Atmospheric Environment Vol. 24B, No. 2, pp. 221 225, 1990. 0957 1272/90 $3.00+0.00 Printed in Great Britain. © 1990 Pergamon Press plc

O N THE T R E N D OF THE TRANSMITTANCE OF DIRECT SOLAR IRRADIANCE IN ATHENS D U R I N G THE S U M M E R

G. S. KARRAS, D. K. PISSIMANIS and V. A. NOTARIDOU Department of Applied Physics, University of Athens, Greece

(First received 11 July 1988 and in final form 13 February 1989)

Abstract--In recent years, high concentrations of air pollutants in Athens have been measured. Since 1962 measurements of the broadband direct solar irradiance, which is directly affected by the aerosol loading in the atmosphere, have been available from the National Observatory of Athens (NOA). Using these data, the year to year trend of the mean value of the midday transmittance with clear sky in summer is evaluated for the period 1962-1985. A pronounced decrease of the transmittance occurred between the years 1962 and 1975. This probably reflects the rapid urbanization of the Athens area.

Key word index: Direct solar irradiance, transmittance, aerosol loading, trend, Athens.

I. INTRODUCTION

During the last few decades, a continuous increase in air pollution has occurred in the city of Athens as a consequence of its rapid development and has led to a significant deterioration in the air quality. Although air pollution episodes were officially recorded for the first time in the early 1970s when a network of stations measuring the concentrations of SO2 and smoke was established in the city, it is more likely that such episodes started to occur many years before. The city lies within a basin (the Athens basin) which is bounded by mountains or hills to the north, east and west and is open only to the south towards the sea (the Saronic Gulf, Fig. 1). In the last four decades the population of the city has increased from approximately 1 to 4 million, while the populated area has spread so much that it has practically covered all the interior of the Athens basin. At the same period the number of circulating vehicles, which initially was only of the order of a few tens of thousands, has vastly increased, now exceeding 1 million and situations of heavy traffic in the main roads leading to the downtown area have become very familiar to the citizens. Further- more, a considerable number of small industrial units has been gradually established in the interior of the basin and in its western vicinity, where the surround- ing hills are very low, their height ranging only from 200 to 400 m. It is also worth noticing that air traffic has also greatly increased at the airport which lies at a distance of only 10 km to the south-southeast of the center of the city. If all these facts are taken into account, there remains no doubt why the concentra- tions of pollutants have considerably increased in the recent years and air pollution episodes do occur whenever the synoptic weather conditions become unfavorable for dispersion.

As has been verified with the help of data of smoke and SO 2 concentrations obtained by the network that

has been mentioned before, such weather conditions conducive to air pollution episodes are more frequent in the cold period of the year, but they also occur quite often in the warm months (Lalas et al., 1983). More- over, the abundant sunshine that prevails at times during the episodes, often causes the production of secondary photochemical pollutants. However even for the early 1970s, the information available on the air quality is poor since by that time the above network consisted only of two stations. While the number of stations gradually increased to five during the first half of the 1970-1980 decade, measurements of the concentrations of photochemical products as well as other pollutants were started only by 1980. It can he concluded from the above that information on the time evolution of air pollution in Athens from direct measurements can be deduced to a certain degree only for the last 15 years. Nevertheless, there still remains the possibility of drawing indirectly some interesting information on the time evolution of the air quality for a longer period in the past by utilizing some other meteorological parameters that have been meas- ured. One such parameter is the transmittance of the direct solar irradiance (~) which is known to be affected by the presence of pollutants in the form of aerosols in the atmosphere. The aerosols cause an attenuation of the direct solar irradiance mainly in the short wavelengths of the solar spectrum, due to scattering and absorption. In general, the reduction of the transmittance of solar irradiance by the pollutants in the urban areas has been verified by several authors (Wesely et al., 1976; Bach, 1973; Hodges, 1977 etc.). When measurements of direct solar irradiance (In) are available, the transmittance can be calculated from the relation

z=(In/loEo) (1)

where 1o is the solar constant and Eo is the reciprocal of the square of the relative radius vector of the Earth. Long term measurements are available at the Natio-

221

222 G.S. KARRAS et al.

O0

Oe OQ

O

5 I 1OKra

v w I

n ' ~ l Iff%m ~ r r l zat,¢ / o ~ r

Fig. 1. Topographical map of the Athens basin and its vicinity with the industrial point sources by 1975, (black dots), as issued by the Technical Chamber of Greece. Contours are drawn every 100 m. Dot size is

proportional to pollutant emission.

nal Observatory of Athens (NOA: q~=37°58 ', ). =23043 ', h=107 m) at 8:20, 11:20, 14:20 and 17:20 Local Standard Time (LST is 2 h ahead of GMT). The broadband direct solar irradiance is measured every day by a Kip and Zonen actinometer on the condition that clouds are not present in the sightpath. Since the transmittance itself depends on the relative optical air mass (and hence on the zenith angle of the sun) as well as on the attenuation caused by the other atmospheric constituents, it is evident that its values that may be obtained from these measurements could not gen- erally serve as a measure of the atmospheric contami- nation. Nevertheless, if groups of measurements could be selected for every year, for which the variation of the air mass and at the same time the variation of the attenuation due to the other atmospheric constituents

could be considered as very small, the long term trend of their mean values could prove to be useful for drawing at least some qualitative information on the evolution of the air quality of the city. For Athens, the most favorable period for the selection of such groups is the summer season in which not only the relative optical air mass changes very slowly, at least at the noon hours, but also clear days are frequent and therefore a large number of measurements exist. Also, as will be explained in the next section, the variation of the attenuation of the direct solar irradiance by the atmospheric constituents other than aerosols, is small in this season for measurements obtained with clear skies.

In this study, an attempt is made to deduce some qualitative information on the time evolution of the

Transmittance of direct solar radiation 223

air quality in Athens for the summer season and with clear sky, for the period 1962-1985, by determining the transmittance of the direct solar irradiance from the measurements of 11:20 and 14:20.

2. THE ESTIMATION OF THE TREND OF THE

TRANSMITTANCE OF DIRECT SOLAR IRRADIANCE

IN SUMMER

The attenuation of the direct solar irradiance during its passage through the atmosphere is caused by three main physical processes which affect the various re- gions of the solar spectrum in a different way. These are: 6) molecular scattering caused generally by the gaseous constituents of the atmosphere; (ii) absorption by the atmospheric gases, (mainly by water vapor, but also ozone and mixed gases, i.e. 0 2, N z, CO 2 etc.) and (iii) the absorption and scattering by aerosols. From the Lambert-Bouger law, the transmittance of the direct solar irradiance for the whole solar spectrum can be expressed as follows:

f ~= ® - A~.m) d2 Io(2)e =0

= ( 2 )

fz L= ~ lo0.)d2 =0

where: lo(2 ) is the spectral irradiance at the top of the atmosphere, A(2,m) is the optical path length of the atmosphere at wavelength 2 and m is the absolute optical mass. By taking into account the main physical processes that cause the attenuation of a solar beam, the optical path length A(2, m) may be expressed as follows:

A(2, m) = maaR(2 ) + ~ mriaw,(2) + mraD(2 ). (3) i=1

In Equation (3) aR(2) is the Rayleigh extinction coeffic- ient per unit absolute air mass, ma is the absolute air mass for the atmosphere (i.e. m a = mr(P/lO13), where P is the pressure at the station and m r the relative optical air mass), awi(2) are the selective absorption coeffic- ients for water vapor, 03 and mixed gases (02, N2, CO2 etc), m, are the relative air masses of the absorb- ing gases and aD(2) is the aerosol extinction coefficient per unit air mass. The relative optical air mass for the aerosol loading is usually considered as equal to that of dry air. It becomes obvious that if groups of measurements of direct solar irradiance for zenith angles of the sun ranging within small intervals are selected, then the optical path length of the atmos- pheric constituents will virtually depend on the absol- ute optical mass in the vertical of each constituent. Further, if the variation of the optical masses of the gaseous constituents are small then, according to Equations (2) and (3), the transmittance will depend on the aerosol loading of the atmosphere. In considering the available data it became obvious that the most favorable period for which the above condition could

be accomplished would be the summer season for the following reasons. First, because the mean cloudiness during this period is very low a considerable number of measurements do exist. On the other hand, the zenith angle of the sun remains relatively small, espec- ially at the hours 11:20 and 14:20. In order to investi- gate the possibility of selecting groups of measure- ments for which the variation of the corresponding relative optical air mass could be considered as negli- gible to a good approximation, the relative optical air masses at 08:20, 11:20, 14:20 and 17:20 were found for all days by using Kasten's (1966) formula:

i.e. mr=(COS Oz+0.15(93.885-Oz)- 1.z53)- 1 (4)

where ®z is the zenith angle of the sun. For this purpose, the local time was converted into true solar time and the zenith angle of the sun was found. The extreme values of the relative optical air masses at the 4 h in each month are given in Table 1. It can be seen that at the hours of 11:20 and 14:20 the variations in the whole summer period are small reaching the values of 0.12 and 0.15, respectively. In addition to the above, the variations of the absolute optical air mass that may be induced by variations of the surface pressure can be considered as negligible because the last are generally very small in Athens during the summer, rarely exceeding 15 mb. Further on, care may be taken to avoid considerable variations of the absolute water vapor mass or precipitable water, by selecting only measurements taken with cloudiness ~< 1/8 in which case the sky is practically considered as clear. With clear sky the precipitable water may vary considerably only with temperature. But in the sum- mer only minor temperature fluctuations are experi- enced. Finally, the month to month variation of 03 is small at the latitude of Athens in summer, being ~<0.03 cm (NTP) (Iqbal, 1983).

By taking the above into account, the transmit- tances (z) were calculated for the measurements of direct solar irradiance that corresponded to clear sky with the help of Equation (1). A value of 1367 W m-2 was taken for the solar constant (Fr6hlich et al., 1981) while the reciprocal of the square of the relative radius vector of the Earth (Eo) was estimated with the help of the equation proposed by Spencer (1971). The time sequences of their mean values in summer were found at 11:20 and 14:20 for the period 1962-1985. A graphic representation of these sequences together with fitted

Table 1. Variation of the relative optical air mass (m,) in the summer months, at 8:20, 11:20, 14:20 and 17:20

(LST)

Time (LST) June July August

08 :20 1 .69-1 .72 1.72-1.86 1.87-2.13 11:20 1 .07-1 .07 1.07-1.10 1.10-1.19 14:20 1 .13-1 .15 1.13-1.16 1.16-1.28 17:20 2 .19-2 .32 2.18-2.34 2.35-3.13

224 G.S . KARRAS et al.

curves is given in Figs 2a,b. It should be noted that the number of data for the whole season ranged between 30 and 66 at 11:20 and 22 and 63 at 14:20 in the 25 year period. It becomes very clear from both figures that from the beginning of the examined period until the early 1970s a continuous drop of the transmittance of direct solar irradiance occurs. From there on, al- though significant fluctuations do occur, the mean value does not show any systematic change except for the last 4-5 years, where a slightly increasing tendency seems to exist. The same behaviour is also exhibited by the trend of the transmittance when the data for each of the two times is examined separately in each month. This can be verified in Figs 3a,b where the trends are given for 11:20 (LST) of July and August. Obviously the greater fluctuations that ap- pear in these cases are due to the fact that the mean values for each year are based on a smaller number of data (the minimum number of observations in a year for the two cases are 7 and 10, respectively). At 8:20 and 17:20 the transmittance trends cannot be found due to the big variation of the relative air mass at these times. Nevertheless, the transmittance trend at the hour of 8:20 in July, where the variation of the relative air mass is small (Table 1), could probably be consider- ed as representative for the summer season at this

0.65

,~ 0 . 6 0 -

8 8 ..~ 0 . 5 5 -

t - 0 . 5 O -

11:20 LST

045 [ I I i I 60 65 70 75 80 85

Yeor

0.65

060

o 0.55

~ (150

0.45

14 : 20 LST

I I I I I 65 70 75 80 85

Yeor

Fig. 2(a,b). Time sequence of the mean values of transmit- tance (T) for the summer period, at 11:20 and 14:20 LST, respectively. A second degree polynomial curve has been

fitted in each case (dashed line).

hour. Figure 4 shows similar behaviour as in Figs 3 and 2, but the transmittance values are smaller from those at 11:20 and 14:20 as expected from the bigger air mass values. The number of observations on which the mean values in each year were based is generally greater in this case since the skies tend,to be clear in the morning in summer. The minimum number of observations in a year is 16.

By considering the general behaviour of the trans- mittance trend as has been presented above, it be- comes obvious that the air quality of the atmosphere has deteriorated in the period 1962-1976. If in the Figs 2a, b the values obtained by the regression lines are

0.65

Q60

o 055

c

(150

JuLy

f

045 I I I I [ 60 65 70 75 80 85

Yeor

0.65

A 0.60

E

o55 <_, .g c

e I-- 05o

August

/

0.45 I I I I I 65 70 75 80 85

Yeor

Fig. 3(a,b). As in Fig. 2(a,b), but at the hour 11:20 LST for July and August, respectively.

055 i

0.543 l -

045

I - 040

JuLy 08 : 20 LST

o3~" I I I I I 65 70 75 80 85

Yeor

Fig. 4. As in Fig. 2(a,b), but at the hour 08:20 LST for July.

Transmittance of direct solar radiation 225

taken into account the transmittance of the direct solar irradiance appears to have decreased by about 0.6 in the clear summer days during this period. Since such a decrease is quite unlikely to have been caused by a climatic change it should be attributed to the rapid urbanization of the greater Athens area in this period. Finally, in the period between 1980 and 1985 a slightly increasing tendency accompanied by rela- tively strong fluctuations seem to appear. These might have been the result of a number of restrictions which have been imposed on the circulation and the in- dustrial activity in this period but which however had not acquired their final form until 1985.

REFERENCES

Bach W. (1973) Solar irradiation and atmospheric pollution. Arch. Met. Geoph. Biokl. Set. B 21, 67-75.

Fr6hlich C. and Brusa R. W. (1981) Solar radiation and its variation in time. Sol. Phys. 74, 209-215.

Hodges L. (1977) Environmental Pollution. Holt, Rinehart and Winston, New York.

Iqbal M. (1983) An Introduction to Solar Radiation. Academic Press, Canada.

Kasten F. (1966) A new table and approximate formula for relative optical air mass. Arch. Meteorol. Geophys. Bio- klimatol. Ser. B 14, 206-223.

Lalas D. P., Karras G. S., Pissimanis D. K., Notaridou V. A. and Kassomenos P. (1983) Models to predict meteorologi- cal conditions conducive to air pollution episodes--Phase II. Ministry of Physical Planning, Housing and the En- vironment PERPA, E.E.C. DG XI. Contract B6612/9.

Spencer J. W. (1971). Fourier series representation of the position of the Sun. Search 2, 172.

Wesely L. M., and Lipschutz R. C. (1976). An experimental study of the effects of aerosols on diffuse and direct solar radiation received during the summer near Chicago. At- mospheric Environment 10, 981-987.

AE(B) 24:2-D