Search

The influence of the large-scale atmospheric circulation on winter precipitation over the Mediterranean region is studied with the use of monthly 2.5° × 2.5° ERA-Interim 500 and 1000 hPa geopotential heights and GPCP-v2.3 precipitation, for the period 1980–2017. At first, Principal Component Analysis (PCA) is applied both to geopotential heights and precipitation datasets and the main modes of inter-annual variations are found. Then, these modes are investigated through Canonical Correlation Analysis (CCA) to identify the centers of action of the lower and the middle troposphere that control precipitation variability over the Mediterranean region. Three statistically significant canonical pairs are found. The first one, in the North Atlantic, controls precipitation over western Asia Minor, northern Libya and the western Iberian Peninsula. Τhe second mode corresponds to the see-saw between Greenland and central Europe and controls precipitation over the northern Mediterranean region. The third one is located north of Britain and controls precipitation over NW Africa and northern France. For each canonical pair, the mean geopotential heights and precipitation anomaly patterns for the four (10%) years with the highest and the lowest canonical scores are constructed, confirming the results of CCA. Moreover, the correlation between the canonical scores and the indices of the large-scale atmospheric oscillations, namely the North Atlantic Oscillation (NAO), the Arctic Oscillation (AO), the North Sea – Caspian Pattern (NCP) and the East Atlantic/West Russia Pattern (EA/WR), was studied.

The seasonal variability of cloudiness in the Mediterranean region is studied, through the examination of the mean day-to-day intra-annual variation of gridded NCEP/NCAR total cloud cover for the period 1948–2014. A dimensionality reduction process is followed by using a multivariate statistical methodology and three main modes of seasonal variation are found. The first mode is characterized by a winter maximum and a summer minimum and prevails mainly over the Mediterranean and Black Sea areas; a weak see-saw teleconnection over the Alps represents the opposite intra-annual variation. The second mode presents maxima in early autumn and late spring and minima in late summer and winter and prevails over the SW Europe and NW Africa inland regions. The third mode shows a maximum in June and a minimum in October and prevails over the eastern part of central Europe. The main characteristics of these modes are attributed to the seasonal variation of: i) the temperature and humidity of the lower and the middle troposphere affecting static stability and ii) the intensity and the position of the main circulation systems affecting the region, as for example the Azores subtropical anticyclone and the Mediterranean depressions. The examination of inter-decadal changes in the seasonal variation of cloud cover in the above core regions reveals a long-term decrease of cloudiness during the 66-year period 1948–2014. A comparison with contemporary satellite MODerate resolution Imaging Spectroradiometer (MODIS) and surface based European Climate Assessment & Dataset (ECA&D) station cloud cover data is carried out for the above core regions, evidencing a similarity among the data sets as to their seasonal variability, which supports the validity of the NCEP/NCAR reanalysis cloud cover results.

The spatio-temporal regime of lightning strokes over the broader Greek area is studied using 15-day ZEUS cloud-to-ground lightning data for a 10-year period (2005–14). The multivariate statistical method Factor Analysis (S-mode and T-mode) is applied on the mean intra-annual variation of lightning in order to reveal: i) the main modes of intra-annual variation of lighting activity that correspond to specific sub-regions of the Greek area and ii) the main characteristic patterns of strokes prevailing during specific periods of the year (“seasons”). According to the results there are three main modes of intra-annual variation: a) the continental one with maximum lightning activity in early summer, b) the Ionian Sea one with maximum in early autumn and c) the Aegean Sea one with maxima in middle autumn and late May. Also, according to the patterns of strokes, three main “seasons” are found: a) the summer one (early May–middle August) with highest lightning activity over the continental regions of the country due to the atmospheric instability associated with the intense land heating and the persistence of cool upper air masses, b) the winter one (middle December–middle February) with the maximum of lightning activity over the sea where instability is high due to the frequent prevalence of cold air masses over the warm sea-surface and the associated passages of Mediterranean depressions and c) the autumn one (early September–middle October) with maximum over northwestern Greece associated with the early cold season low pressure systems, the warm sea body of Ionian Sea and the essential role of topography.

An objective definition of climatologically homogeneous areas in the southern Balkans is attempted with the use of daily 0.25° × 0.25° ERA5 meteorological data of air temperature, dew point, zonal and meridional wind components, Convective Available Potential Energy, Convective Inhibition, and total cloud cover. The classification of the various grid points into climatologically homogeneous areas is carried out by applying Principal Component Analysis and K-means Cluster Analysis on the mean spatial anomaly patterns of the above parameters for the 10-year period of 2008 to 2017. According to the results, 12 climatologically homogenous areas are found. From these areas, eight are mainly over the sea and four are mainly over the land. The mean intra-annual variations of the spatial anomalies of the above parameters reveal the main climatic characteristics of these areas for the above period. These characteristics refer, for example, to how much warmer or cloudy the climate of a specific area is in a specific season relatively to the rest of the geographical domain. The continentality, the latitude, the altitude, the orientation, and the seasonal variability of the thermal and dynamic factors affecting the Mediterranean region are responsible for the climate characteristics of the 12 areas and the differences among them.

The connection between atmospheric circulation over Europe and snowfall in Ioannina, Northwestern (NW) Greece, is examined. The study deals with 160 snow events, referring to the 56-year period 1956–2011. For each of the events, the patterns of 500 and 1000 hPa geopotential height, 500 and 850 hPa air temperature, 1000–500, 1000–700 and 700–500 hPa thickness are constructed for the European area, whereas static stability K-index, relative humidity at 500, 700 and 850 hPa levels and finally relative vorticity at 500 and 1000 hPa are also used for the Southeastern (SE) Mediterranean region, for the previous day (D-1), the starting day (D) and the day following the cessation of snowfall (END). Factor analysis and cluster analysis are applied to the above datasets and the evolutions of the above circulation parameters are classified into eight clusters. Snowfall in NW Greece is generally associated with a low pressure system over the Ionian Sea combined with an anticyclone over Western or NW Europe, causing cold northwesterly flow in the lower troposphere over NW Greece. In the middle troposphere, a deep 500 hPa trough usually extends from Western Russia to Italy and the Ionian Sea. The eight atmospheric circulation structures differ mainly in the exact location and/or the intensity of the depression and/or the anticyclone, in the orientation of the 500 hPa trough axis and in the trajectories of the above synoptic systems between D-1 and END days.

The 3-dimensional structure and the evolution of atmospheric circulation favoring snowfall in Athens are examined. The study refers to 61 snow events, which occurred during the period 1958–2001. For each one of the events, the patterns of MSL pressure, 850 hPa and 500 hPa air temperatures, 500 hPa geopotential height and 1000–500 hPa thickness are constructed for the European region, for the day before (D-1), the first day (D) and the day after the end of the event (END). A statistical methodology involving Factor Analysis and Cluster Analysis is applied to the above data sets and the 61 cases are finally classified into five clusters. These clusters are generally characterized by a north-easterly flow in the lower troposphere over the Athens area. This flow is associated with the presence of a low pressure system around Cyprus and an anticyclone over Europe. The position, the intensity and the trajectories of the surface and the upper air systems during D-1, D and END days are generally different among the five clusters.

In this study the impact of the Indian summer monsoon on the large scale variability of the atmospheric circulation over the Mediterranean is investigated on an inter-annual time scale. Composite and correlation analysis results are presented, outlining different circulation patterns in the upper and lower troposphere for strong and weak monsoon years respectively. For this purpose ERA-40 Reanalysis monthly mean data at various isobaric levels together with the standardized All India Rainfall Index for boreal summer (June–July–August–September) of a 44-year period were employed. During strong monsoon years many atmospheric circulation systems appear strengthened over Eurasia, resembling a well-organized Rossby wave train over the area. In the upper troposphere a meridional shift of the jet streams over the examined area was also identified during extreme monsoon years. On the other hand, in the lower troposphere enhanced northerlies (Etesians) appear to dominate over Eastern Mediterranean along with intensified subsidence during strong monsoon years.

Cases of atmospheric circulation evolution favouring the occurrence of desert aerosol episodes (DAEs) over the broader Mediterranean region were investigated using an objective and dynamic algorithm, with daily satellite data for the period 2000–2013. After identifying strong and extreme DAEs, at a 1°×1° geographical-cell level, 255 dust aerosol episode days (DAEDs) and 148 cases of consecutive DAEDs, namely desert aerosol episode cases (DAECs), are defined. For each DAEC, the evolution of the lower tropospheric circulation 1 and 2 days before, during the initiation and after the cessation of the DAEC, is considered. S-mode factor analysis and k-means cluster analysis are applied on mean sea-level pressure and 700 hPa geopotential height fields obtained from the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) Reanalysis Project, classifying the 148 cases of atmospheric-circulation evolution into six homogeneous and discrete clusters. The mean intra-annual variation of the DAECs reveals a primary maximum in May (18.9%), and their mean annual number is equal to 11.4 DAECs. On a seasonal basis, the highest percentage of the DAECs is found in spring (51.4%). Maximum duration of the DAECs is 7 days, with 58.8% lasting 1 day. Annually, the mean monthly number of the DAEs varies from 35.8 (September) to 58.0 (April). The western parts of the Mediterranean are affected by the DAEs when cyclonic conditions prevail in the western Mediterranean and northwestern Africa. In contrast, the central and eastern parts of the study region are affected by dust storms when a low-pressure system in the central Mediterranean or central Europe and an anticyclone in the eastern Mediterranean prevail. As to the mean regional intensity (aerosol optical depth at 550 nm) the strong DAEs vary from 0.67 to 0.77, while the extremes vary from 1.14 to 2.06. Generally, strong DAEs are more frequent than extremes (in five out of six clusters).

The seasonal variability of total cloud cover in the Mediterranean region is studied, using NCEP/NCAR daily grid point data. The mean day to day intra-annual variations of total cloud cover during the hydrological year (1 October–30 September) are extracted for the 66-year period 1948–2014. A dimensionality reduction method is applied on the 324 time series and three main modes of seasonal variation of cloud cover over the Mediterranean region are found. The first one is simple, characterized by a maximum in winter and a minimum in summer, prevailing mainly over the sea. The second one is double showing maxima in early autumn and late spring and minima in late summer and winter and prevails over SW Europe and NW Africa inland regions. Finally, the third one shows a maximum in June and a minimum in October and prevails over the eastern part of central Europe. The main characteristics of these modes are attributed to the seasonal variation of the various thermal and dynamical factors affecting the region. An examination of the seasonal variability of cloud cover in the above core regions during sub-periods of the total period 1948–2014 reveals some long-term changes during the last 66 years.

The authors examine the impact of low-frequency atmospheric forcings on the air–sea heat fluxes over the Aegean Sea. The correlation between the air–sea heat flux components and three established [North Atlantic Oscillation (NAO), east Atlantic–western Russian pattern (EAWR), and North Sea–Caspian pattern (NCP)] and two testing climatic indices of potential effect over the eastern Mediterranean Sea region underlines significant discrepancies between the radiative (shortwave and longwave radiation) and the turbulent (sensible and latent heat) components. The NAO index affects the air–sea heat fluxes over the Aegean Sea region much less than the two innovative indices, the “Mediterranean index” and the “Eastern Europe index,” which play more effective roles. Moreover, the influence of the sea level atmospheric pressure (SLP) variability over an extended area (Europe and North Africa) on surface fluxes regime is investigated. The SLP anomalies are corroborated as a prominent regulating factor of the air–sea heat fluxes over the Aegean Sea region, especially during the cold season of the year. The analysis of the extreme values in the heat exchange anomalies for the period 1958–2001 highlights the role of SLP field on determining the air–sea heat fluxes regime, mainly during winter, when, occasionally, large amounts of heat loss from the sea surface trigger the mechanism of intermediate- and deep-water formation. It is suggested that wind regime and turbulent components are the modulators of the net air–sea heat flux anomalies throughout the year.

The characteristics of the atmospheric circulation over Europe and the Mediterranean associated with the formation and the dissipation of fog in Greece are examined. The data used consists of: i) 3-hourly meteorological observations recorded at 16 meteorological stations in Greece and ii) daily (00:00 UTC) 2.5×2.5 grid point values of mean sea-level pressure, 500 hPa geopotential height, 850 hPa and 500 hPa air temperatures and 1000–500 hPa thickness over Europe for the period 1957–2002. 1055 fog events are extracted from the 3-hourly meteorological observations. A specific methodology scheme including S-mode Factor Analysis and k-means Cluster Analysis is applied to the grid point data sets for the first day of a fog event (D day), the day prior to D day (D-1 day) and the day that follows the last day of a fog event (END day) and the 1055 evolutions of the atmospheric circulation associated with fog events in Greece are classified into 10 clusters. The mean patterns of MSL Pressure, 850 hPa and 500 hPa air temperatures, 1000–500 hPa thickness and 500 hPa geopotential height show that in most of the clusters, the presence of anticyclonic conditions over the Balkans, a warm front passage, or a weak, humid southerly flow induced by the presence of a shallow depression over the western Mediterranean favor fog formation in Greece, while the dissipation of fog occurs when drier air masses are transferred over the Balkans. The main differences among the 10 clusters refer to the exact position, the intensity and the specific evolution of the surface and the upper air systems, the season of their predominance and the area of the Greek territory that mainly refer to.

Some aspects on the intra-annual variability and the seasonal peculiarities of the atmospheric circulation in the Mediterranean region are examined. T-mode factor analysis is applied on the long-term (1948–2003) 5-day mean values of relative geostrophic vorticity for the 73 5-day periods of the year. The analysis is applied for sea-surface and 500-hPa levels. For sea level, two circulation types are found to be dominant: the type of the warm period and the type of the cold period, characterizing the so-called Mediterranean climate; they are strongly dependent on the thermal and humidity characteristics of the earth's surface and the position and strength of the sub-tropical (Azores) anti-cyclone. For 500-hPa level, two extra but weaker circulation types are added, prevailing mainly during the transitional seasons and modifying the smooth intra-annual variation of the atmospheric circulation. In this case, the thermal influence of the earth's surface appears to be weaker, while other circulation features, such as blocking activity and/or cut-off lows, seem to play an important role too. Copyright © 2007 Royal Meteorological Society

Based upon sea-surface temperature (SST) data, recently corrected for bucket-intake observational change, the SST fluctuations since 1873 in the Mediterranean are studied and compared with air temperature (AT) fluctuations at some land stations of this area and for the Northern Hemisphere (ATNH). Instead of the actual temperature, we used the frequency difference (per cent) of the number of the warm minus cold months, (fw—fc) annually and seasonally, smoothed using 10-year moving averages. This frequency difference was found to be a linear function of the actual temperature. For the whole Mediterranean, the trend of the SST is simple enough, with a minimum in about 1910 and a double maximum in about 1940 and 1965. A secondary minimum follows in 1975–1980, very strong in the eastern Mediterranean, with a rise after that to the present. This trend when compared with ATNH presents some differences: in the ATNH a minimum appears earlier, in 1890, but this is shown only slightly in the Mediterranean SST. The second SST maximum, in 1965, is not shown in ATNH, and the general rise of recent years starts in the ATNH about 10 years earlier. When considering the AT of the Mediterranean stations, it can be seen that the rise, after the minimum of 1910, stops 15 years earlier than the SST and ATNH. Then, an anomalous and slow decrease follows until about 1975–1980, followed by a rise to the present, which is delayed in the eastern Mediterranean.

In this work, the relationship between atmospheric circulation over Europe and precipitation in Greece is studied during high winter (January-February). Specifically, the 850 hPa relative vorticity centres, affecting or related to precipitation in the various areas of Greece, are defined and the corresponding temporal variability of both parameters is examined. Factor analysis is first applied on both fields, in order to. reduce the dimensionality of the original data sets. Canonical correlation analysis is then applied on the factor analysis results, in order to interrelate the two parameters. It is found that: (i) precipitation over western and northern Greece, the eastern Aegean islands and the western Turkish coasts is mainly controlled by vorticity over Italy and the Gulf of Genoa; (ii) southern Aegean Sea and Crete precipitation is controlled by relative vorticity west of Crete; and (iii) eastern mainland areas precipitation is significantly affected by a vorticity seesaw between Tunisia and the northern Aegean. The temporal variations of precipitation and relative vorticity parameters over the above areas show a precipitation reduction over most Greek stations, combined with enhanced anticyclonic activity over almost the whole Mediterranean during the late 1980s and the early 1990s, a period characterized by high North Atlantic oscillation index values. Finally, cluster analysis is applied on the factor scores time series of precipitation, in order to classify the winter precipitation anomaly patterns into objectively defined clusters. Six main precipitation patterns were revealed and the corresponding relative vorticity mean anomaly regimes over southern Europe indicate the atmospheric circulation characteristics causing these specific patterns. Copyright (C) 2003 Royal Meteorological Society. International Journal of Climatology

This paper presents RISKMED, a project targeted to create an Early Warning System (EWS) in case of severe or extreme weather events in the central and eastern Mediterranean and specifically in southern Italy, northwestern Greece, Malta and Cyprus. As severe or extreme weather events are considered, cases when the values of some meteorological parameters (temperature, wind, precipitation) exceed certain thresholds, and/or a severe weather phenomenon (thunderstorm, snowfall) occurs. For an accurate weather forecast, selected meteorological models have been operated daily, based on a nesting strategy using two or three domains, providing detailed forecasts over the above mentioned areas. The forecast results are further exploited for the evaluation and prediction of human discomfort and fire weather indices. Finally, sea wave models have also been operating daily over the central and eastern Mediterranean Sea. In case a severe or extreme weather event is forecasted within the next 48 or 72 h for selected target areas (sub-regions defined by their morphological and population characteristics), the local authorities and the public are informed via a user-friendly graphic system, the so-called RISK MAP. On the web page of the Project (http://www.riskmed.net), additional information is provided about the real-time values of some meteorological parameters, the latest satellite picture and the time and space distribution of lightning during the last 24 h. The RISKMED project was financed by the EU and the Ministries of National Economy of Greece, Italy, Malta and Cyprus, in the frame of INTERREG IIIB/ARCHIMED programme., peer-reviewed

In this paper, using 20 year mean monthly precipitation totals from 54 stations, the general statistics for the time and space distribution of precipitation in northwest Greece is at first described. Precipitation varies from about 1000 mm at the coast to 2500 mm on the mountains, with a large variability. Then, the intra-annual variation is studied, using Fourier analysis. This variation is described well by two harmonics, explaining about 90% of the total variance. The first harmonic, over 80% of the variance, shows a maximum from late December (coastal areas) to early January (continental areas) and is caused by the northwest Mediterranean depression activity. The second harmonic exhibits maxima in May and November, also delayed in the continental areas, and is mainly caused by upper air troughs and instability. The spatial distribution of precipitation is then studied by using factor analysis. The 12 months of a year are classified in two parts corresponding to two statistically significant factors, explaining 88.5% of the total variance. Finally, the combination of positive and negative standardized scores of the two factors showed that northwest Greece is characterized by four precipitation regimes, depending on the combination of depression activity, the effects of the sea and of topography. Copyright (C) 1999 Royal Meteorological Society. International Journal of Climatology

In this work, the divergence and convergence of the wind field over Europe, the Mediterranean and North Africa is studied for the 60-year period 1950–2009. Mean monthly values (s−1) are utilized at 273 grid points, spaced by 2.5° × 2.5°, at 10 atmospheric pressure levels from 1,000 hPa up to 100 hPa. It is found that on a seasonal basis considerable differences appear among the various European regions both near the surface of the Earth and in the upper atmosphere. During winter, near the surface, high values of convergence appear in the cyclogenesis areas and along the tracks of depressions. Specifically: in the Mediterranean, with maximum values in the Gulf of Genoa, the Black Sea and from the North Sea up to the Baltic Sea. On the contrary, over the rest of the areas divergence prevails. This situation is reversed above approximately 600 hPa with positive values (divergence) over the Mediterranean and the Seas of northern Europe and negative (convergence) over central Europe. During summer, over the seas, where air is cooler than over land, positive values are observed (divergence). This difference is more intense over lower latitudes. Similarly to winter, in the upper atmosphere, the convergence-divergence field becomes smoother changing signs at about 600 hPa.

This paper is just the three-page introduction to a Special Issue of Advances in Meteorology focusing on desert dust. It provides a paragraph each on 13 accepted papers, most relating to the used of satellite data to assess attributes or distribution of airborne desert dust. As guest Associate Editors of this issue, we organized the papers into a systematic whole, beginning with large-scale transport and seasonal behavior, then to regional dust transport, transport history, and climate impacts, first in the Mediterranean region, then India and central Asia, and finally focusing on transport model assessment and the use of lidar as a technique to constrain dust spatial-temporal distribution.

The seasonal characteristics of atmospheric static stability in the Mediterranean region are examined, for the 60-year period 1948-2007 and for the four 15-year sub-periods 1948-1962, 1963-1977, 1978-1992 and 1993-2007. S-Mode and T-Mode Factor Analysis are applied to the mean 5-day values of K static stability index over the Mediterranean region. Three dominant modes are revealed for both, the intra-annual variation and the spatial distribution of K-index. It is found that these modes are connected to the seasonal characteristics of the main atmospheric circulation systems affecting the region and the thermal properties of the Earth's surface (land or sea). The differences among the results of the four sub-periods partially reflect the inter-decadal variations of the strength of the above factors. Theoretical and Applied Climatology

In the present study, the synoptic conditions that favour the occurrence of aerosol episodes over the broader Mediterranean basin are investigated. For this purpose daily satellite measurements of aerosol optical depth (AOD) at 550 nm from MODIS-Terra (Collection 5) for the period from 1 March 2000 to 28 February 2007 were used together with mean-sea-level pressure and geopotential height at 700 hPa obtained from the NCEP/NCAR Reanalysis Project. An objective and dynamic algorithm was developed in order to determine strong and extreme aerosol episodes in the study area. According to the applied methodology and the relevant criteria, 322 days with aerosol episodes have been identified. The atmospheric circulation of each day was objectively classified by the implementation of Factor Analysis and Cluster Analysis, resulting in eight representative synoptic conditions (clusters). The aerosol episode days (AEDs) are most frequently observed during the dry period of the year, and especially in July–August (108 days or 33.5%) and April–May (106 days or 33%) and give rise to desert dust, anthropogenic pollution and biomass-burning episodes. The most frequent synoptic conditions are those of Clusters 4 (30.1%) and 5 (21%), inducing episodes that occur primarily in the western and secondarily in the central Mediterranean. In the former case, which is observed in summer, Europe is under the influence of the extended subtropical anticyclone of the Atlantic (Azores), while the eastern Mediterranean is under the influence of the extended thermal low of south-west Asia. In the latter case, in spring and summer, anticyclonic conditions prevail over central Europe and over the east Atlantic Ocean, and low pressures at the northwestern parts of the broader Mediterranean basin and across the Sahara. During the aerosol episode days of the various clusters, the mean regional AOD can reach values up to 0.8 ± 0.2 and 2.3 ± 1.3 for strong and extreme episodes, respectively. Copyright © 2011 Royal Meteorological Society

Atmospheric circulation patterns that are conducive to extreme ocean heat loss are investigated at four sites of special interest in the Mediterranean Sea. The Gulf of Lions, the South Adriatic Sea, the Cretan Sea, and the Levantine Sea are areas where anomalously high winter heat loss may lead to deep- or intermediate-water formation. At each of the above sites, the atmospheric circulation during such events is derived by averaging the sea level pressure (SLP) fields during the lower decile of the wintertime series of the net heat exchange.A relatively simple SLP pattern dominated by an anticyclone over northwestern Europe with a weaker cyclone to the southeast is found to be associated with strong heat loss in the selected sites with minor variations in pattern structure depending on the site. The SLP composite pattern reflects the combined effect of different atmospheric modes of variability and the authors consider the impacts on heat loss of a number of these modes (North Atlantic Oscillation, east Atlantic pattern, east Atlantic–west Russia pattern, and Scandinavian pattern), together with the North Sea–Caspian pattern and the Mediterranean index. The extremes in heat loss are strongly connected with the intensity and the positions of the poles of these patterns that modulate, through the necessary SLP gradient and associated northerlies, the transfer of cold and dry air over the areas of dense-water formation. Analysis of air–sea temperature difference, specific humidity, and evaporation anomalies indicates that the extremes of the net heat fluxes are primarily due to the latent and sensible heat flux components, This study was funded by the European Community’s Seventh Framework Programme (FP7/2007-2013) under Grant 287600 with the Policyoriented marine Environmental Research for the Southern European Seas (PERSEUS) project.MichelDe´que´ is acknowledged for running the ARPEGE-Climate model in its ARPERA configuration. The HIPOCAS dataset has been provided by Puertos de Estado (Spain).

Using sea-surface temperature data for the post-war period, the Mediterranean Sea is divided objectively (principal components factor analysis) into subareas with similar covariances. This division is not the same in every season of the year. In summer, two subareas are found, the west and the east. In spring and autumn three, and in winter four significant subareas denote physical oceanographic and meteorological processes. The scores of the analyses show clearly the minimum of the late 1970s and then the warming of recent years somewhat delayed in the eastern Mediterranean.

In this work, the extreme precipitation events in NW Greece are studied. The data used are daily precipitation totals recorded at the meteorological station of Ioannina University for the period 1970–2002. 156 days with precipitation totals above 35 mm (5% upper limit) are only considered. It is seen that, a minimum frequency of extreme precipitation events appears in the period 1986–1991, which is characterized by a high positive NAO index. For each of the 156 extreme precipitation days, at first, the mean sea level pressure pattern over Europe is constructed by using 273 grid point values. Using Factor Analysis, the dimensionality of the 156×273 data matrix is reduced to 156×5 (84% of the total variance) and then, Cluster Analysis is applied on the results of Factor Analysis. Thus, the 156 cases are grouped objectively to 11 clusters, revealing the main pressure patterns, which favour extreme precipitation in NW Greece. Seven of the patterns are encountered in winter and autumn, while three of them cover a period from autumn to spring and one appears mainly in summer. In all of them the cause of the extreme precipitation event is a low pressure system centred west of Greece or a low pressure trough extended eastwards or southwards up to Greece. In some cases the depression is so strong and extended that it covers the whole Europe and the Mediterranean. In the single summer pattern, rainfall is caused by an extension of the SW Asia thermal low up to the central Mediterranean.

The variability of sensible and latent heat fluxes in the Mediterranean Sea and their effect on precipitation in the Greek area during the winter months are investigated for the 39 year period 1959-97 by using multivariate statistical methods. First, factor analysis is applied, mainly as a data reduction tool, and then canonical correlation analysis is applied on the factor scores time series of the two pairs, i.e. sensible heat flux-precipitation and latent heat flux-precipitation, to reveal their interrelations. The results show there is a statistically significant relationship between sensible and latent heat fluxes over the western Mediterranean Sea and precipitation in the western windward areas of Greece. The sensible and latent heat fluxes in the western Mediterranean Sea contribute significantly to depression development. The depressions formed advect cold and dry air masses over the western Mediterranean, reinforcing these sensible and latent heat fluxes. Furthermore, these depressions lead to a southerly or southwesterly airflow over Greece. This flow, along with the existence of potential instability, significantly contributes to precipitation formation in the windward areas of western Greece and also in the islands of the eastern Aegean Sea. Copyright (C) 2004 Royal Meteorological Society. International Journal of Climatology

Five-minute data of global and diffuse irradiance and illuminance along with zenith luminance from Bratislava and Athens for a period of 5 years are used in order to define the prevailing sky luminance distributions at the two sites in the winter period. The classification into 15 sky types is mainly based on the value of the ratio of zenith luminance to diffuse illuminance with solar altitude as parameter (SSLD method). This is applied for the first time in a research work. It is found that the most frequent sky type in Bratislava is the overcast (with steep luminance gradation and slight brightening towards the sun); while in Athens it is the cloudless polluted sky (with a broad solar corona). That is, the winter daylight climate is quite different in Central Europe in comparison to the Mediterranean. An analysis for various solar altitudes reveals the frequency distribution of the 15 sky types suitable for the application in daylight climate studies. (C) 2002 Published by Elsevier Science Ltd. Journal of Atmospheric and Solar-Terrestrial Physics

The spatial and temporal covariability between the lower troposphere and sea surface temperatures (SSTs) are Studied in the Mediterranean basin for the period 1958-98. Monthly air temperature anomalies for the 850 hPa pressure level (T-850hPa) at 2.5degrees x 2.5degrees grid points and SST anomalies in 5degrees x 5degrees grid boxes are utilized. As a first step, factor analysis is applied oil both sets of data in order to reduce their dimensionality. Then, canonical correlation analysis is applied and this leads to one statistically significant pair of canonical variates for winter and to two pairs for summer. In winter, a teleconnection (see-saw) between western Europe and the eastern Mediterranean at the 850 hPa level is revealed, and a corresponding weaker one between the areas of central-west and eastern Mediterranean for SST. The correlation between T-850hPa and SST appears higher over the eastern Mediterranean. In summer, the first pair of canonical variates reveals a covariability between T-850hPa and SST in the western Mediterranean, and the second one shows a covariability in the eastern Mediterranean, without the existence of any strong spatial teleconnection. The analysis is repeated, using time lags of I month, or longer, in order to detect any possible non-synchronous relation. Statistically significant results are found only when T-850hPa leads SST with a time lag of I month. In particular, the results are statistically significant for winter only, and the findings are similar to those of the first analysis. Therefore, the existence of a I month time scale SST persistence is detected for winter months. Copyright (C) 2002 Royal Meteorological Society. International Journal of Climatology

This paper deals with the surface pressure covariability over the Altantic/European sector of the Northern Hemisphere. using monthly grid point data for 100 year period 1890 1989. Factor analysis is applied to 90 grid point time series for January, February. July, and August. The initial 90 pressure variables can be reduced to 7-8 factors in winter and 10 in summer. A winter teleconnection was identified. known as the seesaw phenomenon, between the Icelandic low and the Azores subtropical anticyclone. In order to define the centers of action for temperature, winter precipitation and summer northern wind frequency (etesian days) in Athens and in the Aegean sea, the variability of the factor scores and of these weather elements is compared. It is shown that the center of action for temperature in Athens is found to be in north and northwest Europe (centered over southern Scandinavia). For winter precipitation, the center of action is located in the west and southwest Mediterranean and northwest Africa. Finally, for the etesian winds frequency variability, this center of action is found over the northern Adriatic and northern former Yugoslavia, while there is no evidence of influence by the southwest Asia thermal low. Theoretical and Applied Climatology

This study deals with the annual variation of pressure over the Mediterranean area using monthly values from 24 stations, for a period of 22 years. The Azores and the Siberian anticyclones as well as the low pressure system of SW-Asia are responsible for the observed distribution pattern of the pressure over the area and for the seasonal changes at each station. Fourier analysis shows that for Central Mediterranean stations, which record two minima and two maxima of pressure during the year, the amplitude of the second harmonic is smaller than the amplitude of the third.