Five years surface ozone behaviour in a semi-rural location at Mohal-Kullu in the north-western Himalaya, India


 Surface ozone is one of the most important Green House Gases (GHGs). Five years (2011–2015) measurements of surface ozone (O3) and one of its precursors- oxides of nitrogen (NOx) were carried out at a semi-rural location, Mohal (77.12°E, 31.91°N, 1154 m amsl) in the north-western Indian Himalaya. The concentration of O3, NO, NO2 and NOx was measured maximum 74.6 ± 23.2 ppb in 2013, 27.5 ± 7.5 ppb in 2013, 51.8 ± 13.2 ppb in 2013, and 60.8 ± 13.2 ppb in 2012, respectively. Seasonally, O3 concentration was highest during summer while lowest in monsoon. The O3 concentration shows unimodal peak while its precursors show bimodal peaks. A reasonable decrease in percent change was found in terms of O3 (-13), NO2 (-6), and NOx (-3) due to imposing regulations imposed by local government in compliance with the order of Hon’ble National Green Tribunal (NGT) of the country in 2015. However, NO (9) is increasing due to vehicular activities in the nearby area which is one of the famous tourist spots. Washout effect due to increasing rainfall by 8% might have also affected O3 concentrations. Biomass burning for heating and cooking purposes and forest fire for tender livestock forage and transported gaseous pollutants from the Middle East countries and Indo-Gangetic Plain (IGP) could be major contributors of surface ozone and its precursors. HYSPLIT air-mass back trajectories drawn for external ozone sources showed the maximum air masses reached the study location either with the western desert countries or IGP polluted regions.


Introduction
Surface ozone (O 3 ) and changing climate are closely linked to each other. Surface ozone in climate change perspective primarily indicates changes in temperature. Air parcels having more ozone may absorb more heat through absorption of sun's ultraviolet radiation and upwelling infrared radiation from the troposphere. The more ozone in the air masses, the more it retains heat. According to the Intergovernmental Panel on Climate Change (IPCC 2001), O 3 in the atmosphere is the third most powerful greenhouse gas after CO 2 and CH 4 .
In many parts of the northern hemisphere, the concentration of O 3 near the surface has been found to be increasing (Nair et al. 2002;Naja and Lal 2002). The meteorological parameters such as sunlight, temperature, and wind speed were the main parameters on which surface ozone concentration depends.
From the polluted regions, ozone rich air-masses could be transported ) and may affect nearby other remote places (Chand et al. 2001;Naja et al. 2004;Kumar et al. 2013;Lal et al. 2013). Due to increasing anthropogenic activities, the surface ozone produced by photochemical reactions is increasing rapidly (Lal et al. 2017). At high altitude, its maximum concentration is normally found in May, while minimum was in August due to conversion of its precursors (Sanchez et al. 2005;Kumar et al. 2010) due to washout effect or other atmospheric dynamism. In the southern and central Asia of the developing countries, there is continuous increase in the emission of O 3 precursors due to higher emissions and weaker legislative policies (Ghosh et al. 2018). Lower concentration of surface O 3 was measured in Paris due to poor layer formation of pollutants in morning, which develops due to daytime boundary (Klein et al. 2019). Also, it shows signi cant seasonal and regional variations which mainly depends upon the stratospheric intrusion (Tiwari and Agarwal 2018). In the North Atlantic Ocean, the long-term trend gives the in uence of European regional NO x and VOC emissions which decrease in winter time and increase in summer time while episodic peaks reduce continuously (Derwent et al. 2018).
While in India especially at low altitudinal rural site, maximum O 3 concentration was found in summer and winter but minimum in rainy season (Debaje and Kakade 2006;Reddy et al. 2011). The temporal variation of NO x concentration is highest in winter  and shows seasonal variability with high NO and CO during summer but low in summer rains or monsoon (Tiwari et al. 2015). Also, it shows highest diurnal variation during afternoon hours and lowest in morning hours (Gaur et al. 2014). In Delhi, high level of variation occurs in O 3 concentration due to pollutants and intense solar ux across Indo-Gangetic Plain (Singh et al. 1997;Jain et al. 2005). Statistically, in Punjab and southern Sindh, the temporal increase raised the ozone columns which are consistent to high population, urbanization, and extensive anthropogenic activities (Noreen et al. 2018). Among all the O 3 precursors such as NO x had shorter lifetime and also considered to be rate-limiting precursor in the atmosphere (Liu et al. 1980;1987).
The result of daytime gas phase (OH with NO 2 ) recombination reaction leads to the emissions of NO x with signi cant increase in tropospheric O 3 which mainly promotes the HNO 3 formation (Finlayson-Pitts and Pitts 2000). In the lower troposphere, O 3 concentrations increased signi cantly (Naja and Lal 1996) and long-term measurement of O 3 and its precursors are actively involved in various groups (Lal et al. 2000;Naja and Lal 2002;Naja et al. 2003;Jain et al. 2005;Reddy et al. 2008;Kumar et al. 2010;David and Nair 2011;Singla et al. 2011;Nishanth et al. 2012). With increasing emissions of anthropogenic pollutants, ozone levels are likely to increase in this region in future (Wild et al. 2012;Stevenson et al. 2013). Long term ozone behaviour is studied in the Central Himalaya at Nainital  and northwestern Himalaya at Kullu, Manali and Kothi (Kuniyal et al. 2007;Sharma et al. 2013). At Anantapur (Andhra Pradesh), it shows a signi cant positive correlation with ambient air temperature and negative correlation with both wind speed and relative humidity (Reddy et al. 2012). In the western and southern Indian region, O 3 concentration during summer was in uenced by transported pollutants from backward air-mass trajectories, meteorological parameters and high insolation, while it is highly in uenced during late autumn and winter (Lal et al. 2000;Naja and Lal 2002;Nair et al. 2002;Beig et al. 2007;Mittal et al. 2007;Sharma et al. 2013). Overall, high nitrate levels were associated with anthropogenic causes, whereas meteorological conditions were associated with sulphate concentrations as against to gaseous emissions (Si et al. 2019). By using the stable orbits, satellites will be able to check quality for climate issues, reference locations and meteorological radiometers (Keckhut et al. 2018).

Site Description And General Meteorology
The simultaneous measurements of O 3 , NO x (NO + NO 2 ) and other meteorological parameters were carried out at Mohal-Kullu (31.9°N, 77.19°E, 1154 m amsl) in India from January 2011 to December 2015.
The sampling site is about 5 km south-south west to Kullu town and its distance from National Highway NH-21 is about 228 m (Fig. 1). It is mainly located in the campus of G.B. Pant National Institute of Himalayan Environment (NIHE), Himachal Regional Centre, Mohal-Kullu (H.P.). The observation site is surrounded by hills of considerable height in a range 3,000-5,000 m in the Kullu valley and is far away from the snowline ). The famous River Beas ow in a middle of the Kullu valley extends up to 80 km long from north to south and, on average, is 2 km wide. It begins from Largi (957 m amsl) in the lower Beas basin and stretches up to Rohtang Pass (3978 m) and Rohtang Crest (4,038 m amsl) in the upper Beas basin ). Due to a bowl shaped valley, cooling occurs as the valley traps cold air in winter due to inversion of temperature. It settles close to a valley oor during night time and further leads to build-up pollutants. Heating by solar radiation destabilizes surface air due to movement against gravitational pull which results in dispersion of trapped pollutants.
On the basis of meteorological conditions, it is suggested that this region's climate has four distinct seasons, namely, winter (December to March) is associated with heavy snowfall and rain during summer (April to June). The air mass ow appreciably in uences the pollutant concentrations from the western and south-eastern directions. The monsoon season (July to September) also receives heavy rain. While the autumn season (October to November) characterised with average temperature and unusual and torrent rainfall (Kuniyal et al. 2007;Sharma et al. 2011).

Material And Methods
Observation of surface O 3 and its precursors, i.e., nitrogen dioxides (NO + NO 2 ) were carried out in a bowl shaped valley (1154 m amsl) in an Environmental Observatory at Mohal. Here, different online analyzers and equipment are functional. Among these are UV photometric Ozone Analyzer (Thermo Fischer Model, 49i), and NO x Analyzer (Thermo Fischer, Model 42i), etc.
Surface Ozone (O 3 ) measurement is based on the absorption of ultraviolet radiation by ozone molecules at 253.7 nm (Singla et al. 2011). The radiation source is a UV-lamp and zero air (clean air without any pollutants) and the sample itself is alternately measured in cells. The minimum detection limit is 0.5 ppb and precision is ± 1 while its ow rate is 0.5 slpm. Calibration of the instrument is done regularly by the built-in ozonator (ozone generator) and zero air generators. The online analysers were operated every day continuously for 24 h. Simultaneously, measurement of nitrogen oxides (NO,NO 2 and NO x ) is based on the chemiluminescence effect which forms peaks at 630 nm radiation. NO 2 is measured by the thermal conversion method from NO 2 into NO. The molybdenum converter is found to have higher sensitivity with 100% conversion e ciency (Winer et al. 1974;Finlayson-Pitts and Pitts 1986). The actual concentration of NO 2 is low as a result NO x may also be lower. The minimum detection limit and precision of NO x analyser is 0.4 ppbv and 1 ppbv, respectively.
The analog signal of gases data is converted into digital values through data acquisition system and stored in the computers by using ENVIDAS software. The raw les are then extracted and separated into individual series. The measurements of O 3 and NO x (NO + NO 2 ) were available from January 2011 to December 2015. Hourly averaged data were used to perform statistical analysis. The data have been used to investigate the diurnal and seasonal patterns of trace gases and their interdependence of these two chemical species. The major pollution sources near the study site are local tra c, biomass burning and use of coal in restaurants and hotels for heating and cooking purposes. In the present study, NO x is the most crucial precursor of O 3 .
The pollutants being contributed from external sources were also examined by drawing seven-days back trajectories using Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT-4). This model is available with NOAA Air Resources Laboratory that has been utilized in both operational and research applications to enhance understanding of atmospheric chemical transport, dispersion, and deposition. The trajectories were calculated at 0700 UTC at 1000 m height. These trajectories show their origin, route or direction of the air masses to the present site. Additionally, the HYSPLIT trajectory model can be run both forward and backward depending on required time, and is therefore useful for source-receptor applications. Relevant to this work, HYSPLIT has been applied in studies of East Asian dust and pollution transport to both Mauna Loa Observatory (MLO) (e.g. Rinsland et al. 1999) and the Western United States (e.g. Weiss-Penzias et al. 2006). Calculations in HYSPLIT are based on a hybrid combination of Eulerian and Lagrangian frameworks (Draxler et al. 1998). Eulerian models focus on a speci c location in space and calculate both advection and diffusion on a xed spatial grid while Lagrangian models are applied to an air parcel moving through space, therefore calculating advection and diffusion separately (Draxler et al. 1998). HYSPLIT utilizes the Lagrangian method for calculating advection and diffusion while chemical concentrations are calculated using an Eulerian xed grid (Draxler et al. 1998). For this study, HYSPLIT was used only for particle trajectory calculations without application of the dispersion or diffusion models. The HYSPLIT model can be run using a variety of available model meteorology datasets that have been processed into the format required by the model. These datasets include the metereological parameters like wind, temperature, etc. from which backward trajectories represent more accurate air masses motion in atmosphere (Nguyen et al. 2016). Concentrated Weighted Trajectory (CWT) also gives direction in concentrated ratio which helps to get the signi cant source of O 3 values for air masses in a concentrated form (Sharma et al. 2017).

Results And Discussion
During the study period (2011)(2012)(2013)(2014)(2015), the average high concentration of surface ozone was observed in the year 2013 measuring maximum as 74.6 ± 23.2 ppb, on the other hand its minimum values stood to be 16.0 ± 5.6 ppb in 2015 (Fig. 2). Diurnal variation of O 3 concentration at Mohal in the Kullu valley from 2011 to 2015 showed uni-modal peak during mid of a day at 15:00 hrs (IST). This is mainly due to a strong process of photo-oxidation in presence of sunshine and its precursors. While O 3 values were found low in early morning and late evening due to absence of sunlight. The concentration of O 3 was found to be increasing gradually after sunrise (07:00-08:00 hrs IST), attaining maximum concentration during afternoon (14:00-16:00 hrs IST). Thereafter, it showed a gradual decrease. Annual maximum concentration of NO was found 27.5 ± 7.5 ppb in 2013, NO 2 51.8 ± 13.2 ppb in 2013 and NO x 60.8 ± 13.9 ppb in 2012. These precursors of ozone are mainly considered to be produced due to anthropogenic activities such as vehicular emission and biomass burning; as a result these are considered to be primary pollutants. The percentage change in O 3 concentration with its precursors (NO,NO 2 , NO x ) shows the highest increase with 244.25% during 2011-13. Analysis of meteorological data in context to O 3 showed its high concentration days associated with a large-scale stagnation, intense solar radiation, and minimum rainfall. Among meteorological parameters, rainfall has a negative correlation with O 3 . Its best example was observed from 2012 to 2015 when rainfall took place more as compared to remaining years. These precursors are mainly considered to be emitted in the present study sites due to primarily anthropogenic activities such as vehicular emission and biomass burning. Further, the difference in peak magnitude of NO x between winter and pre-monsoon season was about two-folds, which possibly indicate the vertical mixing as an important factor for the observed diurnal variations as photo-chemical loss is much lower in winter than that of the pre-monsoon season. The surface ozone, NO, and NO 2 show a signi cant diurnal variations at an urban site at Delhi ). The concentration of NO x shows highest magnitude in winter season followed by post-monsoon, pre-monsoon and monsoon season. Lowest concentration of NO x was observed during afternoon hours when O 3 showed a peak. This clearly illustrates the production of O 3 via photo-chemical oxidation of carbon like compounds by their reaction with OH radicals in presence of NO x . The daytime O 3 concentration builds up is a pronounced feature of urban polluted sites (Lal et al. 2000). The largest diurnal peak concentration was observed in summer (61.1 ppb), followed by winter (45.7 ppb), monsoon (27.1 ppb), autumn (35.6 ppb) and pattern is similar to that of other locations in India (Pulikesi et al. 2006;Reddy et al. 2010;Ahammed et al. 2006;Beig et al. 2007). But unlike the Northern China (Meng et al. 2009), O 3 concentrations were lowest in monsoon season. The larger variability in monthly averaged NO x was noticed during the winter season indicating larger anthropogenic sources near-surface and evolution of boundary layer height. On the other hand, the mean concentration of O 3 shows higher variability in summer season when photochemistry in presence of precursors and strong sunlight remains at peak. At this time, the process of O 3 production is faster than the process of destruction. The intercept shows low value during winter season, which are probably attributed to less active photochemistry (Table 1). In contrast to NO x as the secondary pollutant, O 3 exhibits a reverse seasonal pattern. It is mainly due to higher emissions related to heating and cooking, and photochemical production due to lower temperature and solar radiation. Surface ozone shows correlation coe cient of (-0.477*) with air temperature (Table 2). On average, the maximum vehicular in ow from Gulaba to Rohtang Pass (3978 m) in the year 2014 was recorded in the month of June (i.e., 44,588) and minimum in December (i.e., 13,696). Out of the total plying vehicles on the way to Rohtang Pass, majority of them were registered petrol operated vehicles followed by diesel operated. Petrol operated vehicles were recorded maximum in June (40,731), followed by October (24,300), whereas their minimum numbers stood in September (12,079). On the other hand, diesel operated vehicles were recorded maximum in August (4,195) and minimum in December (676) (Kuniyal et al. 2014). The vehicular activities reduced in the area only after the order of Hon'ble National Green Tribunal (NGT) wherein a regulation on diesel and petrol vehicles on Rohtang Pass area of Himachal Pradesh issued on 29th May, 2015. Here, total 1200 vehicles (i.e. 800 petrol and 400 diesel) were allowed in compliance to the NGT order. In between, there remains public resentment especially among the taxi operators because of limiting their livelihood options. In view of the same as well as accommodating more tourists, the previous limit was relaxed by NGT on June 19, 2017. At this point of time, daily limit increased from 1200 to 1300 (i.e. 60% petrol and 40% diesel) out of which 25% of the total were supposed to be reserved for the natives of Himachal State and 75% of the total allowed vehicles for the tourists outside Himachal Pradesh state. Daily 860 petrol, 440 diesel vehicles and up to 9 CNG buses are run to reduce the vehicular pollution activities in this ecologically sensitive and topographically fragile area. The main reasons for high O 3 concentration are both-the local as well as external sources. The local sources are supposed to be from the immediate area or region, while the external pollution sources could be transported from outside the region. Figure 5 represents the back trajectory ending at study site at 1000 m altitude. From the pathways of air-masses, it is made clear from the trajectories that in all the seasons, north-west, south-west, and south-east winds were found to be dominant. Most of the times, airmasses passed through Saudi Arabia, Iran, Iraq, Thar desert, Sahara and sub-Saharan region and polluted regions of Pakistan, Punjab, Haryana and Delhi. During observational time, the air masses reaching at Mohal (present study site) in different seasons were from northwest and west direction in winter, north, northwest, west and southwest direction in summer, southwest and east to southeast direction in monsoon and north-western direction in autumn during observation years from 2011 to 2015. In a cluster analysis, 4 clusters were grouped into back trajectories (Fig. 6). From 2011-2015 years, most of the trajectories were corresponding to the western ows (regional) contributing to 38% of the total air masses followed by 33% from the south-eastern, 20% from the westerly moderately moving and 8.3% from the north-western directions. Mean concentration of ozone associated with fast moving westerly and southerly air masses are signi cant only in monsoon season compared to those observed for other clusters. The CWT analysis has indicated the heavy or pollution laden air mass ow from the Arabian and Middle East countries.
Surface ozone has been continuously monitored since 2007 under ISRO-GBP with Atmospheric Chemistry Transport and Monitoring (AT-CTM) programme throughout India. Its observations from year to year at different observational sites have been reviewed and compared with our location. The seasonal and diurnal variation of surface O 3 over Mohal shows a simple pro le of a semi-rural area, which is changing season to season similar to other places of the country. Seasonally, O 3 concentration observed at Mohal-Kullu is lower than that of one of the hill station-Nainital which shows maximum O 3 during wintertime (Table 3) In winter, slow chemical loss is due to low temperature, minimal solar radiation and higher anthropogenic emissions from burning of fuel wood for cooking and heating in the surrounding villages. There has been a decrease in the percentage change of O 3 (-13) and NO 2 (-6), NO x (-3), and BC (-14) during 2011-2015 (Table 4). However, as an exception, there has been increase in NO (9) concentration and rainfall (8) amount. These results may be due to washout effect of rainfall, anthropogenic activities, forest res, automobiles, fuel and coal burning, etc.

Conclusions
During reporting period (2011)(2012)(2013)(2014)(2015), higher concentration of surface ozone was observed in the year 2013 indicating maximum as 74.6 ± 23.2 ppb, while its minimum values remained 15.8 ± 5.1ppb in 2015. Diurnal cycle of O 3 concentration at Mohal in the Kullu valley showed a unimodal peak during mid of a day at 15:00 hrs IST. This is mainly due to relatively a process of photo-oxidation in presence of strong sunshine and its precursors. While O 3 values were found low in early morning and late evening due to absence of sunlight. Concentration of O 3 was found to be increasing gradually after sunrise (07:00-08:00 hrs IST), attaining maximum concentration during afternoon (14:00-16:00 hrs IST). These gaseous pollutants, as considered to be primary pollutants, are mainly attributed to be produced due to anthropogenic activities such as vehicular emission and biomass burning. The percentage change of NO concentration shows its highest values up to 244.25% in 2011-13. There is a decrease in percentage change of O 3 (− 13), NO 2 (-6), and NO x (-3) from 2011 to 2015. At contrary, it is only NO that has shown a reasonable increase (9) due to vehicular activities and washout effect as rainfall increased by 8%. Analysis of meteorological data in context to O 3 showed its higher concentration days associated with a large-scale stagnation, intense solar radiation, and minimum rainfall. Among meteorological parameters, rainfall has a negative correlation with O 3 . In addition, the region has also a long range transport sources outside the region such as the Indo-Gangetic Plain and Middle East countries. Nevertheless, such longterm continuous measurement of trace gases and meteorological variables are crucial to better understand the surface ozone behaviour for half a decade in the areas which are still considered to be having limited information on the subject and are considered to be more prone to climate change.
Declarations Figure 1 Study site-Mohal in the Kullu valley, north-western Indian Himalaya. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 2
Diurnal variation of O3 and its precursors (NO, NO2, NOx) at Mohal in the north-western Indian Himalaya.   Concentrated Weighted Trajectory (CWT) analysis for identifying sources of surface ozone. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 6
Cluster analysis of air mass at the receptor site-Mohal-Kullu for ve years (January 2011 to December 2015). Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.