Analysis of Ground Level O₃ and No_x Measured at Kannur, India

Air pollution will be an important and challenging problem in the coming decades that will affect sustainable development of nations [1-3]. The rapid growths in industrial activities due to globalization result in vast amounts of potentially harmful gases and particles being emitted into the atmosphere on a global scale [4]. Prominent air pollutants are sulphur compounds (SO₂ and H₂SO₄) nitrogen compounds (NO_x and NH₃), carbon monoxide (CO) and organic compounds (hydrocarbons, VOCs, halogenated compounds). These trace gas species in the atmosphere produce other secondary pollutants via thermal, chemical and photochemical pathways [5,6]. Surface O₃ is identified as one of the air pollutants that effect human health and the ecosystem [7-9]. Surface level O₃ concentration has steadily increased since the 19th century [10,11]. The sources of surface O₃ are in-situ photochemical production, transport from other in-situ production sites and downward transport from the stratosphere [12-15].

The photochemical production of O₃ is enhanced by the chemistry involving CO, CH₄, Non-Methane Hydro Carbons (NMHC) and other Volatile Organic Compounds (VOC) and occurs in the presence of sufficient amount of NO_x and depends on the concentrations of hydroxyl and other radicals which act as catalysts [16-21]. NO_x and VOCs are emitted from anthropogenic sources such as fossil fuel combustion, power plants, industrial activities and transportation, as well as from natural sources such as lightning, and vegetation (biogenic VOCs such as isoprene). Recently, elevated concentrations of O₃ have been observed even over large rural areas in developed countries due to anthropogenic activities and biogenic emissions where industrial activities are limited [22-26]. Two mechanisms have been proposed to account for the high rural O₃ [27-29]. One is the transport of O₃ from urban areas, and the other is the transport of its precursors like NO_x and non-methane hydrocarbons [30], followed by photochemical O₃ production. Ozone is a strong oxidizer in the atmosphere and its chemistry is complex. However, the relationship between O₃ and its precursors represents one of the major scientific challenges associated with urban air pollution [31,32]. Further, O₃ being a strong greenhouse gas, increasing levels of O₃ can warm the troposphere, and induce significant changes in the climate over an extended period of time [33,34]. Thus, a long-term investigation of surface O₃ and its precursors is essential to understand the ambient air quality over a location. Meteorological parameters including wind speed, wind direction, temperature, solar radiation and relative humidity have substantial influences on O₃ concentrations [35,36]. Generally, O₃ is observed to be higher in both winter and summer at many measurement sites over the globe [37].

Surface O₃ in the atmosphere has been monitored sporadically at various sites around the globe for the past several decades [38,39] Many studies have been carried out on the photochemical characteristics of O₃ and its precursor gases in the United States [40,41] and most networks in Europe were implemented at the beginning of 1980s. Atmospheric concentrations of O₃ and its precursors like NO_x and VOCs are increasing rapidly in Southeast Asia as a result of the fast growth of the transportation and industrialization [42] found that the transport sector contributed 37% of the NO_x emissions, while 27% came from power generation and 18% from industry. Measured and modeled O₃, NO_x and VOCs in Shanghai, China by [43] shows that O₃ concentrations are higher in rural area than in center of the city, suggesting that there are O₃ depression processes in center of city and air pollutant emissions are not favorable for the O₃ chemical formation. The studies on the significant role of O₃ influenced on the ambient air quality were reported by [44,45]. Further [46] found that Surface O₃ at a coastal site in Hong Kong is strongly influenced by the outflow of Asian continental air during the winter and the inflow of maritime air from the subtropics in the summer. Modeling studies have been emerged as one of the promising methods to evaluate the production and transport of O₃ confined in a region. Such studies were carried out by many groups [47-50] which were quite promising.

In the Indian continent, available observations [51-54] and modeling studies [32,55] show higher O₃ levels during spring winter and summer extending until May. Such variations over India are due to higher levels of precursors and the availability of abundant solar radiation flux during these seasons. In India, various groups have been involved in carrying out long-term measurements of surface O₃ and its precursors; their observations are quite relevant in understanding the regional air quality [28,32,45,52,56-66]. In order to explore O₃ chemistry and its transport at Kannur, a tropical coastal site in the state of Kerala in India, a continuous observation has been on-going with the support of Atmospheric Chemistry Transport and Modelling (AT-CTM) project under Geosphere Biosphere Programme (GBP) of the Indian Space Research Organization (ISRO). This study describes the key observations of both seasonal and diurnal variations of surface O₃ and its precursor NO_x and their variations with meteorological parameters. The location is rural with limited industrial activities. The observation of ambient air quality over this location for a period of one year from November 2009 to October 2010 revealed that the seasonal variation of O₃ was influenced by meteorology and air mass movement. This manuscript begins with a brief introduction.

Monitoring site

The location of the sampling site at Kannur University campus (KUC) (11.9^°N, 75.4^°E 5m) is shown in Figure 1. Observation site is in the northern part of Kannur district in Kerala state which is surrounded on one side by a river basin and the other side by the Arabian Sea. This site is in the valley of the mountains of Western Ghats. It is a rural location with no major industrial activities except a few small-scale industries such as plywood and mattress manufacturing units. The distance to the seashore to the west is 4 km and that to the Western Ghats on the east is 50 km. The land area of Kannur is about 3000 km² with an average population density of 1000 per km². KUC is situated in an open land which receives plenty of sunshine (peak average of 1 ± 0.3 KWhr/m²) throughout the day, and the land is surrounded with green vegetation throughout the year.

Figure 1: Location of observation site.

Meteorological background

A prominent meteorological characteristic of this location is the intense monsoon rainfall occurring in two spells every year. The southwest monsoon is active during the months of June-August, and this intense rainfall has been classified as the Monsoon Season. This is followed by the return monsoon or northeast monsoon in November. Hence September, October and November are earmarked as the post monsoon (autumn) season with scattered showers accompanied by heavy thunder and lightning. The intensity of summer is masked by the southwest monsoon season over the entire region. About 80% of the total rainfall occurs from June to August which constitutes the main monsoon season. The months, December, January and February with scant rainfall and relatively low humidity constitute the winter season. From March to May the season is summer with scorching heat and high convective flow. Thus, we have Easterly wind during winter months from December to February, the Westerly and South-Westerly winds during summer, and the monsoon and post-monsoon season. The period from January to March records the maximum sunshine hours of more than 9.1 hours/day and between June and August records the minimum sunshine due to cloudy skies. Figure 2 shows the monthly variations of meteorological (wind speed, temperature, relative humidity and rainfall) in Kannur during the period of observation. The temperature was high between March and May and was low during June through August. The average monthly high temperature ranged from 29.6 to 37.1^°C and low temperature ranged from 22.9 to 25.8^°C.

Figure 2: (a) Monthly mean variations of wind speed, (b) temperature, (c) relative humidity, (d) total rainfall observed during the period of observation.

The maximum humidity was measured during monsoon (June, July, August) and minimum was observed during winter months (December, January, February). The maximum monthly average relative humidity ranged from 55.5% to 88.4% and minimum of 45% to 80% at this location. The maximum rainfall was recorded during monsoon, while minimum was observed during the winter season. Figure 3 shows the monthly mean wind speed and wind direction at 1000 hPa during the study period at the observation site. The wind speed is steady during monsoon season and minimum during winter season. It was pretty high during the period from June to September and low from December to April. The maximum average wind speed ranged from 2.4 to 5.9 km per hour and the minimum from 1.3 to 4 km/hr during the period of observation.

Figure 3: Polar diagram showing wind speed and wind direction at KUC.

The meteorological parameters like temperature, relative humidity, wind speed and wind direction were collected from the local automatic weather station, which is one of the stations managed by the Meteorological and Oceanographic Satellite Data Archival Centre (MOSDAC) and established by the Indian Space Research Organization (ISRO).

Measurement techniques

The concentrations of surface O₃ and NO_x were measured simultaneously with the aid of respective gas analyzers from Environment S.A., France. The ambient air collecting sampling inlet hood was placed at a height of 6 m from the ground and data were continuously recorded from November 2009. The concentrations of O₃ were obtained using the ozone analyzer (Model O₃ 42M) and the details of the analyzer and its principle were described elsewhere [67-69] Calibration of the analyzer was performed at regular intervals using ozone free air and pure ozone produced by the appropriate gas generators in the analyzer. It has a lower detectable limit of 0.4 ppbv with a minimum response time of 20 seconds. This analyzer is provided with an in-built correction for temperature and pressure variations as well as intensity fluctuations of the light source. Similarly total NO_x was monitored with the aid of a standard chemiluminescent analyzer (Model AC32M). The method was found to have higher sensitivity and specificity [70-72]. The lower detectable limit was 0.4 ppbv with a response time of 30s. The calibration of the analyzer was carried out using reference gas cylinders.

Diurnal variation of ozone

Eight hour averaged (09:00-17:00 h) surface O₃ concentration measured from November 2009 to October 2010 on a day-to-day basis is shown in Figure 4. Despite considerable short-term variations, a seasonal change was seen for the entire period. Among the maximum mixing ratio of O₃, eight-hour average highest and lowest values were observed in December (46.9 ppbv) and in July (16.4ppbv). From the Figure 4, it is clear that O₃ levels were observed to be the highest in winter and lowest during the monsoon season.

Figure 4: Day time average (09:00-17:00 h) O₃ concentrations for one year from November 2009 to October 2010.

The enhanced O₃ mixing ratio observed during winter season was mainly due to the easterly air flow that favours advection of precursors from inland locations which induces active photochemistry. Figure 5 shows the diurnal variation of O₃ for different months over a period of one year from November 2009 to October 2010. The vertical bars shown in the Figure 5 are 1σ standard deviation. The highest maximum O₃ concentration was observed in December and minimum in July.

Figure 5: Diurnal variation of O3 for different month at KUC.

The daytime increase in O₃ mixing ratio is mainly from the photo-oxidation of industrial and anthropogenic hydrocarbons, carbon monoxide, and methane in the presence of sufficient amount of NO_x [58,73,74] With the onset of sunshine, O₃ concentration gradually increases and it reaches a maximum value at 14:00 h noontime, due to its photochemical formation through the photolysis of NO₂ via the following set of reactions.

NO₂ + hν (λ< 420 nm) →NO + O (R1)

O + O₂ + M → O₃ + M (R2)

It was observed that the mixing ratio of O₃ started to decline after 16:30 h in the evening on all days. The low concentration of O₃ observed during nighttime is due to the absence of photolysis of NO₂ and loss of O₃ by NO through the following titration reaction and surface deposition.

O₃+ NO → NO₂ + O₂ (R3)

Loss of O₃ also occurs due to both dry and wet deposition that results in a minimum concentration at sun rise. [28] Have observed a similar diurnal variation at Gadanki, a rural site in south-east India (13.5^°N, 79.20E, 375m asl). To substantiate the annual surface O₃ variation, seasonal average daytime, nighttime O₃ mixing ratios and the rate of change during morning and evening are presented in (Table 1).

Table 1: Monthly average daytime and night time O₃, maximum and minimum O₃ and the rate of change of O₃ at KUC.

From the Table 1, it is evident that the observed higher O₃ mixing ratios were in December (46.9 ppbv) while lower mixing ratios were in July (16.4 ppbv). This may be due to the variations in precursor gases and the influence of changing meteorological parameters. The daytime high average O₃ mixing ratio observed in December (31.9 ± 4.7) ppbv was due to the lower boundary layer that occurs in winter and the low relative humidity. The daytime low average O₃ mixing ratio was in July (11.0 ± 3.4) ppbv, which is mainly due to the intense southwest monsoon which removes pollution over this location. The highest rate of O₃ increase (9.2ppbv/h) was observed in December between 08:00-11:00 h Indian Standard Time (IST) and lowest (2.4ppbv/h) was found in August. The O₃ removal rate was the highest in December (-14.0ppbv/h) between 17:00-19:00 h IST and the lowest was in June (-2.4ppbv/h).

Seasonal variation of O₃

Figure 6 shows the seasonal average diurnal variation of O₃ during 2009-2010 at KUC. The variations in O₃ concentrations for different seasons namely winter (DJF), summer (MAM), monsoon (JJA) and post-monsoon (SON) are shown. It is observed that the variation of O₃ during all seasons shows a similar pattern. Among the maximum concentration, the highest value of O₃ mixing ratio was found in winter (44 ± 3.1) ppbv and the lowest during the monsoon season (18.46 ± 3.5) ppbv. The lowest magnitudes of O₃ mixing ratios observed were (4.0 ± 1.4) ppbv in summer and (2.24 ± 1.26) ppbv in monsoon seasons. The highest O₃ observed in winter is attributed to a lower mixing height that results in trapping of pollutants near the earth’s surface due to temperature inversion [25]. Similar patterns of seasonal variations have also been observed at Gadanki [28] Anantapur [75] and Trivandrum [66] in south India. The minimum and maximum mixing ratios of O₃ found at this site during monsoon were (2.3 ppbv, 18.5 ppbv) and post monsoon were (3.3 ppbv, 28.0 ppbv) respectively. From this it is evident that O₃ exhibits a distinct seasonal variation over this site, which may not only be controlled by solar radiation, but also by atmospheric dynamics as well. There is also a possibility of enhanced transport of O₃ from the stratosphere during the winter season [64,74]. Further the active photochemistry of NO_x under clear sky conditions in winter season would enhance the O₃ abundance.

Figure 6: Seasonal average diurnal variation of O₃ at KUC.

This in turn offers a suitable environment for the production of added O₃ during winter season at this site. Moreover, the meteorological parameters have a strong influence on the seasonal changes in the concentration of O₃. An enhancement in the boundary layer height and increased cloud cover over this location during summer induce a relatively small mixing ratio observed in summer than in winter. A relatively low amount of surface O₃ was observed during monsoon seasons than summer and winter seasons. During the monsoon period, the absence of sunshine and high relative humidity reduce the O₃ concentration at this location. Seasonal average variations of O₃ during day time and night time are shown in (Table 2).

Table 2: Seasonal average day and night time O₃, maximum and minimum O₃.

From the Table 2, it is clear that daytime average O₃ concentration became a maximum (29.2 ±4.4) ppbv during winter while nighttime O₃ became higher during summer (5.3 ±1.8) ppbv due to the land air mass influence in winter and oceanic influence in summer as observed from the backward trajectory depicted in of Air trajectories during the period of observation. The minimum and maximum observed mixing ratios of O₃ during winter were 2.9 ppbv, 44.0 ppbv and that during summer was 33.4ppbv, 4.1 ppbv respectively. From this it is evident that O₃ variation is not only controlled by solar radiation, but also by atmospheric dynamics. The favorable conditions for photochemical production of O₃ are satisfied at KUC during winter months. Moreover, the meteorological parameters have a strong influence on the seasonal changes in the concentration of O₃. A relatively small mixing ratio observed in summer than in winter is due to high air temperature induced convective activities and increased cloud cover. The minimum and maximum observed mixing ratios of O₃ during monsoon were 2.3 ppbv, 18.5 ppbv and that for post monsoon were 3.3 ppbv, 28.0 ppbv respectively.

Figure 7: Diurnal variations of annual hourly average NO, NO₂ and NO_x.

Figure 8: Diurnal variations of NOx for different months at KUC.

Figure 9: Scatter plots showing correlations between (a) NO and O₃ (b) NO₂ and O₃ in winter.

Figure 10: Correlations observed between (a) NO₂ and NO_x (b) NO and NO_x.

Figure 11: Scatter plot showing the daytime correlation between NO_x and O₃.

Figure 12: Diurnal variation of (a) OX, (b) day time variation of [NO₂]/ [O_X] with NO_x.

Figure 13: Variation of O3 with air temperature and relative humidity for the study period.

Figure 14: Backward air trajectories at KUC using CGER-METEX reanalysis data.

Diurnal and seasonal variations of NO, NO₂ and NO_x

Figure 7 shows the annual average diurnal variations of NO, NO₂ and NO_x observed at KUC. The morning peak of NO₂ appears 2h after the NO peak. Subsequently, after the morning peak, NO diminishes until it reaches its lowest value at 15:00 h and has a gradual increase till midnight. During nighttime, surface emission of NO were limited inside the nocturnal planetary boundary layer, and NO reached its second highest peak at midnight. Likewise, a decline in NO₂ mixing ratio observed during daytime was due to the enhanced photolysis by which O₃ is formed at this site. It was observed that NO_x mixing ratio started declining from (3.1 ± 0.28) ppbv at 09:00 h IST and it became a minimum of (1.53 ± 0.26) ppbv at 16:00h and began to regain in magnitude during nighttime. This reduction in NO_x concentration boosts the production of O₃ through photolysis induced by sunlight in daytime. At night, the relatively low air temperature near the surface could prevent the vertical dispersion of NO_x, contributing to its accumulation and resulting in high nighttime mixing ratio. Higher levels of NO_x during early morning and night hours are due to the combination of anthropogenic emissions, boundary layer processes, chemistry as well as local surface wind pattern [28,75]. During nighttime, the boundary layer descends and remains low till early morning, hence pollutants get trapped in the shallow surface layer and show higher concentration. In order to attain high levels of O₃ during noon, large amounts of precursors are needed, and this may leads to low levels of NO_x during daytime [57].

At this site, the annual average mixing ratio of NO_x is found to be about 2.5ppbv, which is relatively small compared to that of rural sites, where mixing ratios can be much higher [28]. The diurnal variation of NO_x for different months is shown in Figure 8. The maximum and minimum concentrations of NO_x observed at KUC were (3.9 ± 0.5) ppbv and (1.0 ± 0.3) ppbv in December and January respectively. The maximum rate of change of NO_x occurs in December in which O₃ concentration is high and the minimum rate is in July in which O₃ concentration is lowest.

The maximum magnitudes of average NO_x concentration (2.9 ± 0.8) ppbv were observed in summer and minimum during monsoon (2.3 ± 0.4) ppbv. During winter the maximum average concentration was (2.5 ± 0.6) ppbv and post-monsoon it was (2.7 ± 0.7) ppbv. Thus the concentration of NO_x did not show significant variation during the period of observation at this site but only had a small seasonal variation. NO_x mixing ratios are generally stronger during early morning and nighttime in all the seasons. The minimum and maximum observed mixing ratios of NO_x were 1.3 ppbv, 3.5 ppbv (winter), 1.6 ppbv, 3.3 ppbv (summer), 1.6 ppbv, 3.2 ppbv (monsoon) and 1.5 ppbv, 3.4 ppbv (post monsoon).

Analysis of O₃, NO, NO₂ and NO_x

The basic process leading to the photochemical formation of O₃ in the lower troposphere and the polluted urban atmosphere involves the photolysis of NO₂ to yield NO and a ground-state oxygen atom. O₃ is formed and destroyed in a series of reactions involving NO and NO₂ as described by (R1), (R2) and (R3) in Diurnal variation of ozone where M denotes a third body like a nitrogen or oxygen molecule that absorbs the excess vibrational energy and eventually stabilizes the O₃ molecule thus formed. In the absence of other chemical species, these three reactions govern O₃ concentration. Figure 9 (a) represents the variation of O₃ concentration with NO mixing ratio within a sample interval of 15 minutes. It is evident that O₃ concentration diminishes as NO mixing ratio increased, which may be due to the removal of O₃ by titration with NO. The most appropriate fit to this distribution is exponential and the empirical relation is (Figure 9)

[O₃] = 97.07 exp (-2.36 [NO]) (1)

Thus, a fast decay in O₃ concentration is observed at high NO mixing ratios from late evening to early morning. Further, the daytime variation between O₃ and NO₂ observed on the maximum number of clear sky days available during winter mornings is shown in Figure 9(b). A linear fit with excellent correlation with a significance level (p < 0.01) that occurs in the winter season reveals the active photochemistry of O₃ production, which starts early morning. The empirical relationship observed between [O₃] and [NO₂] is

[O₃] = 3.6[NO₂] + 3.72 (2)

Figure 10 (a) represents the variation of NO₂ and NO_x using the mean observed data. A linear fit with a good correlation is obtained and the observed empirical relationship is

[NO_x] = 0.82[NO₂] + 1.17 (3)

and it shows the active competition between NO₂ and NO_x on the chemistry of O₃ at this location.

The correlation between NO and NO_x shown in Figure 10 (b) which yields a linear relationship with an empirical relation

[NO_x] = 1.29[NO] + 1.21 (4)

and this enhanced influence of NO mixing ratio is the primary reason for the nighttime reduction in O₃ concentration at this site. However, the wet and dry depositions of O₃ to land and marine surfaces are significant due to the proximity of Arabian Sea and with fairly good vegetation in the area. Figure 11 shows a negative linear correlation between daytime O₃ and NO_x observed during the whole period of observation suggesting the possible VOC-sensitive characteristics of the study location as described by [76,77] and the linear relation is as

[O₃]= 52.41-12.37 [NO_x] (5)

The coefficient of this correlation is -0.74, which is quite significant.

The organic emissions from nearby wetlands and biogenic VOCs from vegetation make this location VOC sensitive assuming that the higher O₃ levels appear when the VOC concentrations increase. Fishing boats and ship emissions are other strong contributors of organic species over the Arabian Sea [78]. The significant contribution to NO₂, SO₂, Black Carbon (BC) and Organic Compounds (OC) is large over coastal zones. Secondary species formed from fishing boat emissions have large chemical lifetimes and are transported in the atmosphere over several hundreds of kilometers. Thus, they can contribute to air quality problems on land [79]. In general, the emission perturbation is most effective in increasing ozone formation over regions with low background pollution [19]. The region in the sea which is adjacent to our sampling location is an estuary with a good catchment area. The marine department of the state of Kerala has estimated that approximately 1425 fishing boats are engaged in fishing from early morning 06:00 h to 14:00 h on all days in a span of 4 km from the shore of the Arabian Sea in Kannur region except during the monsoon season. It is estimated that about 8 liters of diesel is consumed per hour by a single vessel and on an average 16 kilolitres of diesel is burnt per day. The sea breeze at noon time transports these organic compounds from the sea to the land. PM10 samples have been collected using a Respirable Dust High Volume Air Sampler (ENVIROTECH, India Model APM 460 NL) at this site and the chemical analysis employing gas chromatography revealed the presence of several organic species during the day time and is listed (Table 3).

Table 3: Organics detected as associated with the particulate matter in air at this site.

Thus the presence of various organic species detected over this location substantiates the existence of a large amount of VOCs in this region arising from both anthropogenic and biological origin. Since quantifications of VOC were not available on site, it was not possible to investigate the complex chemistry between VOC-O₃- NO_x in detail. However, the analysis of NO_x and O₃ was conducted with the understanding that O₃ production was influenced by NO_x concentration. Nevertheless, a detailed analysis of VOC-O₃-NO_x chemistry is to be explored through future investigations.

Diurnal variation of OX and the correlation between [NO₂]/ [OX] with [NO_x]

Surface O₃ and NO₂ are inextricably linked due to their strong chemical combination. Therefore, the response to reductions in the nitrogen oxides emissions is remarkably not linear [80] and any resultant reduction in the level of nitrogen dioxide is invariably accompanied by an increase in the concentration of ozone. In addition, its interplay with VOCs represents a complex mechanism whose detailed characterization requires a significant number of measurements of source emissions and meteorological variables. At night, O₃ reacts with NO to form NO₂, which further reacts with O₃ to yield NO₃ and N2O5 and in daytime in the presence of sunlight NO₃ and N2O5 reconvert to NO₂ and O₃, these compounds may be grouped as a single chemical family, odd oxygen or OX [81-83]

Thus, OX (nocturnal) = O₃+NO₂ + 2NO₃+ HNO4+3N2O5 (R4)

For daytime, OX is only the sum of NO₂ and O₃ since the daytime abundances of NO₃, HNO₄ and N2O5 are quite meager. The interconversion of O₃, NO and NO₂ under atmospheric conditions can be summarized by the primary reactions (R1), (R2) and (R3) mentioned above. These equations constitute a cycle. The overall effect of reactions (R1) and (R2) is the reverse of Reaction (R3). These reactions represent a closed system in which the NO_x and the OX (O₃+NO₂) components are related. In addition, the increasing ozone background concentration influences on the local levels of O₃ and NO₂ and on the efficiency of local emission controls. As a result, different authors [6,84-90] studied the relationships between ambient levels of NO and NO₂, O₃ in order to improve the understanding of their chemical coupling. Figure 12 (a) shows the diurnal variation of OX with one sigma standard deviation. Similar to the variation of O₃, OX concentration shows a mid day maximum and minimum during night. The OX concentration gradually increases after sunrise and reaches a maximum at 1500 hrs in the afternoon and decreases slowly and reaches a minimum during night time. The photochemical activities are quite intense during midday hours, the O₃ production is higher while the concentrations of NO₂ and NO starts declining by which OX shows a maximum during noon time. A key factor to be considered on the photochemical activity and the behaviour of atmospheric pollutants including OX is the mixing processes caused by the heating of the ground layer and the variation in height of the planetary boundary layer. Vertical convection is a principal mechanism for transporting air from close to the surface to the free troposphere [76,91].

The combination of fresh traffic emissions that occurring from the nearby highway during the early morning, mixed with aged pollutants due to convective activity and the variation of the height in the boundary layer throughout the day, and the atmospheric cycle between O₃, NO and NO₂ with intense photochemical production of O₃ at these hours leads to a daily maximum of OX around noon time. Further, the level of OX, an oxidizing level indicator of the NO_x, O₃ regime, could be increased in two ways: (1) the net increase of NO₂ through NO_x emissions and then oxidization of NO to NO₂ by radical species such as RO2 (where R is a hydrogen atom or carbon containing fragment), OH etc. [5] and (2) the net increase of O₃ via regional background input. The concentrations of NO, NO₂ and O₃ in the photo stationary state are related by the following expression

                                                    (6)

Where J1 is the rate of NO₂ photolysis and k₃ is the rate coefficient for the reaction of NO with O₃. Coefficient J₁ is a function of solar radiation intensity and k₃ is function of the temperature (T) [76] propose the following expression for k₃:

k₃ (ppm-1min-1) = 3.23x103 exp [-1430/T] (7)

The variation of the mean values of J₁/k₃ estimated applying Equation (6) to the observational data of NO, NO₂ and O₃ concentrations. On the basis of the photo stationary state relationship it is possible to infer an expected variation of day light average [NO₂]/[OX] values with [NO_x]. A scatter plot between [NO₂]/ [OX] and [NO_x] is depicted in Figure 12 (b) to reveal the significant variations of O₃ and its precursors NO and NO₂ at this location. A gradual decrease observed in NO₂ and NO produce a rapid change in O₃ during morning and evening could result an exponential relationship. Thus the most appropriate fit to this variation is through an empirical relationship is

[NO₂]/ [OX] = 0.018 exp (0.783) [NO_x] (8)

In this scatter plot, points with higher values represent the mixing ratios of NO₂ and O₃ during early morning and late evening of a day. The low values of O₃ and higher values of NO₂ are the main reasons for this. While in the noon time, O₃ is found to be increasing and NO_x starts declining. Hence, the points with lower values are found to be in the noon time.

Influence of humidity and temperature on ozone

Comparison between O₃ and meteorological parameters like temperature and humidity during winter, summer, monsoon and post monsoon season has been made at KUC. Figure 13 elucidates the relationship between O₃, relative humidity and temperature for the period of observations. The correlation coefficient between O₃ concentration change with relative humidity, wind speed and temperature were -0.84, -0.85 and 0.87 respectively during the period of observation. In winter, summer, monsoon and post-monsoon, the correlation coefficients between O₃ and humidity were 0.82, -0.85, -0.83 and -0.86 respectively Likewise, the correlation between O₃ and temperature in these seasons were 0.88, 0.86, 0.84 and 0.90 (p< 0.01) respectively [92].

It is evident that O₃ variation is directly correlated to temperature and is inversely related to humidity. Monsoon season is associated with the maximum number of cloudy days and minimum sunshine hours; and it also provides minimum temperature and maximum humidity. Thus, in the presence of low temperature and high humidity, O₃ production is kept minimum. In the post monsoon season, the morning sky is clear followed by few showers in the afternoon. The existence of clear sky after the monsoon offers an environment to build O₃ and aerosols at this location. The positive correlation between O₃ and temperature is due to the fact that the radiation controls the temperature and hence the photolysis efficiency will be higher. When the humidity becomes higher, the major photochemical paths for removal of O₃ will be enhanced. Moreover, higher humidity levels are associated with large cloud cover and atmospheric instability, the photochemical process is slow down and the surface O₃ is depleted by deposition on water droplets in this location, hence the O₃ concentration has a strong dependence on humidity.

Air trajectories during the period of observation

To validate further the seasonal variation of surface O₃, monthly average backward air trajectories were retrieved from CGER-METEX reanalysis data (http://db.cger.nies.go.jp/metex/trajectory.html) along the surface of this location during one year and shown in Figure 14. The Meteorological Data Explorer (METEX) model is used for backward trajectory calculations [93].

The model was developed at the National Institute for Environmental Studies (NIES), Japan and Global Environmental Forum, Japan. From this, it is clear that the air mass movement during winter and post monsoon months was from east of this site and confined from west during summer and monsoon seasons. During winter season, long-range transport of air mass contributes to the observed high O₃ levels in addition to its photochemical formation. It is further observed that air masses appear to originate from the eastern part of Kannur that is influenced by the transport of pollutants from nearby industrialized land during winter. While, during post monsoon, short-range air masses at low altitude (500m) appear to originate from the east of this site inducing the transport of precursors, which seems to have little contribution. Hence during post monsoon, in addition to photochemical production, inventories within a short range may effectively contribute O₃ formation at this location. During summer and monsoon seasons, the movement of air mass trajectories originated over the Arabian Sea and traversed through a smaller area of landmass before reaching Kannur. Since, these air masses have a strong marine influence, the observed enhancement of surface O₃ is only at the expense of photochemistry rather than the transport of pollutants. During monsoon season, winds were strong in the westerly direction, back trajectories were of oceanic origin, and marine air mass prevailed over the Indian subcontinent. Marine air mass was relatively clean compared to continental air. Owing to this oceanic influence, the air mass enriched with hydroxyl radicals (OH) may trigger the removal of O₃ to a larger extent. This may be one of the reasons for the reduction of O₃ mixing ratio observed during summer season at this location in spite of high solar flux. A similar pattern of back trajectory was observed in Trivandrum during 2007- 2009 [56]

Comparison with other measurements in india

Seasonal variation of O₃ observed at KUC show a typical pattern of rural site, which is highly influenced by the seasonal changes, like some other locations in Indian sub continent. Figure 15 shows a comparison of monthly average surface O₃ at KUC and other observational sites in India.

Figure 15: Comparison of monthly average ozone at Kannur with other observational sites in India.

From the Figure 15, it is evident that Nainital, a high altitude site located in the central Himalaya region and in Delhi, the capital city of India recorded the highest O₃ mixing ratio during late spring. At Anantapur, Pune, Joharapur and Tranquebar O₃ concentrations show distinct maxima in summer season whereas at Dayalbag, Agra and Gadanki it became high during summer/ winter season. At Ahmadabad and Mt. Abu, O₃ concentration reaches its peak during autumn/winter and at Trivandrum [66], a tropical coastal site shows a maximum O₃ concentration during winter/post monsoon seasons. Among the minimum O₃ mixing ratio recorded at these locations, Agra, Gadanki, Dayalbag and Tranquebar register lower concentrations. It is further observed that the O₃ minima were found during monsoon and post monsoon seasons at all locations except Gadanki, while the maxima changes with season. Variations of surface O₃ concentration recorded at different locations in the Indian subcontinent were mainly due to latitude/longitude variation, population, meteorological parameters, availability of intense solar radiation, differences in the concentrations of precursor gases and anthropogenic activities.

This paper describes the diurnal and seasonal variations of surface O₃ and its precursor NO_x observed for one year from November 2009 through October 2010 at Kannur University, a rural site lying along the west coast of India. The higher surface O₃ concentration observed in the midday and lower concentrations during nighttime was in tune with the solar UV flux. A significant seasonal variation for O₃ and NO_x mixing ratios at this site was observed. The average O₃ mixing ratios were maximum during winter and minimum during the monsoon period. O₃ mixing ratio was higher in the winter months at KUC. This may be due to the dominant transport of biogenic VOCs in winter, from the Western Ghats lying east of this site since the air mass movement is from that direction. While in summer, in spite of higher solar flux, O₃ concentration was low at this site. This may be due to the cloud cover and the higher abundance of humidity due to proximity with the Arabian Sea. During the southwest monsoon period, O₃ concentration was found to be a minimum and it gradually builds up in the post-monsoon season. The seasonal changes in O₃ levels are found to be different due to the atmospheric pollution levels. We found good correlations between O₃, NO and NO₂ which revealed the chemistry of O₃ production from NO₂ and titration with NO at this location. Further a linear relationship between NO and NO_x and NO₂ and NO_x obtained could be useful in unfolding the chemistry of O₃ at this location. The net effect of NO_x on O₃ concentrations was negative with a decaying exponential correlation indicating a possible VOC sensitive location, which reveals the prominent role of abundant biogenic VOCs on the production of O₃ even at lower levels of NO₂ present over KUC.

A good correlation between O₃, temperature and humidity were obtained which shows that the seasonal variation of O₃ has strong dependence on meteorological parameters. This observation further reveals that the marine influence is well defined in this location and other prominent precursors is essential in the O₃ chemistry over this location. These observations are useful in understanding the ambient air quality of a location with high population density and the enhanced anthropogenic activities that lead to deterioration in the air quality. The elevated levels of O₃ in an industrially less developed area can also affect human health. This will be further ascertained through an analysis of indoor air quality in the area. The results of this study are preliminary and need confirmation with more observation to explore the O₃ chemistry and transport with the aid of other gas analyzers like CO, CH₄ and VOC.

The observations are carried out under the consistent support of ISRO-GBP through AT-CTM program. Authors are grateful to Professor Shyam Lal, Physical Research Laboratory, India for his keen interest in this work. They are grateful to Professor R Ravikrishna, Department of Chemical Engineering, Indian Institute of Technology, Madras, India for analyzing and identifying the organic compounds. Authors thank Dr. Jiye Zeng (NIES, Tsukuba) for providing the METEX model data for calculating air trajectories. One of the authors, (Nishanth T) expresses his gratitude to Indian Space Research Organization for awarding Senior Research Fellowship under ISRO-GBP (ATCTM) programme. Authors wish to thank Editor, Managing Editor of the Journal and the anonymous reviewers for their helpful comments on our manuscript which improve the original scientific contents.