Annular Solar Eclipse on 26 December 2019 and its Effect on Trace Pollutant Concentrations and Meteorological Parameters in Kannur, India: a Coastal City

This paper highlights the variations of surface ozone (O3), total column ozone (TCO), oxides of nitrogen (NO and NO2), carbon monoxide (CO), sulphur dioxide (SO2), ammonia (NH3), volatile organic compounds (Benzene, Tolune, Ethyle Benzene, Xylenes (collectively called BTEX)), particulate matters (PM10 and PM2.5), and meteorological parameters at the time of an annular solar eclipse on 26 December 2019 at Kannur town in Kerala, South India. The maximum solar obscuration has resulted a decrease in solar radiation by 93%, air temperature by 16.3%, wind speed by 36.1% and an increase in relative humidity by 27.1% at this coastal location. Along with the reduction in solar radiation, the concentration of surface O3 (61.5%) and total column O3 (11.8%) have been observed to decrease at the maximum phase of solar eclipse. CO and NO2 concentration were found to be increased by 28.9% and 42.2%, respectively, while NO exhibited its typical diurnal variation. Further, a decrease in concentrations of SO2 by 17.6%, PM10 by 18.5%, and PM2.5 by 11.3% were observed. NH3 and BTEX were found to be higher than 11.3% and 22.6% of the concentrations in control days. All of these deviated parameters could be seen returning to their normal state after completing the eclipse episode. The variation of photodissociation coefficient j(NO2) values were theoretically calculated from the observed data, which shows a good agreement with the model simulated j(NO2) reduction. This is an extensive second observation on the variation of trace pollutants on solar eclipse, after the partial solar eclipse observed on 15 January 2010 at Kannur.


INTRODUCTION
Solar eclipse is a natural phenomenon that provides a distinct reduction of solar radiation for a fixed duration due to the obscuration of the solar disc by the moon.Being a rare astronomical event, it provides a unique occasion to investigate the spontaneous changes in photochemistry, dynamics, and boundary layer height (BLH) in the atmosphere due to the variations in the ascending and descending phases of solar radiation (Stoev et al., 2005;Szalowski, 2002;Koepke et al., 2001;Winkler et al., 2001;Fernandez et al., 1996;Manohar et al., 1995).Several investiga-tors have reported the effects of various atmospheric parameters at the time of a solar eclipse in different parts of the globe (Gröbner et al., 2017;Dutta et al., 2011;Krezhova et al., 2008;Aplin and Harrison, 2003;Dolas et al., 2002;Foken et al., 2001;Vogel et al., 2001;Hanna, 2000;Prenosil, 2000;Anderson, 1999;Eaton et al., 1997;Segal et al., 1996).Various researchers have explored the pronounced variations in the atmospheric boundary layer (ABL) height at the eclipse event (Ratnam et al., 2010;Founda et al., 2007;Krishnan et al., 2004;Kapoor et al., 1982;Raman, 1982).Subrahamanyam et al. (2012) observed a double mixed boundary layer between 700 and 1500 m during the solar eclipse on 15 January 2010 at Thiruvananthapuram, which was due to the horizontal advection of air mass.The atmospheric turbulence decay during the total solar eclipse of 11 August 1999 was examined by Anfossi et al. (2004), and their observations were in good agreement with the theoretical model of turbulence decay.
Studies were focused on the eclipse induced surface air temperature variations and reported a sharp decrease in air temperature at the maximum phase of the eclipse (Calamas et al., 2019;Nymphas et al., 2009;Kolarz et al., 2005;Krishnan et al., 2004;Sethuraman, 1982).Wang and Liu (2010) observed a warming of 7°C at 17 km altitude due to thermal contraction, and cooling of 2°C in the troposphere due to reduction in solar heating during the total solar eclipse on 22 July 2009.Likewise, a temperature reduction of 2°C was observed around the tropopause over Thumba by Subrahamanyam et al. (2011) at the annular solar eclipse of 2010, and revealed that these changes were associated with a significant reduction in turbulent kinetic energy, momentum flux, and sensi ble heat flux at the peak period of the eclipse.Observations of land surface temperature changes over Europe during the total solar eclipse on March 2015 by Elizabeth (2016) found to drop up to several degree Celsius, and this temperature drop was correlated with solar intensity variation.Muraleedharan et al. (2011) reported an abnormal warming in the upper troposphere due to decrease in tropopause height on the solar eclipse day 2010 at Goa, India.
Generally, humidity increases with the reduction in surface air temperature (Bhat and Jagannathan, 2012), and this cooling alters the stability of boundary layer in the lower atmosphere (Gorchakov et al., 2008;Praveena et al., 2004).Jokinen et al. (2017) identified the formation of new particles caused by aerosol loading and lack of oxidized organic compounds during the solar eclipse in 2015.Apart from the variations observed in meteorological parameters and atmospheric dynamics, solar eclipses induced substantial impact on the stratospheric and surface O 3 photochemistry, and several studies have been carried out in this direction (Gerasopoulos et al., 2008;Zanis et al., 2007;Tzanis, 2005;Fabian et al., 2001).Observations of total column O 3 using ground based devices during eclipse events by different groups show a decrease in O 3 concentration during the maximum phase of eclipse (Chudzynski et al., 2001;Zerefos et al., 2000;Mims and Mims, 1993).
As the sunlight is obscured for few minutes, the weather parameters fluctuate, resulting in a drastic change in the concentrations of atmospheric O 3 and nitrogen dioxide (Amiridis et al., 2007;Fabian et al., 2001).Manchanda et al. (2012) have reported a decrease in the O 3 concentration in the stratosphere and troposphere instantly after the maximum phase of the eclipse due to temperature cooling, on the solar eclipse of 15 January 2010 at Thir uvananthapuram.Only very few studies have been reported on the changes in the surface O 3 and their photochemistry during the eclipse event in India (Girach et al., 2012;Reddy et al., 2012;Vyas et al., 2012;Nishanth et al., 2010;Sharma et al., 2010;Naja and Lal, 1997;Srivastava et al., 1982).In addition, they also validated the model simulation response, which was in agreement with the observed effect.This is the second study on the variation of air pollutants on solar eclipse, after the partial solar eclipse observed on 15 January 2010 at Kannur; in which we could investigate the variations of surface O 3 , NO x and meteorological parameters at Kannur University campus.Here we describe the results on the variation of surface O 3 , NO, NO 2 , CO, SO 2 , NH 3 , particulate matters (PM 10 and PM 2.5 ), Volatile Organic Compounds include Benzene, Tolune, Ethyle Benzene, Xylenes (collectively called BTEX), total ozone column (TOC) and meteorological parameters (total solar radiation, temperature, relative humidity, wind speed and wind direction) at Kannur Town.Further, a detailed account of the inter-correlation between observed trace pollutants and meteorological parameters is discussed.In addition to these, theoretical calculation and model simulation for the photodissociation coefficient j(NO 2 ) is discussed in this paper.This investigation shows that at Kannur town, which is highly polluted than Kannur University Campus, the synoptic movement of atmospheric trace species plays a crucial role in the progress of eclipse induced changes in the surface O 3 and other trace pollutant.

1 Description of the Eclipse
The annular solar eclipse is a phenomenon in which the sun is partially or completely obscured when the moon comes between the sun and earth, causing the sun's outer surface appears as a ring.The annular solar eclipse of December 26, 2019 started from Hofuf in Saudi Arabia, Qatar, and UAE.It was then crossed over India, Sri Lanka, Malaysia, Indonesia, Singapore, and ended at Guam in the China Sea.Fig. 1(a) shows the path of solar eclipse over Indian region.At Kannur, the annular solar eclipse began at 08:04 IST, reached total obscuration at 09:26 IST and completed at 11:05 IST.The total eclipse lasted for 3 minutes and 12 seconds with a magnitude of 0.98, when the Moon came close to the centre of the Sun.The duration of the eclipse was 3 Hrs 1 minute at Kannur.

2 Observational Site
Kannur is the northern district of Kerala state, lying between the Arabian Sea coast in the west, and Western  The timings for the solar eclipse viewed over Kannur is shown in the Fig. 1(c).As the eclipse occurred in winter season, the environment was less humid with clear sky and wind was from north-west at this site.

3 Experimental Setup
Measurements of trace pollutants were carried out using the respective ground based gas analyzers from Environment S.A France.The measurements of surface O 3 were made using a continuous O 3 analyzer (Model O342e) with a detection limit of 0.2 ppbv.Its working principle is based on O 3 detection by direct absorption in UV light.O 3 absorption spectrum is intense in the 250 and 270 nm wavelength range.Thus, it corresponds to Total column O 3 (TCO) concentrations and the variations in solar UV radiation were recorded at regular intervals using the Microtop II ozonometer (Solar Light Co., Inc, USA).The MICROTOPS-II consists of two sets of instruments, namely sun photometer for aerosol optical depth measurement at five discrete wavelengths and an ozonometer, for measuring the total column O 3 (TCO) using three UV channels.This sun photometer measures the total O 3 at three different wavelengths 305.5, 312.5 and 320 nm in the UV region of the solar spectrum by using differential optical spectroscopic technique.
NO, NO 2 and NH 3 were measured with the aid of gas analyzer (Model AC32e) with a detection limit of 0.2 ppbv.Its working principle is based on the NO chemiluminescence in the presence of highly oxidizing O 3 molecules.The NO in the ambient air is oxidized by O 3 to form excited NO 2 molecules.The concentrations of NO, NO 2 were measured based on the spectrum of the radiation emitted by NO 2 molecules at the excited level.Particulate matters (PM 10 and PM 2.5 ) were measured by using suspended particulate beta gauge monitor (Model MP101M).Its working principle is based on the particle measurement by beta radiation attenuation.The measurement consists of calculating the absorption difference between a blank filter and a loaded filter, knowing that the beta ray absorption follows an exponential law and is independent of the physiochemical nature of the particles.
Measurements of CO were made by using an analyzer (Model CO12e) with a detection limit of 0.05 ppm.Its working principle is based on CO detection by absorption in infrared light.VOC's (BTEX) were measured, based on gas chromatography coupled with a PID detector by using (VOC72e) analyzer.SO 2 measurements were made by using a UV fluorescent Sulfur dioxide analyzer (Model AF22e) with a detection limit of 0.4 ppbv.The ambient air to be analysed is filtered by a hydrocarbon removing aromatic molecule device.The hydrocarbon molecule free sample to be analysed is sent to a reaction chamber, to be irradiated by an UV radiation centred at 214 nm, which is the SO 2 molecule absorption wavelength.All the gas analyzers have been calibrated by using sample gases on a regular basis.The total solar radiation was measured by LSI LASTEM Italia (DPA870) pyranometer and surface air temperature measured by an external Pt100 sensor.

1 Variation of Solar Radiation, Surface and
Total Column O 3 In order to study the influence of solar eclipse on the The intensity of solar radiation regained its normal value by the end of the eclipse (11:05 IST).It was found that the intensity of solar radiation declined from 376± 101 Wm -2 (average value on control days) to 26 Wm -2 on the eclipse day at 9.26 IST with a reduction of 93%.Likewise, surface O 3 concentration started to decrease from 9.3 pbbv at the beginning of the eclipse (08:15 IST) and reached its minimum value of 6.5 ppbv at 09:26 IST, and after full obscuration it started to increase till the end of the eclipse.During control days and eclipse day, surface O 3 concentration reaches its peak value at the noon time, 30 minutes after the maximum intensity of solar radiation, while the O 3 decline was in phase with solar radiation on eclipse day.Thus, a sharp decrease in O 3 concentration (61.5%) was noticed during the maximum phase of the eclipse at 09:26 IST when compared to the respective values in the control days.This decline of O 3 was due to the slow rate of photolysis of NO 2 in the declining phase of the solar intensity as shown in the figure.This is in tune with the similar observations at the maximum phase of solar obscuration during the eclipse on 15 January 2010, at Kannur University campus (Nishanth et al., 2011), and at Thiruvananthapuram (Girach et al., 2012).
Studies carried out at various locations during different solar eclipses have shown different magnitudes of change in the concentration of surface O 3 .Naja and Lal (1997) have observed a decrease of 18-21% at Ahmedabad during the solar eclipse of 24 October 1995 and Abram et al. (2000) have reported a decrease of 8% of O 3 at Silwood Park, Ascot during the solar eclipse of 11 August 1999.Tzanis et al. (2008) have observed a higher rate of change O 3 one hour after the maximum phase of the solar obscuration over four observation centres in Greece during the solar eclipse of 29 March 2006.
The day time variations of total column O 3 on the eclipse day and the control days are shown in Fig. 3. Column O 3 concentration was found to increase during day time when photochemical reactions occurs in the presence of solar radiation, N 2 O 5 , and nitrogen oxides such as NO, NO 2 and NO 3 (Kostadinov et al., 1999).Actinic flux in troposphere may have a strong reduction by stratospheric O 3 absorption by which photochemistry in the troposphere depends on stratospheric O 3 concentrations (Madronich and Flocke, 1999).In high NO x environments, higher photo-dissociation rate coefficients (j) result in tropospheric O 3 production; while in low NO x regions, O 3 destruction is enhanced (Fuglestvedt et al., 1995;Madronich and Granier, 1992;Liu and Tranier, 1988).
The average concentration of total column O 3 varied between 248±4.4 to 285±8.2DU during control days, and it showed a lower concentration on eclipse day.From the figure, it is clear that the total column O 3 concentration was found to be decreasing from the beginning of the eclipse (08:04 IST) and reached the minimum during the maximum phase of the eclipse and then returns to normal value after the eclipse event.With respect to control days, TCO was decreased from 262±6.6% to 231±5.3% in a time of 10 minutes after the maximum phase of solar obscuration.Thus, a change of 31 DU which correspond to 11.8% change in TCO was observed.A similar variations were reported at Thiruvananthapuram (Kumar et al., 2014) and New Delhi (Chakrabarty et al., 2001) during the eclipse events in 2010 and 1995.Chakrabarty et al. (1997) observed a sharp decrease in total column ozone before 10 minutes of maximum obscuration of the Sun over Ahmedabad during the solar eclipse of 1997.Using MICROTOPS Sun photometer, Kazadzis et al. (2007) reported that a decrease in 30-40 DU total column ozone in Kastelorizo, Greece on the day of eclipse on 29 March 2006.The measurements of total column ozone using MICROTOPS II Ozone Monitor by Girach et al. (2012) show a decrease of 14 DU at the maximum phase of eclipse at Thiruvananthapuram during 2010.Further they obs erved a fluctuation of columnar ozone about 20 minutes after the maximum phase of eclipse due to the presence of wave activity.Thus, the observed variation in TCO over Kannur, during the eclipse needs a detailed investigation incorporating fluid dynamics, and photochemical models to explore the boundary layer dynamics and associated vertical mixing.the prominent precursors of O 3 , their concentrations were found to be lower in the afternoon (around 14:00 IST) when O 3 concentration became a maximum.Several researchers reported a similar diurnal vari ation O 3 and its precursors in many urban areas in India (Verma et al., 2017;Mahapatra et al., 2014;Beig et al., 2007;Naja and Lal, 2002).
The day time variations of CO, NO and NO 2 for control and eclipse days are shown in the Fig. 4(a, b, c), respectively.During control days, the concentration of  CO varied from 140±56 to 633±145 ppbv, and that of NO and NO 2 varied from 2.01±0.58 to 6.78±1.44 ppbv, and 3.34±1.1 to 10.54±1.2ppbv, respectively.This clearly indicates the photochemical production of O 3 in the presence of NO 2 from CO, CH 4 and VOC's (Nair et al., 2018;Tyagi et al., 2016;Nishanth et al., 2014).A considerable reduction in solar radiation during the eclipse delayed the atmospheric chemical reactions which resulted in a reduced photochemical production of surface O 3 by NO 2 .The instant consequence of eclipse is earmarked by a decline in j(NO 2 ) in tune with the in solar radiation.Thus, the theoretical calculation along with a photochemical box model simulation were used to determine the j(NO 2 ) variation for control and eclipse days and the details are discussed in section 3.6.An increase in CO concentration was observed from the beginning of eclipse (590 ppbv), and it reached a maximum value (728 ppbv) during the maximum obscuration.Thus, a sharp increase in CO concentration (28.9%) was observed during the maximum phase of the eclipse (09:26 IST) when compared to the respective time in the control days.Further, the concentration of NO 2 was found to be increased by 42.2% during the maximum phase of the eclipse with respect to control days, while NO concentration remained unchanged during the maximum phase of the eclipse.Similar variation of NO, NO 2 and CO were observed at Thumba in Thiruvananthapuram (Girach et al., 2012) and Mangattuparamba in Kannur (Nishanth et al., 2010) during the eclipse in 2010.This increase in NO 2 and CO may be influenced by the geography, morning traffic and it may have significant influence on the variation of boundary layer dynamics.This needs to be analysed in detail with the aid of spatially separated observations during the  The day time variations of SO 2 , NH 3 and BTEX for control and eclipse days are shown in the Fig. 4(d, e, f), respectively.The diurnal variation of SO 2 exhibited two peaks during morning and evening hours of the control days due to heavy traffic and it varied from 1.2±0.52 to 3.51±0.62ppbv.Natural sources of sulphur in the atmosphere comes mainly from hydrogen sulphide and dimethyl disulphide.As the solar intensity attains its maximum value during day time hours, SO 2 concentration shows relatively low level due to efficient oxidation to sulphate molecules through the oxidization of O 3 and OH radicals (Margitan, 1984).With respect to control days, the concentration of SO 2 was low during eclipse day.Fig. 4(d) shows a decline of SO 2 from 3.13±0.82to 2.58 ppbv (17.6%) with respect to control days at the maximum phase of eclipse episode.This decrease in SO 2 concentration observed during the eclipse may be one of the causes for the decrease observed in O 3 concentration.It was further observed that the concentration of SO 2 regained its same concentration as that of the control days, nearly after four hours of completion of the eclipse event.
Daily NH 3 exhibited a temporal variation with higher concentrations during early morning and late evening hours, due to the accumulation of air pollutants by lower wind speeds (Zhao et al., 2016;Wang et al., 2015).During control days, on diurnal basis the mixing ratios of NH 3 varied from 5.06±0.38 to 6.16±0.42ppbv.During eclipse day, the concentration of NH 3 started increasing from the beginning of the eclipse, and reaches its higher concentration during the maximum phase of eclipse.With respect to control day's average concentration, NH 3 increased from 5.32±0.32 to 5.92 ppbv during the maximum phase of solar obscuration.It leads to a change of 11.3% of increase in NH 3 concentration at the maximum phase of the eclipse.The night time chemical reaction may have been taking place to increase the concentration of NH 3 during the eclipse episode.This may be the prime reason for the observed enhanced concentration of NH 3 during the eclipse episode.Further, the lowering of boundary layer height during the eclipse episode may also contribute higher mixing ratio of NH 3 over the observational site.tha (2012); Zhao et al. (2016), have reported that the concentration of NH 3 in the night is increasing in urban and rural areas.Benzene, toluene, ethyl benzene, xylenes (BTEX) are the major hydrocarbons found in the atmosphere in the category of volatile organic compounds (VOC's) which play a vital role in the production and photochemical reaction of O 3 and other oxidants in the lower atmosphere (Sahu and Saxena, 2015;Yassaa et al., 2011;Srivastava et al., 2005;Sillman, 1995).VOC's are able to enha nce the oxidising capacity of the atmosphere by pho to chemical reactions with hydroxyl radicals (OH) in the presence of NO x (Lee et al., 2002;Atkinson, 2000).
At a high ratio of VOC to NO x environment, OH will react mainly with VOC's and at a low ratio, OH will react with NO x (Senfield and pandis, 2006).The diurnal variation of BTEX exhibited an increasing trend during late night to early morning and morning hours.Further, it shows a decreasing trend during afternoon (12:00 to 18:00 IST) hours.A similar type of variation was reported by Tiwari et al. (2010) in Yokohama city, Japan, Torre et al. (2015) in Monterrey, Mexico and Yurdakul et al. (2017) in Bursa, Turkey.The observation revealed the fact that the variation in total concentration of BTEX showed not much changes during eclipse period.The concentrations were found to be increased from 8.08± 1.52 to 9.92 ppbv at the maximum phase of solar obscuration and this enhancement resulted a change of 22.6% with respect to control day's average value.The observed high concentrations of BTEX at the maximum phase of solar obscuration may be due to stable atmospheric condition and the absence of photochemical reactions.

3 Variations of Particulate Matters
Particulate Matters (PM 10 and PM 2.5 ) are considered to be one of the significant components in the atmosphere, present in the form of solid and liquid suspensions.The primary source of PM in a non-industrialised location is the exhaust of automobiles (Liu et al., 2015;Srimuruganandam and Nagendra, 2011;Cheng et al., 2006;Latha and Badarinath, 2005).Concentration of particulate matters are depending on the variation of boundary layer height, wind speed and direction, surface solar radiation, cloud cover, sea level pressure, and precipitation.Larger boundary layer and stronger surface wind gust help speedy dispersion of particulate matters (Kim and Kim, 2020;Yang et al., 2016).From night to early morning hours, atmosphere is almost stable without significant mixing which provide a favorable condition for the accumulation of particulate matters.With an increase in solar radiation in the afternoon, different environmental features mentioned above tend to accelerate the dispersion of particulate matters (Miao and Liu, 2019;Kim et al., 2017).Day time variations of PM 10 and PM 2.5 observed during eclipse day, and control days over Kannur are shown in Fig. 5.Diurnal variations of PM 10 and PM 2.5 shows higher concentrations during morning (09:00 to 12:00 IST) and evening (16:00 to 20:00)  hours due to the peak traffic and late-night hours due the accumulation of pollutants by shallow boundary layer.During eclipse day, the concentrations of PM 10 and PM 2.5 started increase from the beginning of the eclipse, and reaches higher concentrations during the maximum phase of eclipse.The magnitude of PM 10 increased from 94.88±18.44 to 112.4 μg/m 3 and PM 2.5 from 62.52± 14.04 to 69.56 μg/m 3 with percentages of increase 18.5% and 11.3%, respectively.This enhancement in PM 10 and PM 2.5 may be influenced by the shallow boundary layer and relatively low wind speed during the eclipse.

4 Variations of Meteorological Parameters
In order to understand the impact of eclipse on terrestrial meteorological parameters, the daily variations of surface air temperature, relative humidity, wind speed and wind direction were analysed in detail.The required meteorological parameters were measured by using an automatic weather station (LSI LASTEM) established in the observation centre during the eclipse day and 6 control days.The diurnal variations in temperature, relative humidity, wind speed and wind direction observed on the control and eclipse days are shown in the Fig. 6.The eclipse proscribes sun light incident on the Earth's surface, and that could significantly affect the surface air temperature, and relative humidity distribution.Consequently, observed air temperature was decreased from 29.13±1.35°C(with respect to control day's average) to 25.04°C with a temperature drop of 4.09°C during maximum phase of the eclipse.This temperature dip was 16.3% with a time lag of 17 minute at the full phase of the eclipse.The diurnal variation of temperature is shown in the Fig. 6(a).The reduction in temperature during eclipse episode induced a corresponding increase in relative humidity as shown in the Fig. 6(b).The relative humidity was found to be increased from 60.41± 4.2% (with respect to control day's average) to 76.82% during the maximum phase of solar obscuration.It resulted a change of 27.1% of increase in RH with a time lag of 18 minutes at the maximum phase of the eclipse.Namboodiri et al. (2011) have reported a temperature dip of 3.5°C, and 19% increase in relative humidity at a peninsular Indian coastal station (Thumba) during an eclipse on 15 January 2010.
The wind speed was found to be decreased to 36.1% with respect to control days at the maximum phase of ecli pse (Fig. 6(c)), while wind direction remained unchanged during eclipse episode (Fig. 6(d)).The observed wind speed was found to decrease during the starting of eclipse, and reached its low at the full obscuration, and tended to increase and regain its normal value after completing the eclipse episode.This decrease in wind speed might be due to the freezing of atmospheric boundary due to the absence of solar radiation (Sharma et al., 2010).Certainly, these deviations in weather parameters will persuade the distresses in the photochemistry of trace gases in the lower atmosphere.
The concentrations of O 3 precursor gases significantly varied during the solar eclipse episode, and the percentage of variations are listed in the Table 1.The observed NO concentrations did not show any substantial variation compared to the control days which reveal the absence of titration of O 3 leading to its destruction.From the table, it is evident that, 93% of reduction in solar radiation resulted in a decrease in temperature by 16.3% and O 3 concentration by 61.5% at the maximum phase of eclipse.At the same time NO 2 , CO, and RH showed an enhancement by 42.2%, 28.9% and 27.1%, respectively.

5 Correlation between Air Pollutants and
Meteorological Parameters Statistical analysis (Pearson's test) is performed, and the resulted correlation coefficient between observed trace gases, particulate matters and meteorological parameters over the study area during eclipse episode is shown in the Table 2.The results point out that O 3 shows a positive correlation between NO, NH 3 , solar radiation, temperature, wind speed and negative correlation with NO 2 , CO, SO 2 , PM 10 , VOC's (BTEX) and relative humidity.A good positive correlation between O 3 and solar radiation (r = 0.90) indicates the production of O 3 by the photolysis of NO 2 .Further, the correlation between O 3 and wind speed (r = 0.68) revealed the transport of O 3 over the observational site.Meteorological parameters such as solar radiation and temperature are in positive correlation with NO, and negative correlation with NO 2 , CO, SO 2 , NH 3 , PM 10 and VOC's.The negative correlations existing between wind speed with SO 2 , NO 2 , CO and PM 10 point out the reduction of these pollutants by the wind.Relative humidity shows positive correlations with NO, SO 2 and negative correlations with O 3 , NO 2 , CO, NH 3 , PM 10 , VOC's and solar radiation.PM 10 shows a negative correlation (r = -0.76)with temperature, which is due to the thermally induced decreased convection of particulate matters and condensation of VOC s.Similarly PM 10 shows a negative corre-lation with (r = -0.67)with relative humidity.

6 Theoretical Calculation and Model Simulation
for j(NO 2 ) Photodissociation rate coefficients (j values) is one of the vital parameters that unfolds the chemistry of air pollution.They depend on absorption cross sections, quantum yield and also the availability of solar radiation at any specific location in the atmosphere.The photolysis rate coefficient j(NO 2 ) at a location is est imated by the equation j(NO 2 ) = ∫F(λ)σ(λ,T)φ(λ,T)dλ, and the temporal variation of j(NO 2 ) is shown in Fig. 7(a).It is found that the j(NO 2 ) started to increase in the morning of the eclipse day, just as in the normal day.Its value gradually reduced in tune with the intensity of solar radiation in its descending and ascending phases of the eclipse and reached its minimum level during the complete obscuration.Thus, a change in the j(NO 2 ) was found in tune with the observed variation of surface O 3 .The photolysis rate coefficient j(NO 2 ) was found to be reduced by about 87.5% during the maximum phase of the eclipse with respect to the control days.This result reveals that the photochemical production of surface O 3 from NO 2 in the presence of sunlight.Therefore, it is further realised that surface O 3 and j(NO 2 ) found a positive relation during day time hours.
The Tropospheric Ultraviolet-Visible (TUV) model was used in the wavelength range 121-750 nm, for calculating the spectral irradiance, spectral actinic flux, and photodissociation coefficients (j-values).The TUV model simulations of j(NO 2 ) were carried out to study the influence of solar radiation on the photochemical production of O 3 from NO 2 over the observational period.More details about the model is described by Madronich and Flocke, (1999).The model simulations were carried out by keeping all the input parameters kept constant in their normal concentrations, and by varying the intensity of solar radiation from lower to maximum.The model simulated of j(NO 2 ) variations are shown in Fig. 7(b) for the control days and eclipse day.From the figure, it is evident, that similar to theoretical calculation, j(NO 2 ) started to decline from its normal value of 1.19 × 10 8 molecules cm -3 sec -1 (beginning of eclipse at 08:06 IST) to 0.245 × 10 8 molecules cm -3 sec -1 (maximum phase of the eclipse at 09:26 IST).The decline in the (jNO 2 ) during the maximum phase of eclipse is calculated to be 91.65% with respect to control days at the same time.After completing maximum obscuration, the j(NO 2 ) again started to increase and reaches its normal value later completing the eclipse episode at 11:05 IST.

CONCLUSION
A total solar eclipse provides a rare opportunity to study the effects of atmospheric changes due to the gradual dimming of solar radiation in the daytime.At Kannur, the annular solar eclipse began at 08:04 IST, reached maximum obscuration at 09:26 IST and completed at 11:05 IST.The highlights of the observation are the following: ▶ At the maximum phase of eclipse, solar radiation displayed a sharp decline (93%), which resulted a strident reduction (61.5%) in surface O 3 concentration due to reduced photochemical activity.The observation of total column O 3 using MICROTOPS II recorded a decrease of 31 DU in a time lag of 10 minutes from the maximum phase of the eclipse.▶ Increasing trends in ground level CO and NO 2 were observed from the beginning of eclipse, and they reached their maximum concentrations during the peak of the eclipse.This enhancement in CO and NO 2 concentration were estimated to be 28.9% and 42.2%, respectively, and this increase might be due to the reduced photochemical reactions, lower boundary layer height, low wind speed and morning traffic.NO concentration remains unchanged during the maximum phase of eclipse.▶ Since O 3 acts as a major oxidising agent for sulphur, consequently a decreased concentration of SO 2 was observed on the day of eclipse.With respect to control day's concentration, SO 2 decreased to 17.6% and NH 3 increased to 11.3% during the maximum phase of solar obscuration.Volatile organic compounds collectively measured BTEX was found to increase 22.6% during the maximum solar obscuration.▶ Meteorological parameters such as surface air temperature, relative humidity, wind speed and wind direction were analysed on the day of eclipse.During the maximum phase of the eclipse, surface air temperature and wind speed were dropped to 16.3% and 36.1%,respectively, and relative humidity increased to 27.1%.The concentrations of PM 10 and PM 2.5 started increasing, and attained a peak during the maximum phase of the eclipse.This upsurge leads to 18.5% and 11.3% for PM 10 and PM 2.5 concentrations, respectively during eclipse episode.
▶ The variation of photodissociation coefficient j(NO 2 ) values were theoretically calculated from the observed data, and a reduction of 87.5% was found.This shows a good agreement with the model simulated j(NO 2 ) reduction of 91.65%.
The results of this study are preliminary at a polluted costal city in north Kerala during a solar eclipse.This study revealed a synoptic movement of atmospheric trace species over Kannur town, a polluted location compared to Kannur University campus in which the earlier observations were made in 2010.The observed abrupt changes in surface O 3 and other trace pollutants should be validated with more observations with the support of other prominent trace species including OH radical and organic samples to understand their role in O 3 chemistry and transport in the coming eclipse events.

Fig. 1 .
Fig. 1.(a) Eclipse path passing over Indian region (source: NASA tracks the path of the annular eclipse on Google Maps), (b) aerial view of observation centre and surroundings at Kannur town, (c) timings for the annular solar eclipse viewed from Kannur (The image taken by using Canon EOS 6D Mark II Camera).
Ghats mountain ranges in the east.Kannur is the sixthmost urbanized district in Kerala, and is one of the most popular tourist destinations in South India.It is a semi urbanized city having few industries including plywood, rub wood, mattress factories, and handloom, while agriculture and fisheries are the key sectors in villages of this district.The observational site is close to the National Highway (NH 17) as shown in the Fig.1(b), which has reasonably heavy traffic during the day and minimum at night.The observational site is surrounded by commercial buildings and residential area, and is about 2 km away from the Arabian Sea.Since the industries are quite far, vehicular emission is the major source of air pollution at this site.Ground based observations were performed continuously from 23 to 29 December 2019 at Kannur town (11.87°N, 75.37°E, 3 m msl) to measure the concentrations of trace gases and particulate matters in the wake of the solar eclipse on 26 December.Kannur district was identified the location in the Indian sub-continent where this annular solar eclipse was first observed.

Fig. 2 .
Fig. 2. Variations of (a) solar radiation and (b) surface O 3 observed during eclipse day and control days.

Fig. 3 .
Fig. 3. Variation of total column O 3 for control days and the ecli pse day.

3. 2
Variations of CO, NO, NO 2 , SO 2 , NH 3 and BTEX Diurnal variations of CO and NO 2 usually exhibited two maxima during morning and evening due to peak traffic hours at Kannur town.Generally, the morning peak occurred at 10:00 IST and evening peak at 20:00 IST.The observed low concentrations during the day time hours and high concentrations during the night time hours are mainly due to the local emissions and the boundary layer height variations.Since CO and NO 2 are

Table 1 .
Percentage of changes observed in solar radiation, O 3 , NO 2 , CO, SO 2 , NH 3 , BTEX, temperature and relative humidity during the solar eclipse episode with respect to control days.

Table 2 .
Correlation coefficients between observed trace gases, particulate matter and meteorological parameters during eclipse episode.