Kerch eddy for the next two months (Fig. 4c). By the end of
August, as the Kerch eddy was disintegrated and the region was replaced by a system
of cyclonic eddies (Fig. 4d), the float started drifting southwest into the interior basin
once again at a maximum speed of 5 cms-1 (Fig. 4e, f), and arrived during the midNovember 2003 at its position very close to that left previously in February 2003. At
the end of November 2003, the float was captured by the Rim Current system (Fig.
4h), and started to move eastward along the Turkish coast at a speed of approximately
3-to-5 cms-1 (Fig. 5b). The float spent about 310 days to reach its last position at
39.5ºE during April 2004, when its northeast and southwest excursions within the
interior basin were excluded. We recall that the float BS634 required only 170 days
to travel the same distance. In other words, the float BS634 traveled almost twice
longer path during the same time period. The average speed of the Rim Current at
750m is then estimated as ~4.0 cms-1, which rougly corresponds to the half of its
speed at 200 m.
The float BS587, launched at the parking depth of 1550m at the same site with
the float BS634 first drifted towards the coast at a rate of 2-3 cms-1 (Fig. 5c) during its
first two months, and then moved along the coast eastward by the Rim Current at a
gradually increasing speed up to 5 cms-1, reaching the Cape Sinop during the midFebruary 2003 (Fig. 3c). Its elapsed time to the Cape Sinop from the deployment site
is almost same with that of the float BS631 at 750m, indicating similar current speeds
of the boundary flow system at intermediate and deep layers. On the eastern side of
the Cape Sinop, the float translation was slowed down gradually up to 1cms-1 during
the subsequent three months prior to the entrapment of the float by the quasipermanent Kizilirmak anticyclonic eddy (Fig. 3c).
Starting by June 2003, it
7
underwent an anticyclonic loop within the eddy until the end of September 2003.
When it approached towards the coast to depths shallower than 1550m, it resided at
the bottom and was able to move only once a week when it came to the surface.
Leaving the Kizilirmak eddy by November 2003, the float continued to move along
the coast further eastward and was captured by another coastal anticyclone in winter
2004.
The topographic control of deep currents is shown remarkably well by the float
trajectories in Fig. 3c. The float, which was deployed at a site having the total depth
of 2200m, immediately started to move along this isobath up to 34oE longitude which
was then deflected towards offshore by joining a cyclonic gyral motion. The flow was
also subject to cross-isobath flow after its cycling within the interior basin. But, once
it arrived at the Kizilirmak coastal anticyclonic region, the strong topographical
control became effective as can be noted by the anticyclonic loop within the canyon
type structure and strong currents along the ridge its downstream. The strong deep
current has been explained by Korotaev (2004) as a result of the resonant
intensification of deep currents due to closed potential vorticity contours, which can
be noted by the topographic structure shown in Fig. 3c.
Since the floats spent 8-12 hours on the surface once per week, it is possible to
estimate temporal evolution of surface currents with respect to those at deeper levels.
The scatter plots of zonal and meridional components of currents at the surface and
200 m (Fig. 6a,b) suggest strong correlation with the magnitude of surface currents
roughly ten times more intense than that on 200m. Time evolution of the current
components (Fig. 7a,b) indicates that the correlations are high for all resolved time
scales (i.e., greater than a week) and therefore suggests vertically uniform current
structure within the upper 200 m. In other words, the current structure at 200m depth
reflects mainly characteristics of the upper layer circulation system. However, good
correlation between currents at the surface and deeper levels holds at low frequencies
only (at seasonal and longer). The scaling coefficients between the surface and
intermediate (750m) and deep (1550m) layers are respectively 0.06 for the zonal flow
and 0.045 for the meridional flow.
4. Conclusions
8
The 1.5 year observations performed by means of three autonomous profiling
floats deployed into the pycnocline layer at 200m, intermediate layer at 750m and
deep layer at 1550m provided new set of information about the strength and
variability of the flow field below the upper layer, which is permanently separated
from the rest of the water column by a strong pycnocline layer. The data suggested
the active role of mesoscale features on the basinwide circulation system at 200m,
similar to the case observed in the upper layer (<100m) circulation system. The
currents reach a maximum intensity of 15 cm s-1 along the Rim Current jet around the
basin. The floats at the intermediate (750 m) and deep (1550 m) layers also reflect the
role of mesoscale eddies on the flow field which has the maximum currents of about 5
cm s-1. The deep layer currents are found to flow along the strong topographic slope
zone by following a constant potential vorticity isoline. The orientation of the currents
is cyclonic at all depths, and no flow reversals and circulation cells in the opposite
directions were noted in the data. These findings reveal a new perception of highly
dynamic intermediate-deep layer circulation, which is contrary to traditional belief of
almost an inert circulation system.
Acknowledgement
Authors are grateful to J. Murray for his efforts in realization of the Black Sea
profiling float program supported by the ONR-Europe NICOP. G. Korotaev
acknowledges partial support by Science and Technology Center in Ukraine “Remote
Sensing of Marine Ecological Systems” for the project number 2241. This work is a
contribution to the Black Sea GOOS and ARENA projects.
References
Afanasyev, Y.D., A.G. Kostianoy, A.G. Zatsepin, and P.M. Poulain, 2002. Analysis of
velocity field in the eastern Black Sea from satellite data during the Black Sea ’99
experiment. J. Geophys. Res., 107, 10.1029/2000JC000578.
Eremeev, V.N., E. Horton, V.S. Motyzhev, S.V. Motyzhev, P.- M. Poulain, S.G.
Poyarkov, S.V. Stanichny, A.G. Zatsepin, 2003. Drifter monitoring of Black Sea in
2001/2002. WMO-IOC Data Bouy Cooperation Panrl Report,
http://www.dbcp.noaa.gov/dbcp/doc/DBCP22/DOCS_DBCP22/09_VNEremeev.doc
9
Gawarkiewicz, G., G. Korotaev, S. Stanichny, L. Repetin, and D. Soloviev, 1999.
Synoptic upwelling and cross-shelf transport processes along the Crimean coast of the
Black Sea. Cont. Shelf Res., 19, 977-1005.
Ginsburg, A.I., A.G., Kostianoy, D.M., Soloviev, S.V., Stanichny, 2000. Remotely
sensed coastal/deep-basin water exchange processes in the Black Sea surface layer.
In Satellites, Oceanography and Society, D. Halperneds. Elsevier Oceanography
Series, Amsterdam, 63, pp. 273-285.
Ginsburg, A.I., A.G., Kostianoy, N.P., Nezlin, D.M., Soloviev, S.V., Stanichny,
2002a. Anticyclonic eddies in the northwestern Black Sea. J. Mar. Syst., 32, 91-106.
Ginsburg, A.I., A.G., Kostianoy, V.G., Krivosheya, N.P., Nezlin, D.M., Soloviev,
S.V., Stanichny, V.G. Yakubenko, 2002b. Mesoscale eddies and related processes in
the northeastern Black Sea. J. Mar. Syst., 32, 71-90.
Korotaev G.K. Intensification of a mesoscale basin deep circulation under the
influence of the bottom topography. Marine Hydrophys. Journal, N2 (accepted, in
Russian).
Korotaev, G.K., O.A., Saenko, C.J., Koblinsky, 2001. Satellite altimetry observations
of the Black Sea level. J. Geophys. Res., 106, 917-933.
Korotaev G.K., Oguz T., Nikiforov A.A., Beckley B.D., Koblinsky C.J., 2002.
Dynamics of the Black Sea anticyclones derived from spacecraft remote sensing
altimetry. Earth Research from Space, No. 6, pp 1-10 (in Russian).
Korotaev, G.K., T. Oguz, A. Nikiforov, C. J. Koblinsky, 2003. Seasonal, interannual
and mesoscale variability of the Black Sea upper layer circulation derived from
altimeter data. J. Geophys. Res., 108(C4), 3122, doi:1029/2002JC001508.
Oguz, T., S., Besiktepe, 1999. Observations on the Rim Current structure, CIW
formation and transport in the western Black Sea. Deep Sea Res.I, 46, 1733.1753.
Oguz, T., P., La Violette, U., Unluata, 1992. Upper layer circulation of the southern
Black Sea: Its variability as inferred from hydrographic and satellite observations. J.
Geophys. Res., 97, 12569-12584.
Oguz, T., V.S., Latun, M.A., Latif, V.V., Vladimirov, H.I., Sur, A.A., Makarov, E.,
Ozsoy, B.B., Kotovshchikov, V.V., Eremeev, U., Unluata, 1993. Circulation in the
surface and intermediate layers of the Black Sea. Deep Sea Res. I, 40, 1597-1612.
10
Oguz, T., D.G., Aubrey, V.S., Latun, E., Demirov, L., Koveshnikov, H.I., Sur, V.,
Diacanu, S., Besiktepe, M., Duman, R., Limeburner, V.V., Eremeev, 1994.
Mesoscale circulation and thermohaline structure of the Black sea observed during
HydroBlack'91. Deep Sea Research I, 41, 603-628.
Oguz, T., P., Malanotte-Rizzoli, D., Aubrey, 1995. Wind and thermohaline circulation
of the Black Sea driven by yearly mean climatological forcing. J. Geophys. Res.,
100, 6846-6865.
Oguz. T, L.I. Ivanov, S. Besiktepe, 1998. Circulation and hydrographic characteristics
of the Black Sea during July 1992. In: Ecosystem Modeling as a Management Tool
for the Black Sea, L. Ivanov and T. Oguz (eds). NATO ASI Series, Environmental
Security-Vol.47, Kluwer Academic Publishers, Vol. 2, pp.69-92.
Sokolova, E., E.V. Stanev, V. Yakubenko, I. Ovchinnikov, R. Kosyan, 2001. Synoptic
variability in the Black Sea. Analysis of hydrographic survey and altimeter data. J.
Mar. Syst., 31, 45-63.
Sur, H.I. and Y.P. Ilyin, 1997. Evolution of satellite derived mesoscale thermal
patterns in the Black Sea. Prog. Oceanogr., 39, 109-151.
Sur, H.I., E., Ozsoy, U., Unluata, 1994. Boundary current instabilities, upwelling,
shelf mixing and eutrophication processes in the Black Sea. Progr. Oceanogr., 33,
249-302.
Sur, H.I., E., Ozsoy, Y.P., Ilyin, U., Unluata, 1996. Coastal/deep ocean interactions in
the Black Sea and their ecological/environmental impacts. J. Marine Systems, 7,
293-320, 1996.
The ARGO Science Team 2000. Report on the Argo Science Team second meeting
(AST-2) March 7-9, 2000, Southampton Oceanography Centre, Southampton, U.K.
Whitehead J.A., Korotaev G.K., Bulgakov S.A. Convective circulation in mesoscale
abbyssal basins. 1998 Geophys. and Astrophys. Fluid Dynamics Vol. 89 pp169-203.
Zatsepin, A.G., A.I. Ginzburg, A.G. Kostianoy, V.V. Kremenitskiy, V.G. Krivosheya,
S.V. Stanichny, P-M. Poulain, 2003. Observations of Black Sea mesoscale eddies
and associated horizontal mixing. J. Geophys. Res., 108, 3246,
doi:10.19/2002JC001390.
11
Figure Captions
Figure 1. The schematic diagram of the upper layer circulation field with its
recurrent/quasi-persistent features.
Figure 2. The mean September 2002 sea level height anomaly field and
geostrophically computed currents. The points A and B denote the deployment sites of
the floats.
Figure3. Trajectories of the profiling floats in the Black Sea for (a) the float BS634 at
200m, (b) the float BS631 at 750m, and (c) the float BS587 at 1550m parking depths.
The dots and numbers represent the position of float during a particular month.
Figure 4. Sea level anomaly maps obtained by merging of various altimeter data
during (a) 05 February 2003, (b) 22 March 2003, (c) 07 June 2003, (d) 23 August
2003, (e) 01 October 2003, (f) 15 November 2003, (g) 31 December 2003, (h) 28
February 2004.05.10.
Figure 5. Temporal evoluton of the velocity components at the depth (a) 200m, (b)
750m, (c)1550m. Solid line represents the zonal component, and the dash line
represents the meridional component.
Figure 6. The scatter plot of the surface velocity and the velocity on the depth 200m.
for the (a) zonal component; (b) meridional component.
Figure 7. Temporal evolution of the currents at the surface (dash line) and at the
depth of 200m (solid line) along the trajectory of the float number 634 for the (a)
zonal component; (b) meridional component. The surface velocity is divided by ten.
12