Oguz, T. (2005b) "Black Sea ecosystem response to climatic variations". Oceanography, 18(2), 122-133.

Oguz, T., Z. Kaymaz, J. W. Dippner (2005) "Climatic Regulation of the Black Sea hydro-meteorological and ecological properties at interannual-to-decadal time scales". Submitted to J. Marine Systems.

Decadal oscillations:

    Our recent analysis of the basin-averaged, winter (December-March) mean hydro-meteorological and ecological data in the Black Sea indicated synchronous oscillations with ~8-to-10 year periodicity. These oscillations are presented here in a composite form by means of three standardized indices (Fig. 1). The atmospheric index (ATI) is constructed by the averages of air temperature, sea level pressure, wind stress magnitude, evaporation minus precipitation time series. The physical climatic index (PCI) represents the averages of sea surface temperature, subsurface cold intermediate layer temperature, 0-30 m layer average salinity, sea level anomaly variations.  Phytoplankton and mesozooplankton biomass, surface chlorophyll concentration, hydrogen sulphite concentration at 150m, and secchi disk depth variations constitute the ecosystem index (ECOI). These indices are standardized by expressing each data set at an interval between zero and one.

Figure 1.Temporal variations of the atmospheric index, ATI, (squares), marine physical climate index, PCI, (dots) and ecological index, ECOI, (stars) during the 1960-1999 period.

    Highly significant correlation of the physical climate index with the atmospheric index (r=0.82, significant at p<0.05) suggests prompt response of the water column physical structure to atmospheric forcing (Fig. 2a).  Similarly, the ecosystem index (ECOI) responds favourably to the marine physical climate index (Fig. 2b) with correlation coefficient r=0.78 (p<0.05). A major implication of such high correlations is synchronous regulation of the Black Sea atmosphere as well as water column physical and biogeochemical structures by some climatic teleconnection patterns. This climatic signal appears to be a highly robust feature of the Black Sea, and it persists in each individual data set even after they are averaged over the basin and over a year/season.

Figure 2. Scatter plots of (a) the atmospheric index (ATI) versus marine physical climate index (PCI), and (b) marine physical climate index versus ecosystem index (ECOI). The presence of high correlation between these three indices suggest an integrated, synchronous response of the Black Sea hydrometeorology and ecosystem to large scale quasi-persistent atmospheric circulation systems.

Teleconnection patterns:  

    It has long been recognized that interannual-to-interdecadal fluctuations in the sea surface temperature and biological properties of the North Atlantic as well as in the adjacent European shelf and marginal seas are connected to the North Atlantic Oscillation (NAO), known to be the dominant mode of variability of the northern hemisphere atmospheric circulation. The NAO refers to a meridional shift in atmospheric mass between the subpolar ‘Iceland low’ and subtropical ‘Azore high’ pressure systems. For the 40 year data (1960-2000), the correlation between the NAO and the Black Sea ATI is 0.45, statistically significant at the 99.9% confidence level (Fig. 3a).

Figure 3. Scatter plot of the Black Sea atmospheric index, ATI, versus (a) the NAO index, (b) the modified NAO index when it was modified by the EAWR index for certain periods.

    Except for the periods of 1961-1965 and 1997-1999, the time series of the Black Sea ATI (shown by red line in Fig. 5) and the NAO index (shown by in black line in Fig. 5) show quasi-synchronous variations. The positive NAO phases are associated with the lower values of ATI indicating colder air temperatures (values closer to zero in Fig. 4), greater evaporation minus precipitation rates (values closer to one in Fig. 4) and higher sea level pressure (values closer to one in Fig. 4). The Black Sea atmosphere is thus driven remotely by strong cold and dry northwesterlies developed in response to anomalous high heat loss of the subpolar North Atlantic and anomalous heat gain in the subtropical North Atlantic. Conversely, lower ATI values coincide with the negative NAO phases associated with weaker meridional sea level pressure gradient between Azore high and Iceland low and southwesterlies giving rise to warmer air temperatures, milder and less dry/more wet atmospheric conditions in the region.

Figure 4. Temporal distributions of winter (December-March)-mean atmospheric temperature (oC) (dots), AT, measured at the meteorological station near the Kerch Strait along the north coast of the Black Sea, basin-averaged evaporation minus precipitation, E-P, (triangles), sea level atmospheric pressure, SAP, (squares), and the atmospheric index (thick line) computed as an average of the atmospheric time series data. The E-P and SAP data are obtained by ECMWF Re-analysis data set in which high frequency oscillations have been filtered by the three point moving average. Their axis, on the right hand side of the plot, is inverted to display in phase variations with that of AT.

Figure 5. Temporal variations of December-January-February mean NAO index (black line), its modified version (yellow line) when the index values for 1960-1965 period are replaced by those provided by the EAWR index, and the Black Sea atmospheric index, ATI (red line).  The original NAO index is based on the difference of normalized sea level pressures between Ponta Delgada, Azores and Stykkisholmur/ Reykjavik, Iceland. It is provided at the URL site: http://www.cgd.ucar.edu/~jhurrell/nao.stat.other.html. The EAWR index data are given at http://www.cpc.ncep.noaa.gov/data/teledoc/eawruss.html. The NAO index time series is filter by the Gaussian filter using the 5 year window size.

    The negative correlation between decreasing ATI values (i.e. cold and dry winters) and negative NAO index values (i.e. wet and mild winters) during 1961-1965 period suggests either (i) occasionally inadequate representation of the NAO atmospheric circulation system by means of two local sea level pressure measurements at Azore and Iceland, or (ii) modulation of the NAO signature by regional quasi-persistent atmospheric systems over the Eurasia continent, or (iii) combination of these factors. For example, even though the winter 1964 sea level pressure distribution (Fig. 6) resemble those of typical NAO positive winters, because the subpolar low pressure center of action is located at considerable distance to the southwest of Iceland (centered at 40-45oW, 55-60oN), the surface pressure difference between Iceland and Azore does not represent adequately the strength of this dipole system and is in fact expressed by a negative NAO index value.

Figure 6. The mean sea surface pressure distribution pattern for winter (December-January-February) period of 1963-1964.  The NAO dipole system is characterized by the Iceland low (L) and Azore high (H) pressure centers over the North Atlantic. The EAWR dipole system is characterized by the high pressure cell (H) over the northwestern Europe and a low pressure cell (L) over the Caspian region. The region in red colour near the eastern boundary of the plot represent Siberian high pressure system.

    Examining 41-year (1958-1998) December-January-February mean precipitation patterns of the Eastern Mediterranean, Krichak et al. (2002) noted modulation of the NAO-based climatic teleconnection patterns by the so-called East Atlantic/West Russia (EAWR) quasi-persistent atmospheric system. They concluded that the NAO system alone can not adequately describe climate variability of the Eastern Mediterranean.

    The EAWR system is characterized by two poles with either negative or positive surface pressure anomaly centers located over the northwestern Europe and the Caspian region. The positive winter EAWR pattern involves decreasing surface pressures from west to east with a low pressure anomaly center over the Caspian Sea region and a high pressure anomaly center over the northwestern Europe. The Black Sea-Eastern Mediterranean system is then affected by arctic air masses providing colder and drier than normal conditions, similar to those observed in the positive NAO years. The alternative EAWR winter pattern involves a positive surface pressure anomaly center over the Caspian region and a negative anomaly center in the northwestern Europe. The Black Sea is then accompanied by warmer and wetter than normal conditions as in the case of the negative NAO index.  

    The cooling cycle defined by decreasing ATI values during the first half of the 1960s fits into EAWR >0 phase as shown in Fig. 6, and quantified by the positive EAWR index values provided at the URL site http://www.cpc.ncep.noaa.gov/data/teledoc/eawruss.html. If the original NAO index is modified by replacing its values with those of the EAWR index for this period, the “modified NAO index” time series (shown by yellow line in Fig. 5) reveal more synchronous variations with the ATI time series. Their correlation (significant at 99.9% confidence level) then increases from 0.45 to 0.59 (Fig. 3b).