Draft Report of the

AEGEAN SEA WORKSHOP

Rhodes, Greece, October 8-10, 2002

 

Sponsored by the Office of Naval Research

and Office of Naval Research International Field Office

 

Report prepared by the Workshop Organizing Committee:

Sarantis  Sofianos ⁽¹⁾

William Johns ⁽²⁾

Alexandros Lascaratos ⁽¹⁾

Stephen Murray ⁽³⁾

Donald Olson ⁽²⁾

Alexander Theocharis ⁽⁴⁾

 

⁽¹⁾ Department of Applied Physics

University of Athens

University Campus, Build PHYS-5

Athens, 15784, Greece

 

⁽²⁾ Rosenstiel School of Marine and Atmospheric Science

University of Miami

4600 Rickenbacker Causeway

Miami, FL 33149, USA

 

⁽³⁾ Office of Naval Research

800 N Quincy Street

Arlington, VA 22217-5660, USA

 

⁽⁴⁾ National Center for Marine Research

Institute of Oceanography

P.O. Box 712

Anabyssos-Attiki, 19013, Greece

Contents

1. Introduction

2. Scientific Background

3. Important Scientific Questions

4. Recommendations for Future Work

5. Acknowledgments

Reference/Appendices

1. Introduction

A workshop was held, on October, 8-10, 2002, at the National Center for Marine Research / Hydrobiological Station of Rhodes, Greece, on the topic "Aegean Sea", sponsored by the U.S. Office of Naval Research and the Office of Naval Research International Field Office.

The workshop focused on primarily on the physical oceanography and air-sea interaction of the Aegean Sea but also considered biophysical interactions in the region. The central objectives of the workshop can be summarized as:

a.       Discuss the present level of understanding of the physical oceanography and related disciplines of the Aegean Sea. Exchange information between scientists with experience in the region and/or similar cases, which can help to summarize the known facts and the outstanding questions of the dominant mechanisms governing the Aegean Sea.

b.      Identify important topics for future study in the Aegean Sea on different time scales (synoptic, seasonal and climatic) and possible approaches (observations and modeling). Establish collaborations between scientists and institutes to help the planning of collaborative international projects.

Over thirty scientists from four countries (Greece, USA, Turkey and Germany) participated in the workshop, including theoreticians, observationalists, and numerical modelers (Appendix 1). During the first day of the workshop thirteen invited talks were presented on various topics of the dynamics governing the Aegean Sea circulation, exchange with adjacent basins, air-sea interaction, as well as the biophysical characteristics of the region (Appendix 2). Following the scientific presentations the workshop was organized in three working group sessions. The central theme of Session 1 was the review of the dynamics and circulation of the Aegean Sea, and the participants broke into two groups focusing on: (A.1) short to seasonal time scales and (A.2) interannual/climatic variability. In Session 2, the discussion focused on the priorities for future research, and the participants broke into the same two groups: (B.1) short to seasonal time scales and (B.2) interannual/climatic variability). Finally, the third working group session dealt with forming recommendations for future work and suggesting possible approaches to this research, in two working groups: (C.1) observational and (C.2) modeling approaches. All these groups were not sharply separated, and there was considerable overlap within each group as well as interaction between the groups (in the plenary sessions and by participation in different groups).

This report contains a summary of the discussion that took place during the three days of the Aegean Sea Workshop. Section 2 highlights the principle features of the Aegean Sea circulation and exchange with the atmosphere and the adjacent basins, in various time scales.  Results from recent experiments (not yet published) are also briefly presented. Section 3 presents the most important scientific questions concerning the dynamics of region, as were summarized by the group discussion sessions, followed by a summary and recommendations for future research in Section 4.

Figure 1. Major islands and basins in the Aegean Sea region

2. Scientific Background

2.1 Geographical Characteristics

The Aegean Sea is one of the four major basins of the Eastern Mediterranean Sea, covering an area of 240,000 km². It is situated in the northeastern Mediterranean, to the northeast of the Ionian Sea and to northwest of the Levantine Sea, bounded to the north and west by the Greek mainland, to the east by the Turkish coasts and to the south by the islands of the Cretan Arc. The Aegean sea is connected with the Marmara and Black Seas through the strait of Dardanelles (average width 3.5 km; sill depth 70 m). At its southern end it is connected with the Levantine and Ionian Seas through a series of six straits (the Cretan Arc Straits), including (from east to west): the Rhodes Strait (width 17 km; sill depth 350 m), the Karpathos Strait (width 43 km; sill depth 850 m), the Kassos Strait (width 67 km; sill depth 1000 m), the Antikithira Strait (width 32 km; sill depth 700 m), the Kithira Strait (width 33 km; sill depth 160 m), and the Elafonissos Strait (width 11 km; sill depth 180 m).

The Aegean Sea coastline is very irregular and its topographic structure very complicated. There are over 3,000 islands of various sizes scattered all over the basin. Three deep basins exist: the North Aegean Trough oriented in the WSW-ENE direction (including the N. Sporades, Athos, Limnos and Saros basins) in the northern Aegean with maximum depths up to 1,500 m, the Chios basin in the central Aegean with maximum depth of 1,100 m, and the Cretan basin in the southern Aegean, which is the largest and deepest one, with maximum depths of 2,500 m. The Chios basin is bounded to the south by the extensive Cyclades plateau and sills (with depths that do not exceed the 350 m), which is identified in the literature as the limit between the north and south Aegean.

2.2 Atmospheric Forcing

The wind field over the Aegean Sea is strong and its direction is largely controlled by orography. The prevailing winds, on the annual mean basis, are of the northern sector. The seasonal wind stress distribution is presented in Fig 1 (May, 1982). The seasonal cycle of the wind field is associated with the strength and the persistence of the northerly winds and has two maxima, one during the winter (December - February) and a second one during summer (July - August). During winter the wind is primarily from the north (during this season, the Aegean Sea occasionally experiences strong southerlies, (Katsoulis, 1970)) bringing cold and dry air through the various river valleys of the Balkan peninsula. During summer the wind field is dominated by the Etesian winds, which are northerlies, dry, and can reach gale force. Moderate diurnal sea breezes are present along the coasts during summer.

The evaporation rate over the Aegean Sea is around 1.3 - 1.5 m/year (da Silva et al., 1994; Jakovides et al., 1989) with maximum values during February and August. The annual mean net heat flux is estimated at 26 W/m² (Poulos et al., 1997), which implies that the Aegean Sea is, on the average, losing heat through its surface. This loss is balanced by advection of the warm Levantine waters through the Cretan Arc straits (while the heat exchange with the Black Sea is negligible (Tolmazin, 1985)).

The Aegean Sea receives the colder and fresher Black Sea Water (BSW) through the Bosphorus, Marmara Sea and the Dardanelles strait. Below that layer there exists an outflow of the more saline Aegean Sea waters. The BSW has a rate of 1,257 km³/year while the Aegean Sea waters a rate of 957 km³/year (Unluata et al., 1990). The Aegean Sea also receives the freshwater outflows from various rivers along the Greek and Turkish coasts. The annual mean evaporation rate (E) exceeds the sum of precipitation (P) and river runoff (R) (0.5 m/year and 0.11 m/year, respectively) (Poulos et al., 1997). However, if the influence of the BSW is taken into account, then the water balance of the Aegean Sea is negative (E-P-R-BSW<0). The excess volume of water per unit area has been calculated between 0.7 and 1.4 m/year (Bethoux and Gentili, 1994; Poulos et al., 1997). The water budget of the region is compensated by the advection of waters through the Cretan Arc straits.

2.3 Circulation and water mass formation

Despite the progress in direct observations and modeling efforts during the last two decades (many of these studies were presented during the presentation session of the workshop, see Appendix 2.), the circulation of the Aegean Sea is yet far from being well defined and understood. The circulation pattern that emerges from observations is rather complex and variable. This is due to many factors, such as the distribution of the various island chains and straits, the irregular bottom topography, the seasonal variability of the atmospheric forcing, the presence of strong meteorological events that may alter the local circulation patterns, and the presence of many different water masses.

If we attempt to summarize the known circulation characteristics (from various cruises covering parts of the region, recent drifter deployment, and modeling techniques), there seems to be a general cyclonic circulation in the Aegean Sea. However, the most active dynamic features are the mesoscale cyclonic and anticyclonic eddies, which can extend to several Rossby radii of deformation (around O(10 km)). The spatial and temporal variability of these features is not really known. Some of these features appear to be permanent (i.e. the cyclonic eddy in the south Chios basin), while others have a transient character. The dynamics responsible for the formation (and decay) of these features is not clear. The wind field, the thermohaline forcing, and the interaction of these features with the very complex topography are suggested as the mechanisms involved in the generation and spatiotemporal variability some of the identified eddies.

The most characteristic feature of the circulation pattern in the north Aegean is the surface outflow of the brackish BSW from Dardanelles, which creates a front with the ambient saltier waters of Levantine origin. These waters have a general westward movement and then southward, being modulated by the presence of mesoscale eddies (Zodiatis, 1994). They cross the various passages towards the westward side of the Cyclades plateau and, after experiencing strong mixing, they appear at the western side of the Cretan Sea in the southern Aegean. The dynamics of the BSW plume and its interaction with the irregular coastline and topography (islands, bays, straits) is not well understood even though it plays an important role in the biochemical processes of the region. The basic water masses in the northern Aegean are three: the BSW, the highly saline and warm waters of Levantine origin (LSW and LIW), and the very dense deep waters that fills the bottom of the various sub-basins. These deep waters are formed locally (both shelf and open sea convection processes have been hypothesized as the mechanisms involved in the North Aegean deep water formation) with strong interannual variability (Theocharis and Georgopoulos, 1993; Zervakis et al., 2000). It should be noted that the local formation of extremely dense waters in the North Aegean takes place despite the fact that the latter is much more stratified than the South Aegean, due to the presence of the BSW layer at the surface. Several studies (Plakhin, 1971, 1972; Zervakis et al., 2000) have commented on the role as an insulator of the surface layer of BSW in the North Aegean, and on the impication it has in deep water formation processes. Thus, the variability of buoyancy input from the Dardanelles could be a significant contributor to the thermohaline circulation of the Aegean.

The southern Aegean (Cretan Sea) and the Cretan Arc straits have been more intensely investigated during recent international projects (Georgopoulos et al., 1989, 2000; Roether and Schlitzer, 1991; Theocharis et al., 1993, 1999). This investigation revealed intense mesoscale variability, including transient and/or recurrent cyclonic and anticyclonic eddies (Fig 4). The exchange between the Cretan Sea and the Levantine and Ionian basins is very complex, presenting strong seasonal and interannual variability, as well as interaction with local circulation features (Asia Minor Current, Mirtoan/West Cretan Cyclone, East Cretan Cyclone, the Ierapetra Anticyclone and the Rhodes Gyre) (Kontoyiannis et al., 1999). Direct measurements revealed a persistent deep outflow of Cretan Deep Water (CDW) (s_q>29.2) with an annual mean of ~0.6 Sv, through the Antikithira and Kassos Straits at depths below 400 m and 500 m, during 1994-95, respectively. However, during 1987-95 the CDW outflow presents large variability with peak values exceeding 1 Sv (Theocharis et al., 1999). The typical water mass structure of the Cretan Sea (until the dramatic change of the late 1980s, see below), included in the upper layer the saline surface waters of Levantine origin (LSW; S~39 psu) and the less saline (S<38.9) surface waters of Black Sea origin (BSW).  The inflow of LSW seems to occur mostly during the warm period of the year and enters the SE Aegean through the eastern Cretan Arc Straits,  while the BSW comes from the North Aegean and affects mainly the Mirtoan basin and western Cretan Sea. Frequently there are also intrusions through the Cretan Arc Straits of the less saline subsurface water of Atlantic origin, the so-called Atlantic Water (AW) originating from the Ionian Sea. Below the upper layer and down to the bottom, two water masses were distinguished with similar characteristics, but slightly denser (σ_θ up to 29.16) than those of the Levantine Intermediate Water (LIW). This relatively dense Cretan water was observed to overflow occasionally through the Cretan straits and contribute to the water masses below the LIW in the Eastern Mediterranean (Schlitzer et al., 1993; Malanotte-Rizzoli et al., 1997).

Observations after the early 1990s revealed a dramatically different structure of the Cretan Sea water column. Extremely dense (σ_θ > 29.2) and very saline (S >39 psu) water, of local origin, started filling the deep Cretan basin and overflowing through the sills of the Cretan Arc straits (Roether et al., 1996; Kontoyiannis et al., 1999; Theocharis et al., 1999; Theocharis et al., 1999a). Due to its high density, the Cretan Deep Water (CDW) displaced water from the deepest parts of the Levantine and Ionian basins of the Eastern Mediterranean becoming a major contributor of deep and bottom water in the Eastern Mediterranean. So, the present state of the Cretan Sea includes three distinctive water masses in the intermediate and deep layers: (i) the Cretan Intermediate Water (CIW), formed locally and located between 50 and 250 dbar, with physical characteristics similar to those of the LIW, but still warmer and slightly saltier, (ii) the so-called Transitional Mediterranean Water (TMW), a mixture between the LIW and the Eastern Mediterranean Deep Water of Adriatic origin found a depths up to 700 dbar, coming from the mid-depths of the Eastern Mediterranean through the eastern and western Cretan Arc Straits, and, (iii) the CDW in the deepest/ bottom parts of the basin (800-2500 dbar), formed by open sea convection and/or in the surrounding shelf areas (Theocharis et al., 1999; Lykousis et al., 2002). This structure is not static but seems to undergo considerable trends and changes. For example, the CDW volume in the Cretan Sea has declined recently, most probably due to lack of very dense water formation.

The coupling between the north and south Aegean through the Cyclades plateau (and other topographic barriers, such as the Sporades island chain) is one of the most unexplored issues of the Aegean Sea dynamics. Zervakis et al. (2000) suggested that during periods of massive dense water formation in the North Aegean the north-south exchange is greatly enhanced and the thermohaline circulation of the Aegean is accelerated. Water masses are exchanged through a very complex series of island chains and straits of different sizes. Exchange mechanisms and enhanced mixing processes that may occur in these regions have not been studied. Semi-diurnal tidal currents in the Aegean Sea are of low magnitude in comparison with the overall flow and, therefore, tidal forcing is unlikely to play a major role in the mixing processes of the region. Information on internal waves is very limited although energetic internal waves have been identified in areas that are characterized by strong flows, complex vertical structure and irregular topography. These oscillations, especially their interaction with the bottom topography, may be having significant effects on the stratification of the region (e.g. disrupt the seasonal thermocline) (Velegrakis et al., 1999; Zervakis et al., 2003).

2.4 Interannual/Climatic Variability

The importance of Aegean Sea to the thermohaline circulation of the Mediterranean Sea has long been a controversial topic (Pollack, 1951; Wust, 1961; Miller, 1963; El-Gindy and El-Din, 1986; Malanotte-Rizzoli and Hecht, 1988). Although several investigators proposed that the Aegean Sea is contributing to the Eastern Mediterranean Deep circulation, it has been generally accepted that the Adriatic Sea was the major source of Eastern Mediterranean Deep Water (EMDW). This picture changed abruptly in the early 1990s, when hydrographic surveys revealed that the circulation in the area was undergoing a major transition (Roether et al., 1996; Theocharis et al., 1992), in  which the Aegean Sea deep waters became a dominant source for EMDW. The density of the deep waters in the Cretan Sea increased dramatically and very dense water (CDW) was able to outflow and sink to the bottom of the Eastern Mediterranean Sea. It was estimated that this event started at the late 1980s (as early as 1988) and that the volume contribution of new water was at least 1 Sv, exceeding the Adriatic source by a factor of three. This event was termed the Eastern Mediterranean Transient (EMT) and has been the focus of several studies during  recent years (Roether et al., 1996; Theocharis et al., 1999; Klein et al., 1999; Lascaratos et al., 1999; Samuel et al., 1999; Malanotte-Rizzoli et al., 1999; Zervakis et al., 2000; Wu et al., 2000; Boscolo and Bryden, 2001; Tsimplis and Josey, 2001; Stratford and Haines, 2002; Tsimplis and Rixen, 2002; Theocharis et al., 2002). Although all investigators agree that climatic transient is a combined effect of temperature and salinity changes, there are several hypotheses proposed by investigators for the causes of this climatic transient. Changes in the water budget of the Aegean Sea (E-P, the role of BSW inflow from Dardanelles, etc.), the presence of strong winters in the late 1980s / early 1990s and changes in the circulation characteristics in the Levantine Sea which caused accumulation of salt in the basin are among the mechanisms proposed as the causes of the EMT.

Although the Aegean Sea experiences strong interannual variability on different time scales, the EMT is the strongest signal of climatic variability, and is most probably connected with many different dynamical aspects of the circulation, water mass formation, and air-sea interaction, including effects on the general circulation of the Mediterranean Sea. There is clear evidence that the EMT is an ongoing process (Lykousis et al., 2002), but the present and future status of the process, its effects on the general circulation of the Mediterranean Sea, and its connection with large scale climatic processes (e.g. NAO) remain obscure. Finally, although we can identify several locations where there are clear indications of dense water formation (including the North Aegean, the Mirtoan Sea, the Cyclades plateau, and the Cretan Sea), the exact locations, formation rates, mechanisms involved, and interannual variability are important points in order to understand the role of the Aegean sea as part of the general Mediterranean thermohaline system.

2.5 Biophysical coupling

The irregular coastline (gulfs, islands, straits) and the complex bottom topography together with the complex hydrology and circulation in the Aegean Sea lead to a great diversity of biotopes and strong spatial variability of the pelagic communities. The Aegean Sea can be characterized as an oligotrophic basin but one that shows strong north-south gradients in plankton biomass and productivity (Fig 5). Recent observations suggest that a microbial food web dominates in the North Aegean (in which copepods are feeding mainly on protozoa), while a multivorous food web (with copepods feeding mainly on on both protozoa and large phytoplankton) dominates in the South Aegean (Lykousis et al., 2002). The North Aegean is characterized as a special environment due to the existence of important rivers and hydrological fronts, including the influence of the BSW. This is also reflected to the higher trophic levels (fish and benthos) of the North Aegean, where a more efficient energy transfer through the pelagic food web is taking place.

Among the physical factors playing an important role in the biochemical processes of the region are the mesoscale variability of the circulation patterns, deep water formation, bottom topography, BSW inflow, and river runoff. The atmospheric forcing and its spatial and temporal variability are also important in establishing regions of coastal upwelling and downwelling, which are mostly associated with the strong Etesian winds. The temporal and spatial variability of the circulation in the Aegean Sea is reflected in the biochemical processes. The simple picture of the mean north-south gradient presented here is significantly obscured by seasonal variability (e.g. seasonal variability of the BSW outflow and river runoff) and mesoscale variability (e.g. there are also strong gradients of biochemical processes within the same basin). Recent advances in ecological modeling of the Aegean Sea (Triantafyllou, 1999) have produced important tools that can help improve our understanding of the physical and biochemical coupling of the region.

3. Important Scientific Questions

In this section we summarize the basic scientific questions posed by the workshop participants as the most interesting and/or unresolved aspects of the Aegean Sea dynamics. The questions are broken into two sections, short time scales (up to seasonal variability) and interannual time scales, following the group discussion sessions that took place during the workshop.

3.1 Short time scales

• Satellite images show changes in surface glitter caused by prevailing northerly winds blowing through gaps in the mostly high islands of the Aegean archipelago.  How do the resulting patterns of surface wind stress around the islands affect eddies, internal waves, and mixing?

• The Aegean is relatively shallow compared to ocean basins and has many sills, ridges, and islands.  In addition, tidal currents are much weaker than in the major ocean basins.  How does the internal wave field change near these bathymetric variations?  From shallow to deep water?  From islands to open water? How is the production of internal waves in a region of complex bathymetry coupled to wind forcing? 

• Owing to the shallow depths and short distances between islands internal waves may be dissipated before they propagate far, precluding establishment of a steady-state background spectrum.  Is there a background internal wave spectrum?  Does the intensity of the internal wave field vary with seasonal changes in wind stress?

• The Aegean contains strong quasi-permanent eddies or gyres with horizontal scales smaller than typical mesoscale eddies in the open ocean.  There are also several strong jets.  How do these eddies and jets modulate mixing?

• What features of the circulation are permanent versus seasonal, or of even shorter lifetime? What are the mechanisms involved in the formation and/or decay of these features (e.g. wind versus thermohaline forcing, topographic effects)? Why does there appear to be increased eddy activity in summer?

• How do circulation features, boundary currents, jets and eddies interact with straits of varying width and depth?

• Atmospheric circulation over the Aegean is neither maritime nor continental.  How does the Planetary Boundary Layer change through the archipelago?  What effects do these changes have on visibility?  On surface wind stress and buoyancy flux?

• How the wind stress pattern is affected by the presence of complex orography and what is the wind shear effect on the exchange through the various straits?

• What are the patterns and dominant processes affecting oceanic convection and restratification?  How are these processes modulated by islands, plateaux, and eddies?

• Where, at what rate, and when, are deep water masses formed in the Aegean Sea? How does mixing across the different plateaux and straits transform the water properties (e.g., what is the role of Cyclades plateau in the strong physical and biological contrast between the North and South Aegean)?

• How do the mixing patterns affect biological production?

• What is the time history of the Dardanelles outflow?  What is the shape of the Dardanelles plume?  Does the plume break into boluses or does it remain connected?

• What are the issues about underwater acoustics in the Aegean and how do they result from the above phenomena?

3.2 Interannual Variability

Although there is already a significant scientific discussion on the causes and role of EMT, there are still a large number of questions concerning the evolution of this event, as well as variability on other time scales. The discussion group focusing on the priorities of the investigation of the interannual variability in the Aegean Sea identified four basic groups of questions.

• What is the interannual variability of the Aegean Sea and how this is coupled to the variability of the whole Mediterranean Sea? Although EMT is a unique oceanic process that needs special attention, variability on other time scales and periods should also be focused in the region. Furthermore, the connection of EMT with other climatic processes may help understand its causes and evolution.

• What are the driving factors of the interannual variability in the Aegean Sea? What is the relative importance of the air-sea heat fluxes, lateral fluxes (exchange between the Aegean Sea and the adjacent basins through the Dardanelles and the Cretan Arc straits), and freshwater budget (evaporation, precipitation and river runoff)? What is the difference in the variability between the North and South Aegean and how these are coupled? It was also made clear from the group discussion that monitoring the interannual variability of the region should include a broader geographical area (e.g. Eastern Mediterranean, Black Sea) since the interaction between basins and feedback mechanisms are important in understanding the dynamics that drive variability on long time scales.

• How does the Aegean Sea respond to changes in the atmospheric forcing and interaction with the adjacent basins? How and why the water mass formation processes differ in the various sub-basins? What is the effect of long term variability on the exchange between the Red Sea, the Black Sea and the Eastern Mediterranean basins? How do extreme events (e.g. very cold winters) affect the long term variability of the Aegean Sea?

• What are the underlying dynamics of the EMT and its relaxation time scales? Are we heading toward the "old" state of deep and intermediate circulation in the Eastern Mediterranean or will the relaxation state be different?

4. Recommendations for Future Work

A strong consensus emerged between the workshop participants that the Aegean Sea can be a useful test basin for the investigation of several oceanic processes since it incorporates many different scales (several basin sizes, islands, straits, forcing mechanisms of different scales, as well as oceanic processes of different spatial and temporal scales). Both wind and thermohaline forcing is strong in the region. The tidal signal is weak in the region and cannot mask the effects of atmospheric forcing and bottom topography in regulating the mixing processes. The existence of strong atmospheric and oceanic numerical model communities with well established operational systems (Poseidon system, ALERMO/SKIRON system, NRL/NAVO system) can also facilitate the understanding  of the Aegean Sea dynamics as well as process studies in the region. Below we summarize recommendations for strategies and methods, as were discussed in the working group sessions (observational and modeling methods, approaches, and opportunities for collaboration).

4.1 Observations

Science issues can be divided loosely into three categories: (1) Long-term monitoring to quantify variability on interannual to interdecadal scales, (2) Exploratory efforts directed at obtaining a baseline picture of variability in under-observed regions and (3) Process studies focused on specific dynamical questions. Long-term monitoring efforts have already begun (e.g. buoys of the NCMR Poseidon System), though existing efforts are of limited scope, and the design of a more comprehensive, cost-efficient systems would benefit from advances in understanding produced by a intensive investigations of critical elements of the Aegean Sea, Black Sea and Eastern Mediterranean. Likewise, efforts to quantify and understand long-timescale variability, along with the findings of exploratory programs, will provide context and focus for investigations of critical dynamical questions. As an initial step, participants highlighted the importance of compiling and analyzing archived oceanographic and meteorological data to generate basic statistical measures of variability. For example, simple metrics such as the frequency, intensity and duration of strong wind events would be useful for interpreting historical data and for planning future measurement programs.

Several aspects of Aegean oceanography were identified as requiring a baseline set of observations. At long timescales, these focused on understanding interannual and interdecadal variability, on general circulation studies focused on the northern basin, and on characterization of deep circulation. Ideally, an extended (decades) measurement program would document basin-wide variability in watermass structure and circulation while attempting to understand the results in the context of variations in both local and remote forcing. Financial constraints shape strategies for achieving an extended program of in situ measurements. These efforts must balance the need to achieve broad, basin-scale scope, sampled at seasonal timescales, with the practical considerations needed to assure continuous, funded operations for periods spanning ten or more years. Discussions turned toward a suite of cost-efficient platforms and integrative measurement techniques. The possibility of continuous SOLO/PALACE float deployments, with deep parking depths designed to keep them from being carried rapidly out of the Aegean, offers a promising avenue that would benefit from the rapidly growing ARGO program. A float program could characterize deep and intermediate circulation while providing basin-wide hydrographic profiles. Surface drifters could be deployed to map surface circulation and provide distributed meteorological measurements. Although Lagrangian technologies are well suited for circulation studies, other methods must be employed to monitor key straits and passages. Sparse moored arrays, augmented with autonomous gliders executing cross-strait sections, offer a cost-efficient system for making extended measurements at the fine lateral scale necessary to resolve these flows. None of these approaches would permit the regular sampling of nutrients and tracers, which can provide unique insights into changing circulation patterns. Such sampling requires regular occupation of a well defined array of hydrographic stations. The prospect of recruiting and equipping voluntary observing vessels also offers a promising avenue for making cost-efficient, sustained measurements. Numerous ferries ply the Aegean throughout the year, many of which might be equipped with ADCPs, hull mounted temperature and salinity sensors or, perhaps, one of the low-cost moving vessel profiling systems presently under development. Such vessels also offer possible platforms for making sustained meteorological measurements (e.g. the SeaKeepers program, which places meteorological sensors aboard recreational yachts). However, there are serious difficulties that must be overcome to produce high-quality meteorological measurements from voluntary observing platforms. Accurate measurements of the full suite of fluxes are costly and require greater care that is typical of voluntary platforms.

The Aegean provides opportunities for a wide range of exploratory and process studies, many of which might proceed as a single coordinated effort. Examples of this approach include the recent ONR-funded Japan/East Sea program, which combined basin-scale float and hydrographic programs with mesoscale studies of eddy and front dynamics to advance our understanding of the dynamics governing circulation within a semi-enclosed marginal sea. At a minimum, exploratory efforts will likely be required to optimize the design of proposed process studies. Strong, episodic wind and buoyancy forcing at extremely small (O(10 km)) lateral scales characterize atmospheric forcing in several recent investigations of marginal seas. We have only limited understanding of the dynamics governing the oceanic response to forcing at scales similar to the deformation radius, though recent theoretical and modeling studies suggest interesting mechanisms for frontogenesis and subduction may be active under these conditions. The combination of intense, small-scale forcing and complex bathymetry will severely impact internal wave generation and propagation and produce spatial and temporal modulation of mixing. Simultaneous investigations of the marine boundary layer would contribute significantly toward advancing our understanding of many of these problems. Although the specific design of any process study hinges on foreknowledge of the interesting dynamics, such efforts in the Aegean will likely share the need to sample at spatial scales on the order of kilometers and time scales of days. For example, many of the straits will require fine lateral resolution, as strait widths are often larger than the deformation radii, admitting complex recirculations that must be resolved. Most of the processes discussed by the working group are clearly three-dimensional, placing additional demands on sampling methodologies. Long sampling periods (obtained from moorings, floats, drifters, long-range AUVs or extended cruises) or highly responsive sampling may be required to capture multiple realizations of response to episodic forcing events. Specific methodologies and instruments cannot be selected without first choosing a problem and designing an actual measurement program. Nonetheless, technologies that the working group found particularly intriguing included:

• HF radar for 3-D maps of surface currents. This technology seemed well suited given the Aegean’s many islands.

• Profiling moorings and ADCPs. Limited fishing activity (relative to many other marginal seas) might ease operational constraints on these platforms.

• Microstructure profiling for quantifying temporal and spatial distribution of elevated mixing.

• Towed, undulating profilers to resolve synoptic, small scale, three-dimensional variability. 

• Traditional hydrographic/ADCP/bio-optical profiling for large-scale coverage, in the context of already scheduled series of research/buoy-maintenance cruises by cost efficient vessels, that cannot be obtained using autonomous sensors.

• Floats and drifters, perhaps deployed in the context of a basin-scale monitoring program, but used to augment specific process studies.

• Long-range autonomous gliders. These could be directed in real time to sample specific features, or used to augment intensive investigations by repeatedly occupying sections of acting as profiling moorings.

All of the investigations envisaged would require combinations of these sampling technologies to adequately resolve the processes of interest. Lastly, although the work group discussed the potential value of including biological, chemical and bio-optical sampling, details were not pursued.

4.2 Modeling

Atmospheric and oceanic modeling of the Eastern Mediterranean, and especially the Aegean Sea, has a long and successful history. There are several "realistic" and "idealized" modeling studies covering parts of the Aegean Sea, the whole basin, or a larger area including the Aegean region, as well as operational/forecasting systems (Poseidon system, ALERMO/SKIRON system, NRL/NAVO system). Despite the progress, there are several limitations of the modeling systems, such as:

• The high complexity of bottom topography and geography in the area points out the need for higher resolution atmospheric modeling and generation of high-resolution forcing fields.

• There is lack of data for the Dardanelles inflow, and its interannual variability. Also, the numerical treatment of the inflow presents difficulties to be resolved.

• Subsurface measurements of velocity, temperature and salinity that can be used as model constraints and provide means for better parameterization of ocean models are very sparse in the Aegean Sea. This is more important in areas of abrupt topographic gradients (straits, islands, etc.).

• Data assimilation in the existing models (e.g. TOPEX/Poseidon data) is very difficult due to the presence of the numerous islands and islets. 

Beyond improving the existing numerical models, the Aegean Sea presents great challenges to modeling studies of various problems. The variety of circulation scales, topographic features, atmospheric and lateral forcing mechanisms introduce many interesting cases where models can help to understand the underlying dynamics while at the same time improving their own skill (e.g. parameterization of mixing and dissipation). Improved modeling of mixing processes is very important as supported by recent results (Spall, 2002, his presentation in the Aegean Sea Workshop) that suggest the basin wide circulation in semi-enclosed basins is largely affected by boundary mixing processes. Some examples of important modeling issues and processes that were discussed in the discussion group session are presented below:

• High resolution modeling studies to investigate the flow response to sudden, small scale wind events and interaction with complex topography and islands.

• Idealized and realistic modeling studies of the flow dynamics through multiple islands connected by narrow sills.

• Investigation of deep water formation in the North Aegean and Cretan Sea in order to identify permanent and occasional sources. This will help in the quantification of deep water production rates, and improve our understanding of the spreading dynamics, and mixing processes.

• Model experiments on interannual time scales to study the influence of freshwater sources on the Aegean and larger (e.g. Mediterranean) scale.

• Modeling studies of plume dynamics (BSW inflow, river runoff, etc.).

• Study the evolution of EMT and its relaxation time scales under various scenarios.

• Investigation of the coupling between atmosphere and ocean on interannual time scales.

• Modeling the Aegean Sea ecosystem dynamics in various scales. Investigate the role of Saharan dust in the region.

5. Acknowledgments

Support for the Aegean Sea Workshop was provided by the Office of Naval Research and the Office of Naval Research International Field office. The workshop was held at the Hydrobiological Station of Rhodes, and the organizing committee wishes to thank the National Center for Marine Research, the Station's director Dr. A. Sioulas and all the staff of the Station for providing a perfect environment for such a workshop. We would like to thank the working group leaders and rapporteurs: D. Olson and V. Kourafalou (A.1), W. Roether and K. Nittis (A.2), M. Gregg and H. Kontoyiannis (B.1), E. Ozsoy and A. Mantziafou (B.2), C. Lee and N. Skliris (C.1), and A. Lascaratos and J. Pullen (C.2).

Bibliography

 

Appendix 1. List of participants

 

Appendix 2. Presentations