Impacts of the solar eclipse of 29 March 2006 on the surface ozone and nitrogen dioxide concentrations at Athens, Greece

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. Impacts of the solar eclipse of 29 March 2006 on the surface ozone and nitrogen dioxide concentrations at Athens, Greece C. Tzanis, C. Varotsos, L. Viras

experimental data demonstrated that the solar eclipse phenomenon affects the surface ozone and nitrogen dioxide concentrations as well as the temperature, the relative humidity and the wind speed near the ground. The reduction of the solar ultraviolet radiation at 312 and 365 nm reached 97% and 93% respectively, while the air temperature dropped, the relative humidity increased and the wind speed decreased. The 10 percentage change (decrease) of surface ozone concentration was maximized one hour after the maximum phase of the eclipse due to the decreased efficiency of the photochemical ozone formation. The surface nitrogen dioxide concentration increased and the time lag of the nitrogen dioxide response to the solar eclipse was found to be different for each station. A plausible cause for the increase in NO 2 concentration may 15 be the conversion of NO to NO 2 through reaction with pre-existing O 3 along with the low photolysis rates of NO 2 as a consequence of the decreased solar radiation during the solar eclipse event. In general, the time response to the eclipse phenomenon was different for each of the aforementioned parameters.

20
The solar eclipse being a rare natural phenomenon gives an opportunity to investigate how the photochemical processes react to the comparatively fast solar radiation changes.
The plausible variations in the stratospheric composition caused by natural processes like a solar eclipse have been among the most attractive issues for many work-25 ers (Bojkov, 1968;Wuebbles and Chang, 1979;Elansky et al., 1983;Burnett and 14332 balance. However the number of studies exploring the effects of solar eclipse on surface temperature and surface winds as well as on surface ozone (SOZ) and its precursors is relatively small (Srivastava et al., 1982;Fernadez et al., 1993;Zerefos et al., 2001;Kolev et al., 2005;Tzanis, 2005).
The experimental data obtained from different observational sites in Bulgaria during the solar eclipse of 11 August 1999, demonstrated that the influence of the phenomenon was manifested with a certain delay and the maximal impact of the eclipse on the meteorological parameters (temperature, humidity and wind speed and direction) and the ground ozone concentration was revealed to be 7 and 10 min after the maximal 15 eclipse, respectively (Kolev et al., 2005).
A decrease of around 10-15 ppbv in SOZ concentration has been observed at Thessaloniki, Greece, during the solar eclipse of 11 August 1999(Zerefos et al., 2001, while the percentage change of SOZ concentration at Athens, Greece was maximized one hour after the solar eclipse maximum and the greater values of SOZ percentage 20 change were observed at the Patision station, an urban station located in the central part of the Athens basin (Tzanis, 2005). The ozone profile measurements over Thessaloniki during the solar eclipse of 11 August 1999 indicated also an ozone decrease up to 2 km with a lag-time between the maximum of the eclipse and the maximum of the induced ozone decrease (Zerefos et al., 2001). 25 The effect of various meteorological parameters on the variability of the surface ozone and its precursors in the Greater Athens area (a site in the Mediterranean region, where very frequent photochemical pollution episodes occur), has been discussed in a number of recent publications (Cartalis and Varotsos 1994;Kondratyev Introduction and Varotsos 2001a,b;Ziomas et al., 1998;Varotsos et al., 2001a,b;Varotsos et al., 2003). Although the most important chemical mechanisms involved in photochemical pollution have been already identified and studied, further investigation is necessary because this atmospheric phenomenon is a very complex process involving meteorological, topographic, emission and chemical param-5 eters (Varotsos et al., 2004).

EGU
In this work we examine the behavior of surface ozone and nitrogen dioxide concentration as well as the variations in various meteorological parameters (temperature, relative humidity and wind speed and direction) during the solar eclipse that took place on 29 March 2006 over Athens, Greece (38 • N, 23 • E).

Data
We have used the measurements of SOZ and nitrogen dioxide concentration along with meteorological measurements, taken at four monitoring stations (Patision, Smyrni, Geoponiki and Zografou) of the local air pollution monitoring network (National Service for Air Pollution Monitoring). The four sites are located into the greater area of the 15 Athens basin as follows: Patission site in a street of heavy traffic in the centre of the city, Smyrni site at about 5 Km SE of the centre of the city in an area of low traffic, Geoponiki site at about 2 Km SW from the centre of the city in an area of low traffic and light industrial activities and Zografou site at about 3 Km NE of the centre of the city in an area of low traffic inside the campus of the Athens University. SOZ and nitrogen dioxide 20 measurements were made by employing conventional analyzers (chemiluminescence and UV absorption, respectively) with time resolution 30 s and accuracy ±1 ppb. The selection of the above mentioned monitoring stations was made on the basis that their respective locations give a representative picture of the urban area since they cover four opposite sections. 25 Solar ultraviolet radiation (SUVR) measurements at 312 and 365 nm were carried out during the eclipse event by using a VLX-3W (Vilber-Lourmat, France) radiometer Introduction EGU equipped with two sensors (CX-312 and CX-365). In addition, the MICROTOPS II sun-photometer (Solar Light Co., Inc.) was used for measurements of SUVR-B at 312 nm. This filter instrument is a continuation of a series of hand-held ozonometers with TOPS at the beginning. The new generation of those instruments is MICROTOPS II, that is a 5-channel sun-photometer with cen-5 tre wavelengths of 300, 305, 312, 940, and 1020 nm for measurements of TOZ, total water vapour and aerosol optical thickness measurements (Kondratyev and Varotsos, 2000). With this instrument TOZ is derived from measurements for three wavelengths in the UV region, given the site's latitude and longitude, universal time, altitude and pressure. As in Dobson instrument, the measurement at an additional third wavelength enables a correction for particulate scattering and stray light . The total water vapour is determined through the measurements at 940 nm and 1020 nm. The angle of view of each of optical channels is 2.5 • and the resolution is 0.001 µWcm −2 . The typical agreement between various MICROTOPS II instruments (accuracy) is within 1-2%. The repeatability of consecutive ozone measurements is 15 better than 0.5%. A 21-months intercomparison of the MICROTOPS II filter ozonometer with the Dobson and Brewer spectrometers resulted that Mtops can measure TOZ with an accuracy comparable to the conventional spectrometers (agreement is better than ±1%), over a reasonable range of µ. Adverse conditions (clouds, haze, and low sun) result in deviations of more than ±2% or even ±3% (Kondratyev and Varotsos, 20 2000).

Discussion and results
The

EGU
Athens basin. The expected values were calculated by applying a 6th degree polynomial fit (best fit) to the observed SOZ concentrations when the eclipse event was absent. All percents in Table 1 were calculated by subtracting the observed SOZ concentrations just before, during and just after the eclipse event from the expected ones. According to Fig. 1 and Table 1, the percentage change of SOZ is maximized one hour 5 after the solar eclipse maximum at each of the stations. The greater values of SOZ percentage change are observed at the Patision station, an urban station located in the central part of the Athens basin. This is in perfect agreement with SOZ measurements before, during and after the solar eclipse of 11 August 1999 at Patision, Geoponiki and Smyrni stations as it shown in Fig. 2 and Table 2 (Tzanis, 2005). The above 10 mentioned behavior of SOZ during the solar eclipse may be related to photochemical processes due to the fact that the gradual decrease in the solar radiation affects the photochemical reactions within the planetary boundary layer. The decrease in SOZ concentration started a few minutes after the beginning of the eclipse and maximized after the solar eclipse maximum as a consequence of the further fall in sunlight inten- 15 sity that decreased the efficiency of the photochemical ozone formation (decrease in the net ozone production rate). The solar radiation started to increase after the eclipse totality while the SOZ concentration started to increase about one hour later and returned to its ordinary behavior several minutes after the end of the solar eclipse. The SOZ variations detected at Athens, before, during and after the solar eclipse of 29 20 March 2006 demonstrated that the influence of the phenomenon was manifested with a certain delay. The observed SUVR measurements at 312 and 365 nm as derived from the MICRO-TOPS II sun-photometer and the VLX-3W radiometer, before, during and after the solar eclipse of 29 March 2006 at Athens are shown in Figs. 3 and 4. As can be seen, the 25 percentage decrease of SUVR-B (312 nm) which was measured by the MICROTOPS II and VLX-3W instruments reached the value of 97% and 93% respectively, during the solar eclipse maximum while the change in SUVR at 365 nm was about 93%. This is in close agreement with the observed reduction in the incoming solar radiation as de-

Printer-friendly Version
Interactive Discussion EGU rived from measurements conducted at Thission station (center of Athens) by Founda et al. (2007). Figure 5 presents the march of surface nitrogen dioxide concentration before, during and after the solar eclipse as well as the nitrogen dioxide variations observed on 28 March 2006, a day with similar meteorological conditions, for the stations of Patision, 5 Smyrni, Geoponiki and Zografou. The data presented in Fig. 5 show that in site of the similar diurnal variability the levels of NO 2 are different, which may be an impact of emission changes due to "eclipse show". The main conclusion from this figure is that NO 2 concentrations start to increase after the beginning of the eclipse event at Patision, Geoponiki and Zografou stations, while at the station of Smyrni NO 2 starts to increase 10 just after the maximum of the solar coverage. After the solar eclipse event NO 2 started to decrease reaching the values observed on 28 March 2006. From Fig. 5 it is evident that the solar eclipse affected the nitrogen dioxide concentration and the time lag of the NO 2 response with respect to the first contact was different for each station. A possible explanation for the above mentioned behavior of NO 2 concentration during the solar 15 eclipse may be the conversion of NO to NO 2 through reaction with pre-existing O 3 along with the low photolysis rates of NO 2 as a consequence of the decreased solar radiation. Comparison of Figs. 1 and 5 shows that the observed reduction of SOZ concentration after the beginning of the solar eclipse is accompanied by an increase of NO 2 concentration at each station. Furthermore, the low photolysis rates of NO 2 20 may be the most possible reason for the reduced surface ozone values during the eclipse event due to the fact that the primary photochemical source of ozone in the troposphere is photolysis of NO 2 in the presence of molecular oxygen (Kondratyev and Varotsos, 2000). It is worth noting that the same conclusions for NO 2 were derived, when applying the analytical method described above for ozone. 25 Figure 6 shows the temporal variation of the relative humidity, temperature and wind velocity during the eclipse at Patision station. Air temperature near the ground decreased from 20.1 • C to 19.4 • C during the eclipse event and after that it began to increase abruptly. Relative humidity started to increase at 13:00 (55%) as a result of the EGU temperature decrease reaching a maximum (58%) at the end of the solar eclipse. The fact that the wind speed started to decrease after the solar eclipse maximum, may be attributed to the cooling and the stabilization of the atmospheric boundary layer. The wind speed continued to decrease for two hours after the eclipse event without any significant change in direction. Kolev et al. (2005)

10
During the solar eclipse of 29 March 2006 measurements of surface ozone and nitrogen dioxide concentration, solar ultraviolet radiation and meteorological parameters (relative humidity, temperature and wind velocity) were performed at four sites (Patision, Smyrni, Geoponiki and Zografou) into the greater area of the Athens basin in Greece. 15 As expected, all the parameters mentioned above were affected by the solar eclipse. The percentage decrease of the SUVR at 312 and 365 nm reached 97% and 93% respectively at the maximum phase of the eclipse.
The SOZ concentration decreased during the eclipse event and the maximum percentage change observed one hour after the maximum of the solar eclipse at all sta-20 tions. The greater values of SOZ percentage change were observed at Patision station (urban station) located in the center of Athens. The decrease in SOZ concentration was attributed to the dramatic reduction of the solar radiation that affects the photochemical reactions.
The surface nitrogen dioxide concentration increased at all stations as the solar 25 eclipse phenomenon progressed while the time lag of the NO 2 response to the eclipse was different for each station. This increase in NO 2 concentration may primarily be at-Varotsos, C., Alexandris, D., Chronopoulos, G., and Tzanis, C.: Aircraft observations of the solar ultraviolet irradiance throughout the troposphere, J.