The new CTD data from the R/V Akvanavt 2002 cruise presented and discussed in this study, show that lateral limits of the Bosphorus lens penetration into the Black Sea interior are substantially wider than the earlier observed ones. The data reveal unexpectedly thick warm lenses in the 150–500 m depth range within the eastern gyre of the Black Sea. Thus, we provide observational evidence that the Bosphorus warm water penetration into the sea interior is not a local phenomenon confined to the southwestern and central parts of the western gyre (the intrusions have been previously observed only there), but rather a fundamental basin-wide process determining the intermediate layer structure in the Black Sea, as was recently assumed on the basis of the numerical modeling results.
 The Black Sea is an excellent natural laboratory for studying high-density water injection into the less dense ambient water body. The injection source is the lower current of the Bosphorus Strait, which supplies the waters of Mediterranean origin.
 It is commonly accepted that very intense mixing of warm and saline Mediterranean Sea Water (MSW, ∼15°C, 36 psu) with cooler and fresher Cold Intermediate Water (CIW, 6.5–8°C, 18.5–21 psu) over the shelf just north of the Bosphorus Strait [Tolmazin, 1985] leads to the formation of different density plumes [Ozsoy et al., 1993; Stashchuk and Hutter, 2003; Stanev et al., 2001]. (The terms “Bosphorus plume” or “plume” are settled in a literature. These terms mean the thin intrusions of Mediterranean origin characterized by anomalous temperature (oxygen content and other characteristics) which propagate in the Black Sea western gyre interior [Stashchuk and Hutter, 2003; Konovalov et al., 2003; Stanev et al., 2001]. Here we also use the term ‘plume’ (simultaneously with terms “intrusion”, “lens”) for the description of a similar phenomenon in the eastern gyre interior in spite of significant thickness of the newly revealed intrusions.) There is still, however, large uncertainty concerning the thermohaline and chemical characteristics of these intrusions and the depths of injections of modified MSW into the interior part of the sea. The lateral limits of the intrusion are also unclear. In this paper, we provide observational evidence that the warm MSW-CIW lenses, propagating within the different density intervals, can reach as far east as the central part of the eastern gyre, thus affecting the vertical thermal structure and feeding the intermediate layer of increased temperature there.
 Instrumental observations have confirmed, however, the appearance of the MSW-CIW plumes only in the southwestern and central parts of the western gyre [Ozsoy et al., 1993; Konovalov et al., 2003; Glazer et al., 2006a, 2006b]. The most distant point where the presence of the MSW intrusions has been observed, is the station (∼42°30′N, 30°45′E) located ∼200 km NE of the Bosphorus Strait [Konovalov et al., 2003]. At present, no evidence for penetration of the Bosphorus plumes/lenses into the eastern gyre is found.
 To date, only a few studies base discussions of the thermohaline structure of the Black Sea intermediate layer on modern CTD data. To describe the main features of potential temperature (θ) and salinity (S) vertical distributions, Murray et al. , Top et al. , Konovalov et al.  and Glazer et al. [2006a] use data from individual stations located in the western gyre interior. Analyses of the θ–S structure of the Black Sea water column provided by Mamaev  and Ivanov and Samodurov  are based on basin-averaged θ–S profiles and model calculation results. In this paper, we present and discuss the CTD data from the 186-nm long hydrographic section spanning the Black Sea eastern gyre.
 This section was carried by the R/V Akvanavt in June 2002 in the eastern part of the Black Sea from (42°27′N, 39°5′E) to (43°51′N, 35°14′E). The station spacing was 15.5 nm. Vertical profiles of θ and S were acquired from the sea surface to ∼20 m above the bottom with a Sea-Bird CTD profiler at each of the 13 stations. The station locations are shown in Figure 1a. The precision of measurements was 0.002°C for θ and 0.003 psu for S. Distribution of θ along the section is shown in Figure 1b; the θ–S diagram and the vertical profiles of θ and S are given in Figure 2.
 Additionally, we use the data from eight CTD casts at the station located at (42°30′N, 30°45′E) near the center of the western gyre in order to characterize the different classes of MSW-CIW intrusions. These data were obtained using a Sea-Bird profiler on board the R/V Knorr in the early summer of 2001. The θ–S plot and vertical profiles of θ and S based on the 2001 data are shown in Figure 3.
4. Results and Discussion
 The relatively warm MSW-CIW lenses in the eastern Black Sea below the main pycnocline (150 m) are revealed in the 2002 section data (Figure 1b). Hereafter, we discuss the characteristics of the observed intrusions and the influence of the associated temperature anomalies on the thermohaline structure of the eastern gyre.
 The distinct θ anomalies are seen in the 150–500-m layer in the northwestern part of the section at stations 1327–1330 (Figure 1b). At sta. 1330, the warm lens is observed immediately below the pycnocline at ∼200 m (σθ = 16.55–16.7 kg/m3, where σθ is potential density referenced to 0 dbar). Within this intrusion, the maximum θ is 0.06°C higher than in the ambient waters (Figure 2). At the other stations (1327, 1328, 1329), the MSW-CIW lenses are deeper, and less pronounced. At sta. 1329, two warm intrusions are revealed at ∼280–320 m (σθ = 16.77–16.87 kg/m3) and 400–500 m (σθ = 16.95–17.02 kg/m3). Relative temperature exceeding in these lenses is 0.015–0.02°C (Figure 2). The warm plume (with θ excess of ∼0.015°C) is also found at sta. 1328 at ∼300 m (σθ = 16.8–16.85 kg/m3, not seen in Figure 1b). A distinct thick lens (θ > 8.83°C in the core) is observed at sta. 1327 in the depth range of 300–500 m (σθ = 16.83–17.03 kg/m3). In the southeastern part of the section, at stas. 1318–1326, no intrusions are detected.
 It should be noted that the thickest lenses at 400–500 m (Figure 1b) are separated by a single vertical profile (sta. 1328) and appear to be individual objects. Due to the data limitations, we cannot claim that these lenses are individual. Likely, there is one large warm intrusion that is intersected by the 2002 section at the two points.
 There are no significant σθ anomalies associated with the plumes/lenses in the section (not shown). Consequently, positive θ anomalies in the lenses have to be compensated for positive S anomalies with respect to density. The θ–S plots (Figures 2a and 3a) show, however, that the observed distinct θ anomalies are not accompanied by any significant S extrema. Even a detailed examination of the vertical S profiles at stas. 1327–1330 does not allow to distinguish the MSW-CIW intrusions from the salinity data. The formal absence of S signatures in the lenses is most likely explained by comparing the magnitude of probable S maxima to the measurement accuracy (±0.002 psu). For instance, θ in the lens core at 420 m (sta. 1329) is 0.015°C higher than in the ambient waters. To maintain of equilibrium (density compensation), S should also reach the local maximum within the lens; calculations show that the positive S anomaly in this case would only be of ∼0.002 psu to compensate the θ maximum.
 In general, the distortions of the vertical salinity gradient associated with the observed MSW intrusions are also insignificant. They are distinct, however, in the warmest intrusion (∼200 m, sta. 1330), where the minimum of the S gradient is revealed. In the core of this plume, the gradient is ∼0.002 psu/m, whereas the one at the adjacent stations is ∼0.004 psu/m (Figure 4).
 Why doesn't the salinity in the observed lenses significantly differ from the salinity of ambient intermediate waters? The vertical density stratification of the Black Sea water column strongly depends on the salinity one, more so than temperature [Murray et al., 1991; Ivanov and Samodurov, 2001], which only influences the stratification by 5–7% (at given θ/S/pressure). Therefore, density equilibrium between a lens and ambient waters can only be reached at depth where the intrusion's salinity does not significantly differ from that of the ambient waters. Upon reaching this depth (density/salinity level), a lens travels away from the formation region, and the θ anomaly gradually diminishes via heat exchange with the ambient waters.
 In order to compare characteristics of the warm MSW-CIW intrusions in the eastern and western gyres of the Black Sea we examined the 2001 data obtained by R/V Knorr in the western part of the sea. Looking for the plumes/lenses at depths of more than 200 m, we used the data from 8 casts of the CTD station located at (42°30′N, 30°45′E) near the center of the western gyre. The θ–S plot based on these data (Figure 3a) shows the significant θ maximum in the 16.55 < σθ < 16.75 kg/m3 range (∼180–250 m). The magnitude of this maximum is only ∼0.02°C, which is three times less than the θ anomaly observed in the eastern gyre at the same depths in 2002 (Figure 2).
 Furthermore we would like to note that the thickness of the observed intrusions in the eastern part of the sea in 2002 is significantly larger than the thickness of the plumes found in the western gyre in 2001.
 At the sta. 1329 of the R/V Akvanavt 2002 section, the lens core with θ > 8.83°C has a thickness of ∼40 m (pressure range of 415–454 dbar, Figure 1b). At sta. 1327, the lens core with the same θ has a thickness of ∼80 m (422–506 dbar), and the θ anomaly associated with this intrusion affects the ∼210-m-thick layer (∼300–510 dbar). Thus, unlike the previous observations in the western part of the sea, which have revealed the thin (<10 m) plume-fingers [Konovalov et al., 2003; Glazer et al., 2006b], our 2002 data show that MSW-CIW intrusion thicknesses can reach tens of meters.
 According to the 2002 and 2001 data, the MSW-CIW lenses in the central Black Sea are warmer than the surrounding waters, whereas the near-Bosphorus observations [Ozsoy et al., 1993; Basturk et al., 1998; Glazer et al., 2006b] show relatively cold intrusions within the 100–500-m layer.
 There is no single explanation for the coexistence of warm and cool intrusions in the Black Sea water column. Mathematical modelling provided by Stashchuk and Hutter  have shown that θ inside the MSW-CIW plumes is mainly determined by the θ–S properties of the incoming MSW. Glazer et al. [2006b] have assumed that positive or negative θ anomalies inside the MSW-CIW intrusions are result of seasonal variations of the CIW properties. Likely, formation of warm or cold intrusions results from a complicated season-dependent superposition of these two factors.
 Several studies [Ozsoy et al., 1993; Ozsoy et al., 2001; Stanev et al., 2001; Stanev et al., 2004; Stashchuk and Hutter, 2003] describe the depth limit of the MSW-CIW intrusions. According to these studies the mixing of the MSW and CIW reduces salinity and density of the inflowing waters, and the majority of the MSW-CIW plumes finally reaches a neutral buoyancy state within the intermediate layer at depths of less than 500–600 m. Stanev et al.  examine the paths of passive tracers penetrating the Black Sea intermediate and deep layers and conclude that plumes do not ventilate at depths below 600 m. Our data do not contradict the latter conclusions, as we only found the MSW intrusions at depths of less than 500–600 m.
 According to the numerical simulations [Stanev et al., 2001], spreading of the MSW-CIW intrusions should occur at ∼200–400 M. The 2002 data show, however, that this range is substantially wider, at 170–500 m.
 The intermediate layer of increased θ permanently exists in the Black Sea at ∼450–500 m above the quasi-isothermal layer (∼500–700 m, Figures 2b and 3b) [see Murray et al., 1991]. Based on numerical simulations, lateral intrusions of the MSW-CIW waters are suggested to be the only source of heat at depths of 100–500 m, taking into account that the geothermal heat flux is negligible there [Ivanov and Samodurov, 2001; Konovalov et al., 2003]. According to the presented 2002 data, the propagation of substantially modified warm Mediterranean waters occurs in the eastern gyre over the quasi-isothermal layer; the lower limit of MSW-CIW lens penetration is ∼500 m (Figure 2b). (This liminal depth has been repeatedly mentioned in a literature as the most probable or maximum depth of MSW-CIW plumes propagation [Stashchuk and Hutter, 2003; Ozsoy et al., 2002; Ozsoy et al., 2001]). Thus, the presented data confirm model results, according to which the MSW-CIW plume injections maintain the increased θ at depths of 100–500 m, and thus support the existence of the layer of increased θ at depths of 450–500 m in the Black Sea interior.
 In our opinion, one of the interesting questions is regarding the pathway of the lenses observed in the center of the eastern gyre of the Black Sea. According to circulation scheme provided by Stanev [2005, Figure 2] the main gyre of the eastern part of the sea is located between 34.5° and 36.5°E and between 43.3° and 44.3°N. In 2002, we observed the warm Bosphorus lenses within this eastern gyre (stas. 1330–1327, Figure 1). The eastern gyre is connected to the western gyre of the Black Sea by the relatively short current paths. It is also known that circulation at intermediate and deep layers repeats the main features of near-surface circulation [Korotaev et al., 2006; Stanev, 2005]. Thus, propagation of MSW-CIW plumes into the eastern gyre is possible. However, we believe that a dynamical aspect of the MSW-CIW lens spreading should undergo further detailed investigation based on a large volume of basin-wide hydrographic and drifter data.
 The observational evidence that the warm MSW–CIW lenses can reach as far as the central part of the eastern gyre is presented. Unlike the preceding works based on individual stations within western part of the sea, we used the data from the complete section occupied in the eastern part of the Black Sea. The MSW-CIW intrusions observed in the eastern gyre at depths of 150–500 m are characterized by more pronounced temperature extrema than those in plumes within the western gyre. The thickness of warm lenses in the eastern gyre interior is also much greater than the thickness of the previously observed intrusions in the western part of the sea.
 Overall, the new hydrographic data discussed in this study provide a telling argument, confirming the assumption, that MSW-CIW plume penetration into the sea interior is not a local phenomenon confined to the southwestern and central parts of the western gyre, but rather a fundamental basin-wide process determining the intermediate layer structure in the Black Sea.
 We thank N. Donoho, K. Egan and the two anonymous reviewers for suggesting numerous improvements to the current manuscript.