Seasonal Chemical Characteristics of Atmospheric Aerosol Particles and its Light Extinction Coefficients over Pune , India

The present study has been conducted to characterize atmospheric aerosol particles in terms of carbonaceous species and ionic constituents for a yearlong period at Pune, India. This study provides the evidence for the ionic chemistry, secondary aerosols formation, temporal variability and its climatic effect in the atmosphere. The average concentrations of PM2.5 and PM10 were 109.6 ± 23.2 and 166.9 ± 4 μg m, respectively, by far exceeding National Ambient Air Quality (NAAQ) and World Health Organization (WHO) standards. Seasonal analyses indicated that PM2.5 and PM10 mass concentrations were higher in the post-monsoon followed by the winter season and lower during the monsoon period. The average concentrations of organic carbon (OC) and elemental carbon (EC) were 31.3 ± 7.4 and 4.2 ± 2.4 μg m for PM2.5, while, 34.2 ± 6.2 and 5.0 ± 2.3 μg m for PM10, respectively. OC and EC data splits into seasons and their mass loadings were in the order of post-monsoon > monsoon > winter > summer for OC and for EC, it was as winter > post-monsoon > summer > monsoon. The overall chemical analysis revealed that particulate matter (PM) consist higher concentrations of OC followed by cations and the lowest one is EC. The ionic composition analysis indicated that cations were the abundant parts of PM in comparison to anions and Na and SO4 were at a higher concentration amongst all the ionic species. The estimated light extinction coefficient (bext) of the aerosol particle was 291.2 ± 55.3 Mm during the study period. Further apportionment of particle extinction coefficient was estimated and the contributions of light scattering coefficient by particles (bsp) were OC (45%), (NH4)2SO4 (17%), NH4NO3 (8%) and coarse mass (12%), while, the contribution of light absorption coefficient by particle (bap) was 18% (EC). This indicates that in the present study the abundance of aerosol particles are more scattering in nature in comparison to absorption. The average value of Aerosol optical depth (AOD) was 0.46 and their positive correlation with anions and relative humidity (RH) showing same properties, while, in the case of EC, it showed contrast nature with respect to climate effect. Trajectory analysis indicated that the air masses appear as a result of long-range transportation during summer and monsoon period while during the winter and post-monsoon seasons local manmade activities showed dominant influence. Keyword: Ionic chemistry; OC and EC; Light extinction; (NH4)2SO4 and NH4NO3; AOD; Sources.


INTRODUCTION
Atmospheric particulate matter (APM) is a multifarious mixture of many various chemical constituents originating from an assortment of natural and anthropogenic sources.These particles are variable by concentration as well as to their physical, chemical and morphological characteristics (Pipal et al., 2011;Sielicki et al., 2011;Satsangi and Yadav, 2014).In the atmosphere, particles are produced by incomplete combustion and non-combustion process such as fossil fuel, biomass burning, soil materials, sea spray and dust; and are also formed from various chemical reactions.
Based on emission sources and formation mechanism, aerosol particles can be classified into primary and secondary aerosols which play a vital role in visibility reduction, precipitation, deterioration of regional air quality, etc. (Vallius, 2005;Xin et al., 2014).In addition to this, they have capacity to change the radiative balance of the atmosphere by absorbing and scattering solar radiation.They also change the microphysical process of clouds and perform as ice nuclei (IN) and cloud condensation nuclei (CCN) (Ramanathan et al., 2001;Wan et al., 2015;Zhang et al., 2015).This phenomenon varies widely depending on particle size, characteristics and chemical compositions of APM (Muller et al., 1999).The carbonaceous matter is one of the large fractions of PM chemical constituent in the atmosphere.It is having a significant fraction (~40-50%) of mass in fine particles (PM 2.5 ) of urban atmosphere (Seinfeld and Pandis, 1998) as elemental carbon (EC) and organic carbon (OC).
Moreover, EC is extensively used as an indicator of air pollutants which are emitted from diesel vehicles to estimate their contribution to fine particles.It is expected that the EC concentration and the ratio of EC to total carbon in atmospheric aerosols can be higher around regions of elevated vehicular density (Pandis et al., 1992;Turpin and Huntzicker, 1995;Pipal et al., 2014;2014a).Apart from carbonaceous species, other chemical species such as Na + , Ca 2+ , Mg 2+ , K + , F -, Cl -, SO 4 2-, NO 3 -and PO 4 3-are also quantifiable in ambient PM due to their atmospheric solubility in water in the atmosphere at definite condition such as high temperature and relative humidity (Pipal et al., 2010;Ram and Sarin, 2011;Aldabe et al., 2011;Kong et al., 2014).Moreover, in order to understand the air quality and atmospheric condition, AOD is a very important parameter and widely used to estimate the concentrations of PM (Gupta et al., 2006).In India, there exists a dearth of studies focusing on chemical constituents of PM 2.5 and PM 10 particles and their effects on light extinction during various seasons.In Indian context, only single study has been conducted by Tiwari et al. (2014a) on light extinction due to chemical species in short period (winter) over Delhi which reported that scattering type aerosols dominate ~76% over the absorbing type (~24%).
In respect of that, the present study has been constructed for a yearlong period covering the above facts hoping to be useful for policy makers and general public to perceive the current baseline of PM and its chemical constituents along with influence on climate as well as on health.Owing to increased urbanization, industrialization, Pune has been selected as a study area of the research subject.The city has witnessed a dramatic rise in the number of residences, office buildings and manufacturing facilities along with the number and density of motor vehicles and migrating population in the past few years attributed to it being a major IT Hub in India.The present study has been conducted with a aim of the following objectives (i) to determine the seasonal trend of PM (PM 2.5 and PM 10 ) mass concentration at urban site of Pune (ii) to characterize PM particles in terms of major cations and anions along with organic and elemental carbon (OC and EC) (iii) to estimate primary organic carbon (POC) and secondary organic carbon (SOC) particles formation in the atmosphere (iv) to estimate light extinction coefficients (b ext ) and its contribution from individual chemical species (v) and to determine the sources and pathways of particulate mass and its associated chemical species at measurement site along with their relationship with AOD.

Description of Study Area
Pune (18°32′ N, 73°51′ E) is located 560 m (1,840 ft) above sea level on the western margin of the Deccan plateau.It is situated on the Leeward side of the Sahyadri mountain range, a barrier between the Arabian Seas and is surrounded by hills and mountains.Pune has a tropical wet and dry climate with average temperatures ranging between 20 to 38°C.The monsoon lasts from July to September with rainfall and temperatures ranging from 10 to 28°C (Pipal and Satsangi, 2015).It is one of India's most important automotive hubs, with some domestic and international auto-giants manufacturing units.Pune also has hundreds of large and small IT companies and has well-established glass, sugar and forging industries.It is rapidly growing city in terms of industrial installations, vehicular population and also urbanization due to the boom in housing industry since last few years.Fig. 1 represents aerosol loading over Pune periphery indicating by AOD during study period along with sampling locations in the map of Pune city.

Sample Collection
PM 2.5 and PM 10 samples were collected with the help of medium volume air sampler (model: APM 550, Envirotech, New Delhi, flow rate: 16.6 L min -1 ) for 24 h during a yearlong period from May 2013-April 2014 (covers all the seasons) at the roof of Chemistry Department, SPPU, Pune.PM samples were collected twice in a week on 47 mm diameter quartz fiber filter (QFF) papers for PM 2.5 and PM 10 separately to perceive their monthly and seasonal variability.The filter papers were conditioned in desiccators at 20-30°C and relative humidity of 20-35% for 24 hrs before and after the sampling to remove the moisture content of the filter papers and weighed on an electronic balance (Shimadzu, AUX 220; sensitivity ± 0.1 mg).The repeated measurement of the filter weights provides an uncertainty of ± 1 mg which corresponds to an overall error of 15% in the aerosol mass concentration.After that, the weighed samples were kept in cassette followed by wrapping of Al-foil sealed in polyethylene zip-lock bags and stored in desiccators till the time of analysis to prevent the degradation of organic compounds due to photo-oxidation.The separate cascade impactor was used to classify aerosols particle depending on their sizes of less than 2.5 µm (PM 2.5 ).The impactor filters were changed after 48 h of sampling or when the filter gets clogged.The quality control in monitoring was made by checking the daily flow rate calculation and to make sure that the fluctuation in flow rate was appropriate.Periodic cleaning of the sampler was done to make the sampler dust free, consequently the dust on the sampler may not be counted with mass concentration of the sample (Pipal et al., 2014).

Analysis of PM Samples (PM 2.5 and PM 10 ) Chemical Analysis
PM 2.5 and PM 10 collected samples were characterized in term of organic and elemental carbon, cations and anions during the study period using different sophisticated techniques.The concentrations of carbonaceous species such as organic and elemental carbon (OC and EC) were determined by the OC-EC analyzer (Sunset Laboratory) (Birch and Carry, 1996;Feng et al., 2009;Satsangi et al., 2012;Pipal et al., 2014).The analytical details for the analysis of OC and EC have been expressed in our earlier publications (Pipal et al., 2014;Pipal and Satsangi, 2015;Tiwari et al., 2016).The calculated detection limit for OC and EC were 0.26 and 0.04 µg m -3 , respectively.Concentrations of anions (F -, Cl -, NO 3 -and SO 4 2- ) and cations (Na + , Mg 2+ , K + and Ca 2+ ) were determined by Ion Chromatograph (DIONEX-2000, USA) using analytical column Ion Pac-AS15, anion micro-membrane suppressor ASRS-II, 38 mM sodium hydroxide/potassium hydroxide as eluent and mili-Q water (~0.05 µS) as regenerator (Bisht et al., 2015).Repeated calibrations through standards were performed (~3-4 times) to distinguish the variations of the ionic species and also to ensure the accuracy and precision of results.The replicate samples were also analyzed at every 10 th injection for accuracy and precision of results.UVvisible absorption spectrophotometer (Shimadzu, UV-1800) was used for the measurement of NH 4 + concentrations.The blank samples were also analyzed in the same manner to finalize the data and its quality control.

Meteorological Status over Measurement Site
As a result of atmospheric processing, meteorological conditions such as wind speed (WS), wind direction (WD), temperature (Temp) and relative humidity (RH) are playing a crucial role in changing physicochemical properties of PM.Therefore, the meteorological parameters were monitored during the study period by automatic wind monitor (Wind monitor 271, Envirotech, New Delhi).Table 1 illustrates the monthly variations of meteorological parameters along with the direction of winds from various locations towards measurement site.Average values of WD, WS, Temp, and RH were 193 ± 4.9 degree, 2.5 ± 1.5 m s -1 , 24.9 ± 3°C and 73 ± 16.8%, respectively during the entire study period.WS was almost constant during the monsoon, postmonsoon followed by the winter and it was higher during the summer season.Correlation analysis was also carried out between the meteorological parameters and PM and its constituents.Very small positive correlations (r ≤ 0.4) were observed among them which indicate the abundance of the local emission sources contributed to mass concentration of PM and its constituents.In the month of March and April, the correlation between PM and wind speed was found to be relatively high which suggest that along with the local sources, long range transport had also contributed during these months(Table 1).This fact was also supported by trajectory analysis (Fig. 2).

Trajectory Analysis for Source Identification over Measurement Site
Season wise trajectory analyses were performed at the height of 500 m during the study period.It was carried out to know the impact of local source and transportation pathways of APM towards the measurement site (Pune).It is based on National Oceanic and Atmospheric Administration (NOAA) Hybrid Single Particle Lagrangian Integrated Trajectories (HYSPLIT) model (Draxler and Rolph, 2003).The back trajectory analysis provides a three dimensional such as latitude, longitude and altitude description of the pathways followed by air mass as a function of time by using National Centre for Environmental prediction (NCEP) reanalysis wind as input to the model.Fig. 2 shows season wise seven day's backward trajectory analysis at Pune which indicate long-range transportation of PM and its constituents from various neighbouring areas in addition to regional local sources.From the Fig. 2, it is also inferred that the air masses are transported from the Arabian Sea covering a few part of Middle Eastern to reach SW part of India during the summer and monsoon period.Whereas in  winter and post-monsoon seasons, aerosol pathways appears from the south-west region of India due to the dominance of local and regional manmade sources such as biomass and fuel burning, vehicular emissions and other local activities (Safai et al., 2014;Pipal and Satsangi, 2015).This indicates that sources of PM and its chemical constituents over measurement site are from different local anthropogenic activities as well as dust transported from other regions during the various seasons.

AOD Measurement
AOD values were obtained at 550 nm wavelength from May 2013 to April 2014 from NASA data source by using Moderate Resolution Imaging Spectroradiometer (MODIS) (http://gdata1.sci.gsfc.nasa.gov/daacbin/G3/gui.cgi?instanc e_id=MODIS_DAILY_L3).The Daily AOD data was averaged (N = 30) for monthly basis to see the monthly variability of AOD during the study period.

Temporal and Seasonal Characteristics of PM 2.5 and PM 10 over Measurement Site
The average concentrations of PM 2.5 and PM 10 were 109.6 ± 23.2 and 166.9 ± 4 µg m -3 , respectively during the whole study period.The observed values of PM were substantially higher than the annual standards stipulated by Indian National Ambient Air Quality (NAAQ) (http://www .cpcb.nic.in/National-Ambient-Air-Quality-Standards.php)(40 µg m -3 for PM 2.5 and 60 µg m -3 for PM 10 ) and World Health Organization (WHO) (http://www.euro.who.int/document/E87950.pdf)(10 µg m -3 for PM 2.5 and 20 µg m -3 for PM 10 ), respectively.Fig. 3 shows the monthly concentration of PM over Pune which indicates that higher concentration was found in the month of Nov for PM 2.5 and in the month of May for PM 10 particles.While lowest PM mass was found in the month of Sept for both sizes of PM.Table 2 shows the mass concentration of PM 2.5 and PM 10 and its statistical parameters such as average, minimum, maximum and standard deviation along with samples frequency during each month of study.
The data of PM 2.5 and PM 10 were divided on seasonal basis (four seasons) over Pune as; winter (Dec-March), summer (April-June), monsoon (July-Sept) and postmonsoon (Oct and Nov).Table 3 depicts the seasonal values of PM 2.5 and PM 10 and their mass loadings were found in order of: summer (PM 2.5 : 123.3 and PM 10 : 219.6 µg m -3 ), winter (PM 2.5 : 106.3 and PM 10 : 156.8 µg m -3 ), post-monsoon (PM 2.5 : 120.6 and PM 10 : 143.3 µg m -3 ) and monsoon (PM 2.5 : 92.9 and PM 10 : 140.7 µg m -3 ).This seasonal analysis indicates that the significant variations were observed in PM 2.5 and PM 10 concentrations during different seasons.This may be due to different regional source activities carried out in various seasons such as dust transportations from the construction site and excess of vehicular activities near the sampling site (Pandithurai et al., 2008, Tiwari et al., 2013).In addition, meteorological parameters also influence the mass concentration of PM during various seasons.
The higher mass concentrations of PMs during the summer was due to the dominance of local sources which transported    1) during this period.This was also confirmed by the trajectory analysis which supports the long range transportation of PMs and constituents during the summer period (Fig. 2).The abundance of PM 2.5 and PM 10 concentrations during winter period was also due to the low boundary level inversion with low temperature.Due to these conditions, pollutants could not disperse quickly, resulting in higher concentrations at ground level.Lower concentrations were identified during monsoon period for PM 2.5 and PM 10 clearly reflecting the wash out of the air pollutants from the atmosphere.
Moreover, the present concentration of PMs (PM 2.5 : 110 and PM 10 : 167 µg m -3 ) when compared with national and international standards, exceeded the limits by ~47% of NAAQS for both size particles while in the case of WHO, it was found to be higher ~83 and 78% for PM 2.5 and PM 10 , respectively.On the basis of this, it is concluded that Pune, a highly escalated city in India, is polluted with respect to PM and its associated chemical constituents due to the emission from anthropogenic sources which are more responsible for respiratory problems of Pune inhabitants (Salve et al., 2006) and climate change (Pipal and Satsangi, 2015).

Organic and Elemental Carbon along with OC/EC
Carbonaceous aerosols are one of the largest fractions of PM accounting to 40-60% of fine aerosol particles (Seinfeld and Pandis 1998;Pipal et al., 2014).Therefore, they play a vital role in global climatic change due to radiative forcing.The average concentration of OC and EC were 31.3 ± 7.4 and 4.2 ± 2.4 µg m -3 for PM 2.5 while 34.2 ± 6.7 and 5.0 ± 2.3 µg m -3 for PM 10 , respectively.Fig. 4 shows the monthly variability in OC and EC concentration reported to be higher during Nov and Jan, respectively while least during the month of April for OC but in the case of EC, it was found to be lower in Sept. The seasonal mass concentration distribution of OC and EC in PM is depicted in Fig. 5 which indicates the increasing order as: post-monsoon > winter > monsoon > summer for OC but in the case of EC, the trend observed was winter > post-monsoon > summer > monsoon.The reported average OC and EC values (OC: 32 µg m -3 and EC: 6 µg m -3 ) at Pune was compared with other worldwide studies along with their adopted analysis method depicted in Table 4.This inferred that the concentrations are similar or higher than the studies reported by Satsangi et al. (2012) at Agra; Rastogi et al. (2009) at Ahmadabad; Rengarajan et al. (2007) at Hisar; Duan et al. (2007) at Hong Kong;Lin and Tai (2001) in Taiwan and Feng et al. (2009) at Shanghai.Whereas, the concentrations of organic and elemental carbon are lower that the studies reported by Tiwari et al. (2013) in Delhi; Ram and Sarin (2010) at Kanpur; Salam et al. (2003) in Dhaka and Pipal et al. (2014) at Agra and at Delhi.
The OC/EC ratios are used to assess the emission and transportation characteristics of carbonaceous aerosols, secondary organic aerosol (SOA) formation and different source origin (Cachier et al., 1996).Therefore, OC/EC ratio was calculated in the present study to determine primary and secondary organic aerosols and their source origin (Gray et al., 1986).OC/EC ratios exhibited an average value of 6.3 and 5.0 for PM 2.5 and PM 10 , respectively during the study period.Fig. 6 shows the levels of SOC and POC along with OC/EC ratios over Pune.
The SOC and POC were estimated by using the EC tracer method which is very suitable for the Indian regions where the source apportionment studies are still lacking, since it requires only ambient OC and EC concentrations ( Docherty et al., 2008;Lin et al., 2009;Pipal et al., 2014).The mathematical equation for quantification of SOC using EC as the tracer for POC which is given below: SOC = TOC -EC × (OC/EC) primary (1) The values of SOC and OC/EC ratio were higher during the month of November but in the case of POC, it was higher during the months of Dec-Jan.The possible sources of carbonaceous species were determined on the basis of OC/EC ratios and also with the help of trajectory analysis.This indicates that carbonaceous aerosols was mostly emitted from the regional local activities such as diesel and gasoline powered vehicular exhaust and biomass burning (garden cutting, fallen leaves and paper burned by the side of roads) which were abundant around the vicinity of site (Schauer et al., 2001;Feng et al., 2009).
Organic matter (OM) was also estimated in present study.It is a conversion factor of OC to OM by average ratio of the molecular weight of organic compounds to the molecular  weight of carbon in the compounds.These compounds are formed with oxygen (O), nitrogen (N), hydrogen (H) and carbon (C).Turpin and Lim (2011) concluded that factor of 1.6 ± 0.2 was a better for estimation of organic matter for urban areas, while the factor of 1.9-2.3 is used for aged aerosols and 2.2-2.6 for aerosols originating from biomass burning.Therefore, in the present study OM was estimated by total carbon multiplied by a conversion factor of 1.6 (as the present site comes under urban environment) (Satsangi et al., 2012;Pipal et al., 2014;Pipal and Satsangi, 2015).
The average concentration of organic matter was 60.2 and 61.3 µg m -3 for PM 2.5 and PM 10 , respectively and was higher during the post-monsoon and winter period.The total organic matter (TOM) was derived as a summation of OM and EC.The average concentration of TOM was 65.3 and 67.7 µ m -3 in PM 2.5 and PM 10 , respectively.This indicates that the total organic carbonaceous matter was contributed by ~60 and 40% of PM 2.5 and PM 10 particles, respectively.This clearly indicates the abundance of carbonaceous matter in fine fraction (PM 2.5 ) of particles in comparison to PM 10 particles.

Characteristics of Inorganic Ions and its Chemistry in PM
The inorganic components are investigated in a detailed knowledge of PM inorganic components, their concentration range, size distribution, major sources and role in secondary particle formation.The average concentration of F -, Cl -, SO 4 2-and NO 3 -were 2.6, 5.8, 6.2 and 3.2 µg m -3 in PM 2.5 while in PM 10 , these were 1.6, 6.6, 7.4, 2.4 and 2.8 µg m -3 , respectively.In addition to this, Na + , Mg 2+ , K + , Ca 2+ and NH 4 + were also characterized and their concentration was found to be 10.3, 3.2, 9.8, 3.3 and 3.1 µg m -3 in PM 2.5, while, in PM 10 they were 11.9, 6.4, 8.7, 6.6 and 3.8 µg m -3 respectively.Ionic species is contributing ~38% of PM where the cations (21%) are dominant in comparison to anions (17%) into total ionic PM chemical constituents.Most of the ions such as SO 4 2-, Cl -and NO 3 -are found to accumulate in fine particles (PM 2.5 ), while, the cations such as Na + , Ca 2+ , K + and Mg 2+ were mostly abundant in PM 10 .
Amongst the ions, Na + and SO 4 2-have the higher concentration which may be due to the acidic species (SO 4 2-and NO 3 -) produced by the oxidation of the gaseous precursors (SO 2 and NO x ) in the atmosphere under certain conditions such as abundance of anthropogenic activities and higher temperature.Further, amongst anions, SO 4 2-and NO 3 -were higher during the study period that may be attributed to the long-range transportation of chemical species carried through their precursor gases in the atmosphere (Seinfeld and Pandis, 1998).The study done by Lee and Sequeira (2002) suggested that SO 4 2-aerosols are the major species responsible for dropping the visibility over Hong Kong.In another study, Guo et al. (2010) found that secondary inorganic aerosols such as NH 4 + , NO 3 -and SO 4 2contributed 55% of the fine aerosol mass over Beijing during summer time was responsible for the district scale pollution.The abundance of SO 4 2-and NO 3 -indicates the formation of HNO 3 and H 2 SO 4 which are water soluble and will be rapidly absorbed into atmospheric water droplets and react with atmospheric aerosol particles forming sulfates and nitrates salts.Moreover, the ratios between NH 4 + with NO 3 -and SO 4 2-inferred that the formation of (NH 4 ) 2 SO 4 and NH 4 NO 3 for rapid reaction with H 2 SO 4 and HNO 3 and gas phase NH 3 .Whereas, the limestone particles (CaCO 3 ) can be converted to CaSO 4 and salt particles (NaCl) of marine origin can be converted to Na 2 SO 4 and NaNO 3 with the displacement of hydrogen chloride gas (Satake and Mido, 1994;Behera and Sharma, 2010).In the atmosphere, ammonia plays a major role for secondary aerosol formation.Atmospheric ammonia reacts with both H 2 SO 4 and HNO 3 acids to form fine aerosol particles such as (NH 4 ) 2 SO 4 and NH 4 NO 3 which is shown by following reactions (Bauer et al., 2007). (2) The overall averages annual contribution of major chemical species such as cations, anions and OC and EC arises towards to PM are shown in Fig. 7.This indicates that OC is the higher contributors followed by inorganic ions (cations and anions) and EC.This further indicates that abundance of scattering type aerosol over measurement site in comparison to absorbing pollutants (details description is given in the later section of light extinction).Table 3 also shows the seasonal concentrations of cations and anions along with OC and EC.This trend shows that the ionic species (cations and anions) and OC concentration was higher during the post-monsoon followed by the summer and lower in winter period.Whereas, EC concentration was higher during the winter period followed by the postmonsoon and lowest during the monsoon season.Higher EC concentration during winter may be due to abundance of combustion activities around the site.

Secondary Aerosol Particles
Due to the abundance of organic matter, as discussed in the previous section, the organic and elemental carbons have diverse optical and chemical properties.Thus, they show the impact on light extinction in different way as organics are scattering of solar light and EC act as an absorbent type species (Ackerman et al., 2000;Wang et al., 2011).Apart from OC, the SO 4 2-, NO 3 -and NH 4 + also produced their secondary particles in the atmosphere.
Secondary particles are mostly composed of OC that forms a major component (up to 90%) while EC shows relatively low contribution ≤ 10% (Ram et al., 2008).Moreover, POC and EC are released directly from fossil fuel combustion and biomass burning while SOC are formed by oxidation of volatile organic reactive species in the atmosphere.Therefore, POC, SOC, SOA and inorganic aerosols such as SO 4 2-and NO 3 -were determined during the study period.The calculated SOC and POC levels in PM are given in Fig. 6.The average concentration of POC, SOC, and SOA were 11.5, 34.8, 20.4 µg m -3 for PM 2.5 and 10.6, 27.0 and 21.9 µg m -3 for PM 10 (details of calculation are given in our previous section), respectively while the concentration of SO 4 2-was found to be higher amongst the anions during study period over Pune.This indicates that the secondary particles are abundant over Pune in comparison to primary particles.This fact was also confirmed by light extinction coefficients analysis of individual chemical species (details are given in later section).These particles in atmosphere varied with size range due to chemical reactions which may occur by the mechanisms resulted in low vapor pressure products by the reaction of gases.

Estimation of Light Extinction Due to Chemical Constituent
At present, the studies of atmospheric aerosols are also widespread interest due to its light extinction and subsequent impacts on the visibility degradation and climate change.Titos et al. (2012) suggested that the light extinction of PM is highly associated with particle size and chemical compositions of aerosols such as SO 4 2-, NO 3 -, OC and EC.The term 'light extinction coefficient' means the loss of light in the atmosphere from a directly transmitted beam due to absorption and scattering of radiations.
In order to know, the effects of the chemical composition of PM on light extinction coefficient (b ext ), it is a better parameter to examine the impact and role of different chemical species on scattering and absorption of solar radiation in the atmosphere.The light extinction (b ext ) could be reconstructed from the mass concentrations of reformed species, the extinction efficiency of the species and the correction factor of humidity.Interagency Monitoring of Protected Visual Environment (IMPROVE) developed an Eq. ( 4) for estimation of light extinction of individual chemical composition of atmospheric aerosol particles in the atmosphere. .The unit of b ext is Mm -1 (inverse megameters), unit of chemical species is µg m -3 and the hygroscopic growth term [f(RH)] is unit less.
The light extinction coefficients was calculated by multiplying the concentrations of each measured chemical constituents as well as the mass of atmospheric aerosols by composition specific light extinction efficiency and sum of all major ionic species (IMPROVE, 2006(IMPROVE, , 2011)).This empirical Eq. ( 1) was used for estimation of b ext and separating the individual contribution of chemical species into light extinction.The adopted equation has been used globally by various authors in previous studies to calculate the light extinction of individual chemical species (Cao et al., 2012;Titos et al., 2012;Tiwari et al., 2014a;Cheng et al., 2015).During the study period, the calculated average b ext was 291.2 ± 55.3 Mm -1 varied from 176.2 (August), 352.1 (December) Mm -1 which is mainly due to chemical species (OC, EC, (NH 4 ) 2 SO 4 , NH 4 NO 3 and coarse mode particles present in the atmosphere.Fig. 8 shows the monthly estimated b ext due to chemical constituents for entire study period indicating higher values in the month of Dec-Jan and lower during the month of Aug.The seasonal analysis trend indicates elevated values during the winter (317.3Mm -1 ) followed by the post-monsoon (289.3Mm -1 ), summer (275.1 Mm -1 ) and least for monsoon (293.9Mm -1 ).This variability in b ext is mainly due to atmospheric aerosols being generally dominated with PM 2.5 particle during winter and post-monsoon, whereas PM 10 particles as well as hygroscopic aerosols are dominated during the summer and monsoon period.Fig. 9 depicts the contribution of each chemical species to extinction coefficient which signified that the OC is the higher (45%) contributor for light extinction coefficient followed by EC (18%), (NH 4 ) 2 SO 4 (17%), coarse (12%) and lower was NH 4 NO 3 (8%).Apart from this, the contributing NO 2 was the minor amount (0.02%) to light extinction coefficient of the course of study.The NO 2 was measured by Aeroqual integrated multi-gas monitor (IQM 60) utilizing Aeroqual's proprietary Analytic GSS Technology integrated with photo-ionization detector (PID) and nondispersive infra-red (NDIR) sensors to achieve precise measurement of NO 2 gas pollutant.
The estimated light extinction coefficients (b ext ) of chemical species of aerosol particles were further inferred that the abundance of light scattering coefficient by particles (b sp ) was 82% and it is due to OC (45%), (NH 4 ) 2 SO 4 (17%), NH 4 NO 3 (8%) and coarse (12%) while light absorption coefficient (b ap ) due to EC (18%).The contributions by each chemical species to b ext along with its comparison with other earlier studies are presented in Table 5.It was found that very few studies have been carried out on b ext due to chemical species of atmospheric aerosol particles.In present study, the contribution of OC was similar to other studies (Zhang et al., 2013;Tiwari et al., 2014a;Xiao et al., 2014) while the higher than the other studies (Jung et al., 2009;Cao et al., 2012;Jun et al., 2012).Of the above discussions, it is clearly inferred that Pune atmosphere is have abundance of scattering type of aerosols (82%: due to OC, (NH 4 ) 2 SO 4 , NH 4 NO 3 and coarse) than absorbing type aerosols i.e., EC; 18%.

Relationship between AOD and Chemical Species of PM
The measurement of extinction of the solar beam due to aerosol particles which prevents the direct sunlight to reach the ground by the aerosol particles is called aerosol optical depth (AOD).The measurement of AOD was also retrieved by using MODIS data source at 550 nm wavelength during the study period.The average mean value of AOD was found to be 0.4 ± 0.2 over sampling site.Fig. 10 shows the average monthly values of AOD during the study period.The monthly trend indicates that the AOD values were higher in July (0.9) and lower in Feb (0.2).The seasonal analysis of AOD followed in the order of monsoon (0.6 ± 0.07), summer (0.66 ± 0.04), post-monsoon (0.38 ± 0.03) and winter (0.32 ± 0.02).The monthly variability in AOD was seen as increasing trend from May (0.4 ± 0.1) to July (0.9 ± 0.1) and decreasing value observed in August (0.5 ± 0.1) and Feb (0.1).AOD increased further in March (0.3 ± 0.1) to April (0.4 ± 0.2).The correlation coefficient was also determined between AOD and chemical constituents associated with PM over measurement site.AOD was found to be negatively correlated with EC (r = -0.6 0.5) which indicates that they have the distinct properties with respect to incoming solar radiation.Whereas, the positive correlation between AOD and anions (r = 0.5-0.2),AOD with RH (r = 0.6) and WD (r = 0.4) was observed.The good correlation between AOD and anions indicates that both variables are showing the same properties  (scattering in nature of light) in the atmosphere.It is to further state that anions such as NO 3 -and SO 4 2-are participating in the formation of secondary particles due to scattering of radiations.Whereas the negative correlation was observed with elemental carbon (EC) thereby indicating diverse properties as EC is a good absorbent of light due to its specific surface properties thus playing an important role in the aerosol climatic forcing and visibility degradation (Jacobson et al., 2001).In view of the above discussion, the study on scattering and absorption coefficients are very important that depends on the chemical constituents and size distribution aerosol particles as mentioned in the previous section.

CONCLUSIONS
This study examines the seasonal tendency of PM and its associated chemical species along with its effects on light extinction coefficients and its contribution from each chemical species.The average concentration of PM 2.5 and PM 10 exceeded ~2.7 times NAAQS and ~8-11 times WHO annual standards of PM.Seasonal analysis of PM indicates that its mass concentration was varied amongst the seasons as; higher during post-monsoon followed by winter and lower during the monsoon period.The carbonaceous analysis of PM reveals that organic carbon (OC: ~34 µg m -3 ) was higher in comparison to elemental carbon (EC: ~5 µg m -3 ) over Pune.The OC followed the seasonal trend as post-monsoon > monsoon > winter > summer while in the case of EC the trend was as winter > post-monsoon > summer > monsoon.The ionic study indicates that cationswere highe contributor to PM in comparison to anions and Na + and SO 4 2-were found to be higher concentrations amongst the all the ions.Total chemical species indicates that OC was the higher contributor towards PM followed by cations and EC.This study also inferred the abundance of secondary particles (SOC: 27-35 µg m -3 ) in comparison to primary particles (POC: 10-12 µg m -3 ) and they are rich in fine particles than inhalable particles.The light extinction analysis inferred that the abundance of more scattering in nature (82%) in compared to absorption nature (18%) of particles constituents.The positive correlation of AOD with anions and RH was observed which indicates that they have same properties while in the case of EC, it was oppositely indicating negative nature of each other.Season wise air mass backward trajectories indicate that the contribution of longrange transportation of particulates and it allied chemical constituents towards the measurement site from India and other countries region including the Arabian Sea.It is extensively accepted hypothesis that physical (mass and size distribution), chemical and optical characteristics of PM along with their climatic nature are necessary to determine the light extinction coefficients and radiative forcing.Therefore, temporal variations of PM and its associated chemical constituents and its role in secondary aerosol particles are essential to perceive the health and climatic problems in versatile environment.Thus, the present study is very useful to environmentalist and policy makers to develop strategies for mitigation of air pollution along with its awareness to people and its forecasting in highly growing environment.

Fig. 2 .
Fig. 2. Season wise seven days backward trajectory analysis during the entire study period

Fig. 6 .
Fig. 6.Month wise concentration of SOC and POC along with OC/EC ratio.

Fig. 7 .
Fig. 7. Contribution of chemical species to PM mass in entire study period.

Fig. 9 .
Fig. 9. Individual contribution of chemical species into light extinction coefficient in atmosphere during study duration over Pune.

Table 1 .
Month wise description of meteorological parameters (WD: wind direction, WS: wind speed, Temp: temperature and RH: relative humidity). Pune

Table 2 .
Monthly mass concentration of PM 2.5 and PM 10 along with their statistical parameters and samples frequency.

Table 3 .
Season wise contribution of chemical species during study period (µg m -3 ).

Table 4 .
Comparison of observed OC and EC values with other studies along with their adopted analysis methods.

May June July Aug Sep Oct Nov Dec Jan Feb MarchApril
Variations in light extinction coefficient due to different chemical species in Pune atmosphere during study period.

Table 5 .
Contribution of Light extinction coefficient due to individual chemical species (%).