This study explores the interdecadal variability and
trends of surface horizontal visibility at the urban area of Athens from
1931 to 2013, using the historical archives of the National Observatory of
Athens (NOA). A prominent deterioration of visibility in the city was
detected, with the long-term linear trend amounting to
Visibility is defined as the greatest distance at which a black object of suitable dimensions (located on the ground) can be seen and recognized, when observed against the horizon sky during daylight (WMO, 1992). Visibility represents one of the dominant features of the climate and landscape of an area. Although it is highly affected by atmospheric circulation and the prevailing meteorological conditions, under clear sky conditions it is mainly determined by the loading in atmospheric aerosols (Davis, 1991; Lee, 1994; van Beelen and van Delden, 2012; Doyle and Dorling, 2002; Singh and Dey, 2012); therefore, visibility can be considered as a strong indicator of air quality over an area. Horizontal visibility has also been introduced in formulas for the estimation of atmospheric turbidity parameters (e.g., in the Ångström atmospheric turbidity coefficients; Eltbaakh et al., 2012).
Aerosols in the atmosphere contribute to light extinction by scattering and
absorbing, thus reducing visibility (Appel et al., 1985; Chan et al., 1999;
Elias et al., 2009; Singh and Dey, 2012). The impact of particulate matter
(PM)
on visibility depends on its physical (e.g., particle size distribution) and
chemical properties (Dayan and Levy, 2005). In particular, visibility is
inversely related to light extinction coefficient, which is determined by
scattering and absorption of light by gases and particles, the latter (e.g.,
sulfate- and carbon-containing particles) being the main contributor (Malm,
1999; Hand et al., 2002; Bäumer et al., 2008; Deng et al., 2011; Wang et
al., 2012). Sulfate- and carbon-containing particles play a major role in
light extinction, while the role of relative humidity (RH) on visibility is
also important (Larson and Cass, 1989; Malm, 1999) because when RH reaches
saturation values, visibility deteriorates due to fog formation and the
hygroscopic growth of SO
Although the use of visibility as a viable atmospheric variable has been
disputed by many researchers due to the numerous biases related to
observational procedures (Davis, 1991), visibility statistics have been
increasingly used as a surrogate for aerosol load (Zhao et al., 2011),
especially since visibility records span quite long-term periods. Today,
there is a large number of studies that use visibility observations to
investigate the spatial and temporal variation of the optical properties of
the atmosphere, mainly in relation to pollutant emissions and aerosol load.
These studies refer to global, regional and local scales. On a global scale,
a decrease of clear sky visibility over land from 1973 to 2007 is reported
by Wang et al. (2009). This is interpreted in terms of aerosol
concentration and its impact on incident solar irradiance. A significant
decrease in visibility was observed over Asia, South America, Australia and
Africa (1973–2007), while over Europe visibility increased after the 1980s
as a result of air pollution mitigation measures. Vautard et al. (2009)
found a significant decrease in the frequency of low-visibility days in
Europe after the 1980s, which is spatially and temporally correlated with
SO
In contrast to European areas, a tendency towards lower visibility is observed in developing countries (e.g., China, South Korea, South Taiwan, India), where it is still difficult to control air pollution (Ghim et al., 2005; Che et al., 2007; Wan et al., 2011; Singh and Dey, 2012; Wu et al., 2012). Along this line, Wu et al. (2012) found strong correlation between aerosol optical depth (AOD) and visibility in China over the period 2000–2009 and an overall decreasing trend in visibility (under sunny conditions) during the last 50 years. Singh and Dey (2012) correlated visibility in Delhi with aerosols composition and reported a rapid decrease of visibility during 1980–2000 and stabilization afterwards.
Urban environments are of particular interest, as air pollution from local sources is superimposed on regional ones, strongly impacting visibility (Davis, 1991; Eidels-Dubovoi, 2002; Tsai et al., 2003, 2007; Dayan and Levy, 2005; Chang et al., 2009; Kim, 2015).
The present study explores the historical observations of visibility in
Athens, which is the oldest time series of visibility in Greece and, to our
knowledge, one of the oldest, uninterrupted time series of visibility in the
Eastern Mediterranean. The records are retrieved from the historical
climatic archives of the National Observatory of Athens (NOA) and span a
period of more than 80 years (1931–2013). In the past, Carapiperis and
Karapiperis (1952) reported on the correlation between the visibility and
the blue color of the Attica sky, while Kanellopoulou (1979) analyzed
visibility in Athens for the period 1931–1977 and reported a pronounced
decrease after the 1950s. Since then, there has been no other study to
address changes in visibility, as well as the factors behind these changes
during the last 40 years, when significant changes occurred in Athens in
terms of urban expansion, traffic load, 2004 Olympic Games construction and
the economic recession (starting in 2008). The interdecadal variability and
long-term trends of visibility in Athens are presented in the study. The
role of meteorology and aerosol load (of local and regional origin) on the
variability and trends of visibility is investigated and discussed, while
the relationship between visibility and aerosol load is investigated
through the analysis of satellite AOD retrievals over Athens as well as
surface measurements of PM
Map of the study area in Greece, including the Athens urban station (NOA) and the non-urban reference station (HER) at Heraklion airport, Crete. The grey surface in the colored map represents the boundary of the Greater Athens Area (GAA).
Athens, the capital of Greece, is the main center of commercial, financial,
societal and cultural activities of the country. The Greater Athens Area
(GAA) (Fig. 1) extends beyond the administrative municipal city limits and
covers a surface of 433 km
In order to compare our findings for Athens with a remote reference site, the visibility records from the Heraklion airport (HER) in Crete Island were used (Fig. 1). Heraklion is located about 330 km south of Athens, while its airport is 5 km east of the city with no significant (or systematic) influence by the urban web.
Mean monthly and yearly values with standard deviations of basic climatic elements in Athens (NOA), calculated from the WMO climatic period (1971–2000)*.
* Climatic means were calculated from daily observations at NOA over the
period 1971–2000. Daily time series are almost complete, with sporadic
missing data in certain variables. In particular, data availability for the
period 1971–2000 is 100 % for
Athens has a temperate climate with warm and dry summers and wet and mild
winters, typical for the Eastern Mediterranean. Table 1 presents monthly and
annual normal values along with standard deviations of the daily mean
(
During summer, the area is dominated by anticyclonic circulation that enhances air temperature and intensifies the urban heat island. Athens has been experiencing a significant warming since the mid-1970s, more pronounced in summer, which is the additive result of regional warming and gradual intensification of the urban heat island (Founda, 2011; Founda et al., 2015). Strong northeasterly winds in summer, known from antiquity as “etesians”, induce a relief from air temperature and air pollution levels in the city.
Air mass origin was identified by applying a 4-day back-trajectory
analysis, calculated daily at 12:00 UT with the Hybrid Single-Particle
Lagrangian Integrated Trajectory (HYSPLIT) model (version 4.9; model data
used for vertical motion) (Draxler et al., 2009). Figure 2a presents the
main sectors related to air mass origin in Athens, based on a 10-year
climatology (2005–2014) of daily air trajectories ending at 1000 m above
ground level, while Fig. 2b presents the seasonal variability of air mass
origin according to the sectors defined in Fig. 2a. The S (south) sector is
linked to transport of air masses from arid areas of N Africa, frequently
associated with dust events that affect the Eastern Mediterranean (Hamonou
et al., 1999; Gkikas et al., 2016), the N (north) sector accounts for the
Balkans and the main continental Europe, while the W (west) sector
corresponds to SW Europe and the W Mediterranean Basin. Note that air mass
transport from the W sector is significantly blocked by the high-altitude
mountain chain of Pindus (
Relative frequencies of surface wind directions for three
wind speed (wsp) categories at NOA, based on hourly values of the period
1971–2000. The integral of the upper curve is 100 %. For instance, the NE
direction occurs cumulatively at a frequency of 17 %, which is the sum of
7.9 % (wsp
Similar conclusions are drawn from surface wind measurements, reported in Fig. 3. Winds from N–NE directions prevail in Athens at a frequency of nearly 38 % (Fig. 3). This sector is also associated with the occurrence of high wind speeds, as shown in the same figure. The second-most-frequent surface winds correspond to S–SW directions (27 %). The frequency of occurrence of this sector is maximum during the intermediate seasons (spring and autumn) and is associated with the occurrence of dust events from N Africa and, in cases of light winds, with sea breezes from the Saronic Gulf (Fig. 1).
A short introduction on the factors that diachronically control air pollution levels in Athens is presented here to facilitate the interpretation of visibility variations in terms of pollutants concentrations.
Air pollution in Athens has been systematically measured since the early
1970s. Road transport, domestic combustion and industrial activity have been
the main sources of air pollution in GAA throughout the years. Downward
trends of sulfur dioxide, black smoke, carbon monoxide and nitrogen oxides
have been reported from the mid-1980s to the late 1990s, attributed to
several anti-pollution measures adopted by the state (e.g., replacement of
the old technology gasoline-powered private cars and the reduction of the
sulfur content in diesel oil) (Kalabokas et al., 1999a). Negative trends of
NO
Measurements of PM had only occasionally been conducted
in Athens before the EU Directive (1999/30/EC) was launched, revealing
increased concentrations of PM
Studying the contribution of local sources vs. regional and the role of
long-range transport over megacities of the Eastern Mediterranean, including
GAA, Kanakidou et al. (2011) summarized that a significant number of PM
exceedances registered in Athens is associated with regional pollution
sources or natural dust transport, clearly highlighting the importance of
regional transport processes. Theodosi et al. (2011) compared simultaneous
mass and chemical composition measurements of size-segregated particulate
matter (PM
Regarding columnar aerosol load and using ground-based AOD measurements in
Athens, Gerasopoulos et al. (2011) showed that the greatest contribution
(40 %) to the annually averaged AOD comes from regional sources (namely
the Istanbul metropolitan area, the extended areas of biomass burning around
the northern coast of the Black Sea, power plants spread throughout the Balkans
and the industrial area in the Po Valley). Additional important contributors
are dust from Africa (23 %), whereas the rest of Europe contributes
another 22 %. Gkikas et al. (2016) found good correlation between
AOD
Vrekoussis et al. (2013) reported an improvement of air quality in Athens during the period 2008–2013 as a result of the economic recession and the subsequent reduction in vehicle use and industrial activity. For the same period, Paraskevopoulou et al. (2014) showed that the massive shift of Athens' population to wood burning for residential heating purposes gave rise to smog episodes characterized by high PM spikes during nighttime in winter. A longer-term (2008–2013) analysis of aerosol chemical composition and sources at a suburban site in Athens by Paraskevopoulou et al. (2015) revealed that the area of Athens is now generally dominated by aged, transported aerosols.
The historical climatic record of the NOA
was used in this study. NOA is located on the Hill of Nymphs (37.97
Visibility data at other stations (e.g., Heraklion, Crete) were extracted from the network of the Hellenic National Meteorological Service (HNMS) and actually represent visibility observations at the airport station, initiated after the mid-1950s. Meteorological data for Athens over the period 1931–2013 were also acquired from the historical archives of NOA. Monthly, seasonal and annual mean values of visibility were derived from the daily observations at 14:00 LST.
The WMO empirical scale for visibility observations, used at NOA.
An empirical scale of visibility classes, as recommended by the World Meteorological Organization (WMO), has been used for visibility observations at NOA (Table 2). Classes are defined based on the greatest distance at which a predefined object can be seen and recognized by the naked eye. The procedure requires that an operator scans the horizon for predetermined objects. In the case of Athens, some historical buildings in the city, as well as certain objects of the surrounding landscape that remained unaltered over the years (e.g., objects on the mountains or islands of the Saronic Gulf; Fig. 1), were chosen to represent visibility classes and relevant distance ranges. The procedure introduces inevitably some kind of subjectivity and bias in the measurements, related to individual eyesight of different operators. It is assumed, however, that the execution of visibility observations by different operators over the years could have possibly had a compensating effect and an overall reduction of biases. More details about the possible errors and validity of visibility observations have been thoroughly discussed by Davis (1991).
The use of the WMO scale introduces a further uncertainty to visibility observations, associated with the amplitude of visibility ranges corresponding to each visibility class. Information on the use of WMO scale and relative uncertainties, as well as the followed procedure for averaging daily visibility observations, is provided in the Supplement.
Long time series of atmospheric pollution measurements in Athens and the selected reference site would enable drawing relationships between visibility and aerosols and would provide evidence for the origin (regional or local) of atmospheric pollution in Athens and its impact on long-term visibility variations. Given that such time series are missing, we used shorter time series of aerosol measurements for a direct comparison between visibility and atmospheric pollution in Athens.
In an effort to explore the relationship between visibility and AOD over
Athens, we used the Terra/Modis AOD at 550 nm, available since 2000. NASA's
Terra satellite is sun synchronous and near polar-orbiting, with a circular
orbit of 705 km above sea level. MODIS is capable of scanning 36 spectral
bands across a 2330 km wide swath. MODIS aerosol products were used in order
to analyze the temporal and spatial variability of aerosols over the wide
area of interest. In this study, we used daily level-2 collection 5.1
MODIS/Terra AOD at 550 nm. Daily overpass data for the specific area were
extracted at a spatial resolution of
In addition, in order to further examine long-term satellite-based AOD
series in the area, we used the longest satellite time series available from
the Advanced Very High Resolution Radiometer (AVHRR). AOD retrievals
PATMOS-x AVHRR level-2b channel 1 (630 nm) provide data over global oceans
at high spatial resolution (0.1
Surface PM
Finally, a dataset of PM
Interdecadal variability of the annual visibility in Athens from 1931 to 2013, along with linear trends for three subperiods: 1931–1948, 1949–2003 and 2004–2013 (red line). The dashed blue line illustrates the population growth in Athens (in millions) since 1930 (Founda, 2011). The long-term variability of the annual relative humidity (RH) in Athens is also shown (upper black line).
Figure 4 displays the long-term development of the annual visibility in
Athens from 1931 to 2013. The population growth in the city of Athens over
the same period is also shown, while the figure also displays the long-term
variability of the RH in Athens (which is discussed
below). It is obvious that the annual visibility in Athens has undergone a
very strong and almost continuous decline over the past 80 years, in
coincidence with the increase in population. The long-term linear trend over
the entire study period was found to be equal to
The separation of the time series into three subperiods, as described above, follows changing trends. In the following, the much longer middle subperiod (1949–2003) was further separated into two parts (1949–1975 and 1976–2003), as it corresponds to substantially different visibility conditions. Figure 5 illustrates the frequency distribution of the different visibility ranges (as described in Table 2) for those four different subperiods.
In the first subperiod (1931–1948), visibility values are almost equally
distributed between the ranges of 10–20 km and 20–50 km, at frequencies of
approximately 35 %. Very high visibility (
A progressive shift of frequency distribution towards lower visibility
categories is observed in the next subperiods. In particular, the frequency
of very good visibility (20–50 km) drops to 13 and 6 % for the periods
1949–1975 and 1976–2003, respectively, while the most frequent visibility
range is 10–20 km (44 %) during 1949–1975 and 4–10 km (41 %) during
1976–2003. The frequency of visibility
The frequency distribution changes dramatically during the most recent
period (2004–2013). In particular, although visibility range of 4–10 km
remains the most frequent (30 %), as in the subperiod 1976–2003, almost
similar frequency (
Relative frequency distribution of different visibility ranges (as defined in Table 2) in Athens for the four subperiods 1931–1948, 1949–1975, 1976–2003 and 2004–2013.
Since visibility is influenced by the prevailing meteorological conditions
(Davis, 1991; Sloane, 1982), it is expected to exhibit a seasonal
variability, depending on the intra-annual variability of climatic
conditions at the study area. Mean monthly values of visibility were
calculated for the subperiods 1931–1948, 1949–1975, 1976–2003 and
2004–2013. Figure 6a–d present the mean monthly values of visibility in
Athens over each subperiod, normalized with the value of the month with the
highest visibility. In the same plot, the mean monthly values of RH coinciding visibility observations at 14:00 LST over each
subperiod are also shown. It is noteworthy that RH at NOA does not exhibit
any significant trend over the years (as already shown in Fig. 4) and its
monthly distribution remains almost unaltered in all subperiods. As one can see from Fig. 6a–d, visibility exhibits a seasonal cycle in all
subperiods, with better visibility occurring in the warm and dry months of
the year. Although seasonality is observed in all subperiods, the pattern
is more evident and robust in the first subperiod (Fig. 6a), with much
higher visibility values (up to 40 %) in the warm and dry months. The
pattern of visibility in this period is almost a mirror image of the pattern
of RH and reflects the influence of RH on visibility and the
anti-correlation between these two variables. The lowest values of RH
correspond to July and August (mean value of RH
The distinct seasonal cycle in the visibility of the first subperiod changed in the following subperiods (Fig. 6b–d). Although the warm and drier months always correspond to higher visibility levels, seasonality is noticeably attenuated and visibility differences between the warm and cold period are much lower. This possibly implies a weakening of the influence of meteorological conditions, as a result of (or in combination with) the stronger effect of air pollution on the visual air quality of the city.
The minimum of visibility is constantly observed in March during all
subperiods. Indeed, March is a month in the transitional season and thus
bears higher values of RH compared to summer months (mean RH at 14:00 LST
The impact of meteorological conditions on visibility has been investigated by different researchers using different approaches, as for instance the classification of synoptic circulation patterns (Sloane, 1982; Davis, 1991; Dayan and Levy, 2005), the application of correction factors on extinction coefficient to account for RH effect (Che et al., 2007), the estimation of correlation coefficients between visibility and meteorological variables (Deng et al., 2011) or simply the comparison of diurnal/seasonal cycles and temporal trends of visibility with the relevant cycles and trends of meteorological variables (van Beelen and van Delden, 2012). Sloane (1982) reported that periods with exceptionally maxima or minima of visual air quality were related (apart from sulfate emissions) to favorable synoptic circulation patterns. Studying visibility in Tel Aviv, Dayan and Levy (2005) reported a strong dependence of visibility levels on meteorological conditions, synoptic weather patterns and air mass origin, with the highest mean values occurring in summer, related to the persistent nature of the summer synoptic weather patterns in the Eastern Mediterranean. Deng et al. (2011) found that RH and wind speed were significantly correlated with visibility at an urban area of China, while Ghim et al. (2006) showed a considerable decrease in visibility in South Korea, despite the observed simultaneous decrease in RH levels. The relationship and possible impact of different meteorological parameters such as precipitation, RH, wind speed and wind direction on visibility in Athens is discussed below.
Long-term variability and linear trends of the annual
precipitation amount and precipitation frequency (number of days year
Precipitation is associated with scavenging of atmospheric particles (e.g.,
Remoudaki et al., 1991a, b), possibly resulting in improvement of
visibility. The PF in particular was found to control
seasonal variability of the total atmospheric deposition of lead in the
western Mediterranean (Remoudaki et al., 1991b). Rainy days, in contrast, are associated with increased RH, resulting in
reduction of visibility. A plot illustrating the long-term variability of
the annual precipitation amount and PF at NOA from
1931 to 2013 was created, for the detection of any significant trends (Fig. 7).
As can be seen in the figure, no long-term trend is observed in the annual
precipitation at NOA from 1931 to 2013, which could have had an effect on
long-term trends in visibility. Precipitation frequency, in contrast,
exhibits an overall negative trend over the same period (
Subsets of data were also produced for the creation of additional visibility time series, accounting for precipitation influence. Figure 8 presents visibility variability during the wet (October–March) and dry (May–September) periods of the year, along with the annual values. Lower values during the rainy and cold period of the year are most probably associated with higher values of RH, resulting in the reduction of visibility. Despite the differences between the time series in Fig. 8, the overall tendency is similar, thus not affecting the validity of our conclusions regarding the long-term visibility impairment in Athens. Additional plots, created from subsets of “rain” and “no rain” days, are provided in the Supplement (Fig. S4).
Variation of visibility at NOA from 1931–2013 during the dry (May–September), wet (October–March) and all year (January–December) period.
Figure 9 presents the running correlation coefficient (15-year window)
between visibility and relative humidity and visibility and surface wind
speed at NOA, over the period 1931–2013. As expected, the correlation
coefficient between visibility and RH is negative, indicating the
anti-correlation between these two variables. High RH enhances water uptake
by airborne particles, leading to higher light scattering and, thus,
visibility impairment. Actually, when RH exceeds a threshold level (e.g.,
Running correlation coefficient and 99 % confidence levels (CL; dashed lines) between visibility and wind speed (blue line) and visibility and RH (green line) in Athens, over the period 1931–2013. A 15-year window was used.
Following Fig. 9, the negative correlation between RH and visibility is statistically significant at the 99 % confidence level almost over the entire study period. However, a progressive weakening of the correlation coefficient with time is observed, indicating a less strong correlation between the two variables over the years from the beginning. Stronger anti-correlation is found until the early 1970s, followed by lower (still significant) values until the mid-1990s. The progressive weakening of the correlation between RH and visibility in Athens possibly suggests a progressive weakening or mask of RH influence on visibility compared to the effect of other factors such as atmospheric pollution (although the influence of RH is enhanced by the presence of certain hygroscopic particles).
On the contrary, the impact of surface wind speed on visibility seems to be
stronger during the late part of the time series (Fig. 9). Higher wind
speeds in this case (positive correlation) are related to the more efficient
city ventilation and dispersion of air pollutants, resulting in visibility
improvement. In other cases, wind speed is also used as a proxy for
long-range transport, but then a negative correlation would be expected.
Lower values of the coefficient in the early part of the time series
possibly demonstrate that the lack of pollutants at that period detracts
from the importance of ventilation. The correlation coefficient increases
progressively over the years. The rate of increase is higher after the mid-1980s, when correlation becomes statistically significant at the 99 %
confidence level. Similar values of correlation coefficient (
Variation of visibility with wind direction (sectors) over
the subperiods 1931–1948, 1949–1975, 1976–2003 and 2004–2013. Visibility is
normalized by its maximum value at a certain sector for each subperiod.
Sector “C” corresponds to calms (wind speed
Apart from wind speed, visibility was also found to be sensitive to wind direction. A distinct variability of visibility with wind direction is observed in Fig. 10, for all subperiods. Lower values of visibility are related to southerly winds, as they bring either dust from Sahara or warmer and more humid air masses from the sea (see also Figs. 1, 2b). Southeasterly winds are, in general, weak winds (see Fig. 3), while southwesterly winds are associated with sea breezes from the Saronic Gulf (Fig. 1). In general, sea breezes and calm wind conditions favor the accumulation of pollutants and the formation of secondary aerosols and photochemical smog in Athens (Colbeck et al., 2002), thus reducing visibility. A number of S/SW events are also associated with strong wind speeds occurring during Sahara dust outbreaks, which enrich Athens atmosphere with dust particles that decrease visibility (Figs. 2, 3). As one can see from Fig. 10, the highest visibility occurs under northwesterly winds and this is robust for all subperiods. One explanation is that air masses originated from northwesterly directions are much drier as they have lost water vapor after passing over the high mountainous basin of the Greek mainland (e.g., Pindus mountain), while air pollution is also blocked within the boundary layer by the mountain chain.
In this section we attempt to interpret the observed interdecadal variability and trends of visibility in Athens in terms of air pollution. As already shown in Fig. 4, the pre-1950 period is characterized by considerably higher visibility levels in Athens. From then on, visibility experienced a rapid decrease, followed by a smoother but continuous decreasing trend until the early 2000s. The period after 1950 signifies the post-World War II epoch but also coincides with the end of a civil war in Greece (1946–1949), which was followed by an important urbanization wave in Athens (Maloutas, 2003). This is in line with the rapid growth of Athens' population, as illustrated in Fig. 4. The greatest rate of population increase is observed between 1950 and 1960, when population in Athens almost doubled. The population growth was associated with a significant increase of construction in the city. However, apart from the intense urbanization in Athens, this period is also characterized by the most prominent increase of anthropogenic emissions on a global and European scale (e.g., Mylona, 1996; van Aardenne et al., 2001; Vestreng et al., 2007, 2009).
Interdecadal variability of the annual average visibility
at NOA (urban) and HER (background) stations. Bold black lines represent the
common period of observations (1956–2009) at the two stations, along with
linear trends and slopes. The solid blue line illustrates historical development
of European emissions of SO
Are the changes in visibility in Athens due to local factors or can they be
considered representative of a more extensive area? To answer this question,
the Athens visibility record was compared with visibility at a non-urban reference
station. From the available stations in Greece with long-term
visibility observations, we chose the station at HER on
the island of
Crete (Fig. 1). Actually, both sites, NOA and HER, are exposed, most
of the year, to air masses of similar origin (from north and northeasterly
directions) traveling over the Aegean Sea, in contrast to other sites of
the country that are strongly affected by the mountainous volumes of the
Greek mainland. Visibility observations at HER are available since the mid-1950s. Figure 11 presents the long-term variation of the annual averages of
visibility at HER along with the annual visibility at NOA. Linear trends of
the time series for their common period (1956–2009) are also shown in the
figure. The time series were found significantly correlated (correlation
coefficient
According to Fig. 11, visibility levels at urban NOA are constantly lower by
a few kilometers (
Segregation of emission trends by air mass origin would further enlighten
their possible effect on visibility variation in Athens. As stated in
Sect. 2.2, air masses from the N–NE sectors dominate in Athens,
contributing by more than 60 % on an annual basis. Following segregation
of European SO
In regard to other types of emissions such as organic carbon (OC) or black carbon (BC), historical data reported by Bond et al. (2007) show an increase of the order of 50 % on a global scale between 1930 and 2000. However, segregation by region indicates that European emissions of OC and BC revealed a slight increase between 1950 and 1970 and decrease thereafter. Decreasing trends are also observed in the former USSR after 1970 (Bond et al., 2007).
A very interesting finding in Fig. 11 is the similar slopes in the negative
linear trends of the annual visibility at the background and urban stations
over their common period of observations (
After the early 1990s, the two time series diverge. Background visibility at HER partly recovers, while visibility at NOA keeps declining at the same pace until 2003 (Fig. 11). Recovering visibility is also found at other Greek areas around the 1990s (Lianou et al., unpublished data), which is in line with visibility improvement in other European areas related to emissions reduction (Vautard et al., 2009; Wang et al., 2009). This last feature suggests that, during this period, local emissions might have a dominant role in the determination of visibility in Athens.
A slight recovery of visibility is observed during the decade 2004–2013 (Figs. 4, 11). This improvement could be attributed to a number of reasons. The years after 2004 correspond to the post-Olympic Games period in Athens. A number of important transport projects were completed prior to the Olympic Games in Athens in 2004. Such projects are for instance the construction of the Attica Ring Road (one of the largest in Europe), the construction of Tramway and the extension of Athens Metro. These projects have contributed to the reduction in the number of vehicles in the city, resulting in less traffic problems and lower air pollution levels. Another possible contributing factor concerns the impact of the Greek economic recession (2008–2013) on air quality in Greece, and Athens in particular. Recent studies provide some evidence on this. For instance, Vrekoussis et al. (2013) found strong correlation between different economic metrics and air pollutants after 2007, suggesting that the economic recession has resulted in proportionally reduced levels of air pollutants in the two biggest cities in Greece. This is further supported by other recent research studies that report a significant reduction in energy consumption after 2008, related to the rapid economic degradation (Santamouris et al., 2013).
The relationship of visibility with AOD over Athens was also explored, using two different satellite-based datasets (AVHRR and MODIS) from 1981–2009 and 2000–2014, respectively (see Sect. 2.5). For the AVHRR AOD at 630 nm, Fig. 12a shows a 1.7 % per year decrease from 1981 to 1997 and a 2.4 % decrease from 1999 to 2009 (1998 data were not available). It is interesting to note the AOD maxima in 1991 and 1992 that are linked with the Pinatubo eruption period. The AOD time series for the MODIS instrument at 550 nm showed a significant and similar to AVHRR (2.4 % per year) decrease from 2000 to 2009 and a further decrease of 7.4 % per year for the period 2010–2014 (Fig. 12b).
To investigate the relationship between visibility and AOD changes, the two
parameters are plotted together after data binning. Visibility and AOD
measurements have been used as follows: visibility at 12:00 UT was used
according to the indices defined in Table 2 and plotted against average AOD
from synchronous satellite overpasses of AVHRR and MODIS separately. The
mean AOD and its standard deviation are presented in Fig. 13. Average AOD
from AVHRR and MODIS are not directly comparable, as they represent
different time periods and different wavelengths. The AOD values are related
to the visibility data, using the middle point of each
visibility bin (range) as the distance in kilometers. Only summertime (June–August) MODIS and AVHRR
AODs
have been used, to keep visibility values unaffected by other atmospheric
parameters like low clouds, rain or relative humidity. It is observed that
for average AOD values over Athens (0.25 using the mean June–August AOD at
550 nm from our MODIS AOD dataset or 0.23 at 500 nm as reported by
Gerasopoulos et al., 2011), visibility varies within the range of 4 to 10 km. Under cleaner conditions (W–NW–N, 0.12–0.17 at 500 nm; Gerasopoulos et
al., 2011), visibility can reach as high as 20 km, while very low visibility
(
Illustrating the relationship between AOD, which consists of a vertically
integrated parameter, and visibility, a horizontally integrated parameter,
requires various assumptions. Using satellite-based AOD and visibility
observations for GAA, when assuming a vertically constant extinction
coefficient and a mixing layer that contains all aerosol load we end up
describing the theoretical relationship (Koschmieder, 1924): Vis
MODIS at 550 nm (green) (2000–2014) and AVHRR at 630 nm (red) (1981–2009), AOD (June–August) mean values and standard deviations for each visibility index. Shaded areas represent visibility ranges (km) for each visibility class (Table 2). AOD averages have been represented here in the average distance from each class.
Visibility as a function of different classes of PM
An additional analysis was conducted to verify the relationship between
visibility and particulate pollution from surface measurements, using a
short dataset of PM
Finally, the variation of the annual averages of PM
Variation of the annual PM
The present work analyses, for the first time, the long historical record of visibility at NOA (Athens) from 1931 to 2013 and interprets its temporal variability and trends in terms of relevant changes in atmospheric properties (related to local or regional processes) and/or meteorological conditions. Since this is the longest record of visibility observations in Greece and one of the oldest in the broader area of the Eastern Mediterranean, the study provides unique information on the atmospheric properties of the area in the past, when air pollution records are missing. The study period was divided into subperiods corresponding to different trends in the time series of visibility, each subperiod being affected by different factors.
The impact of meteorological conditions on visibility was investigated in
different ways. Visibility in Athens was found to follow a seasonal cycle,
with higher visibility corresponding to the warm and dry months of the year.
Seasonality is more distinct in the first subperiod of the time series
(1931–1948), while after the 1950s the seasonal cycle attenuates.
Visibility was found to be negatively correlated with RH, the correlation
being stronger in the early part of the time series and attenuating over the
years. In contrast, a positive correlation between visibility and wind
speed was found, statistically significant during the late part of the time
series, suggesting the increasing role of winds on the cleanup of the
atmosphere from air pollutants. Visibility was also found to be sensitive to
wind direction, reflecting the influence of air mass origin. Lower
visibility levels are constantly observed under southerly winds,
corresponding to sea breeze circulation as well as to African dust outbreaks.
The study demonstrated that visibility in Athens has undergone a prominent
impairment since the early 1930s. The overall trend of the annual visibility
averages was found equal to
The comparison of the annual averages of visibility in Athens and at a non-urban
reference site (HER) in Crete, revealed similar and statistically
significant negative trends at both sites, suggesting the major contribution
of long- and regional-range transport of natural and anthropogenic pollution
in the GAA. An improvement of visibility at HER around the 1990s was not
associated with synchronous improvement of visibility in Athens, where
visibility deterioration continued until the early 2000s. Although negative
trends of main gaseous air pollutants are reported in Athens at that period
(Kalabokas et al., 1999a), the direct effect of such pollutants on light
extinction is negligible compared to suspended particles and particularly to
fine particles (
The relationship between AOD and visibility in Athens was also examined in
the study, using MODIS and AVHRR satellite data (Figs. 12, 13), and confirmed
their negative correlation. Also, a strong anticorrelation was found between
visibility and PM
The analysis showed a recent stabilization (or even slight improvement) of
visibility in Athens, consistent with the observed decreasing trends of
PM
The 82-year-long time series of visibility in Athens unfolded for the first time information on the atmospheric conditions over the area, for periods when atmospheric pollution measurements are missing. Although the analysis is subject to several limitations and assumptions, mainly associated with the methods of visibility observations, the results are robust and statistically significant, showing an outstanding degradation of the visual air quality in the city from the 1930s to the 2000s.
All data are available upon request to the authors.
The study is a contribution to the ChArMEX (The Chemistry-Aerosol Mediterranean Experiment) work package on variability and trends. The study was supported by the Excellence Research Program GSRT- Siemens (2015–2017) ARISTOTELIS “Environment, Space and Geodynamics/Seismology 2015-2017” in the framework of the Hellenic Republic-Siemens Settlement Agreement. The authors are grateful to the Editor François Dulac and the two anonymous reviewers for their very useful comments and suggestions on this study. The authors would also like to thank the Hellenic National Meteorological Service (HNMS) for the provision of visibility data at Heraklion (Crete) and the Air Quality Department of the Ministry of Environment and Energy of Greece for the provision of air pollution data. The contributions of F. Pierros (NOA) and D. Koutentaki (NOA) in the digitization of visibility data of NOA and of G. Kouvarakis (University of Crete) in the analysis of air trajectories are also acknowledged. Finally, the important contribution of all operators in the acquisition and maintenance of uninterrupted, continuous and reliable historical climatic records of NOA is acknowledged.Edited by: F. Dulac Reviewed by: two anonymous referees