Introduction

Atmospheric aerosols play an important role in climate change and atmospheric chemistry in the troposphere because of the direct effect on the radiative balance of the Earth’s atmosphere by scattering or absorbing incoming shortwave radiation and through indirect effect by acting as Cloud Condensation Nuclei (CCN) (Andreae and Crutzen 1997; Rastogi and Sarin 2006). Carbonaceous aerosol, which is a large fraction of atmospheric aerosols, played a crucial role in air pollution, adverse health, visibility reduction and climate effects. They constitutes a significant fraction in fine particles (PM2.5), and it could be accounted for up to 40 % of mass of PM2.5 in urban atmosphere (Seinfeld and Pandis 1998), however, in rural atmosphere it accounted up to 77 % (Ancelet et al. 2013). The carbon fractions composing ambient particulate matter are commonly divided into two types, depending on their interaction with light and their volatility (Chow and Watson 1993; Birch and Cary 1996; Schauer et al. 2002). Light absorbing particles are referred to as elemental carbon (EC), and non-absorbing particles are referred to as organic carbon (OC) (Chow et al. 2001). A significant fraction of OC is water soluble, making it important into particle–cloud interactions (Decesari et al. 2000) and influence many heterogeneous reactions involving atmospheric particles and trace gases (Begum and Minar 2012). Although, the study of EC and OC are important in atmospheric chemistry and physics, information concerning their spatial and temporal variability is quite limited. Carbonaceous aerosol, which consists of OC and black carbon (BC) or EC, are of the major research interest due to their ubiquitous nature and abundance, particularly in fine-mode (PM2.5) (Lonati et al. 2007; Dutkiewicz et al. 2009). On the other hand, OC is either emitted directly from sources (primary) or formed from the condensation of low volatility products of the oxidation of hydrocarbons (secondary). Carbonaceous particles (elemental and organic carbon) both terms represent rather complex species since even the elemental carbon contains a number of functional groups such as alcoholic, phenolic, carbonylic, carboxylic (Akhter et al. 1985). Moreover, the surface of elemental carbon particles contains numerous adsorption sites which are capable of enhancing catalytic processes. It is the second most important component of global warming, after CO2 (Hansen et al. 2000; Jacobson 2001). Organic carbon represents a large variety of organic compounds such as aliphatic, aromatic compounds, acids, etc. On the other hand, OC is mainly a scattering medium and exerts a negative climate forcing influence (Houghton et al. 2001). Thus, carbonaceous studies demand more attention and research, thus measurement of OC and EC is important.

A growing body of evidence suggests that both primary and secondary organic aerosol (POA and SOA) undergo multistage heterogeneous oxidation and evaporation condensation processes as they age (Robinson et al. 2007; Donahue et al. 2009; Weitkamp et al. 2009). EC, POA, and VOCs are emitted in the course of the combustion process. A significant portion of POA is semi-volatile and evaporates on dilution as it moves away from the source and reacts with oxidants in the gas phase, forming less volatile material that condenses back to produce SOA (Robinson et al. 2007). POA containing carbonyl compounds also participates in heterogeneous oxidation reactions in the aerosol phase to produce more polar and water-soluble compounds (Weitkamp et al. 2007, 2009). Thus, a significant portion of POA is converted into SOA during its atmospheric life span (Donahue et al. 2009). SOA is oxygenated material containing a large fraction of water-soluble organic compounds (Blando and Turpin 2000; Matsui et al. 2009; Asa-Awuku et al. 2010) and is continuously oxidized in the atmosphere via heterogeneous reaction mechanisms as it ages (Robinson et al. 2007; Donahue et al. 2009).

Secondary particles are formed in the atmosphere through a combination of chemical and physical reactions among primary pollutants such as the direct emissions of gaseous pollutants such as oxides of sulfur, nitrogen, ammonia, etc. It is difficult to predict ambient concentrations of these secondary air pollutants because they are also affected by long-range transport, and their formation is dependent on unstable factors such as ambient temperature and humidity. Analysis of the carbonaceous component of particles has traditionally distinguished between EC and OC where the EC is loosely associated with the terms soot and OC is everything else. There is a separable issue of whether the material collected on filters is representative of carbonaceous material in the ambient atmosphere, as it is well-known that semi-volatile organic compounds, for example, can lead to significant variations in collected material depending on the collection conditions (e.g., Turpin et al. 1993; Eatough et al. 1993).

Agra and Delhi are located in Indo-Gangetic Plain (IGP) region, which is one of the most polluted regions in northern India and is influenced by various emission sources such as fossil fuel combustion from industries, vehicular, thermal power plant, post-harvesting of agricultural residue burning, and a lot of minor sources such as textile mills, brick kilns etc. In thermal power plants, they are using coal which is also affecting carbonaceous aerosols concentrations (Schwarz et al. 2008; Salma et al. 2004). Apart from this, due to its topography and being situated in the foot hills of Himalayas, long-range of transport of mineral dust imparts significant variability of aerosols over this region. Due to above said importance, the present study has been conducted in northern part of India at two different well-known places, i.e., Agra which one of the oldest city “world heritage city” in the presence of the Taj Mahal and Delhi, which is the capital city of India with huge population (~18 millions). Thus the objectives of this study are: to determine the mass concentration level of fine particulate matter (PM2.5) at Agra as well as Delhi, to characterized fine particulate matter in terms of carbonaceous species (OC and EC), to determined the relationship between PM2.5 and its associated carbonaceous species, and tried to find out their sources over study region of India. During the same period, soot particle called black carbon (BC) was also studied and data was compared with EC.

Experimental methods

Site description

Agra (27°10′ N 78°02′ E), which is located in the northern part of India and is about 204 km of south of Delhi in the state of Uttar Pradesh. It is one of the most famous tourist spots in India because of the presence of Taj Mahal, situated on the west bank of river the Yamuna. The climate during winter is cooled with moisture; the temperature ranges from 3.5 to 30.5 °C and downwind West–North–West, i.e., WNW 9.4 % and North–North–West 11.8 %. The atmospheric pollution load was higher when wind blows from WNW because oil refinery at Mathura (50 km far from the centre of the city) which is located there and pollutants may be transported. Agra has about 1.3 million inhabitants and the population density is about 21,150 persons km−2 with 386,635 vehicles registered and 32,030 generator sets. In Agra, 60 % pollution is mainly due to vehicular emissions from three national highways (NH-2, NH-3, and NH-11) cross the city (Pipal et al. 2011). Delhi (28° 35′ N; 77° 12′ E; 217 masl), the capital city of India is situated between the rich rain-washed Gangetic plain in the east and semi-arid tracts of Rajasthan to the southwest which is about 1,100 km away from the nearer coast of Arabian Sea and considered as the fourth most polluted city in the world with respect to suspended particulate matter having ~18 million inhabitants (Goyal and Sidharta 2002, Tiwari et al. 2012). In addition to this, 4.8 million registered vehicles, three coal-based thermal power plants, and 125,000 small- and medium-scale industrial units are situated in Delhi (Srivastava et al. 2009). The major sources of pollution over Delhi are power plants, road transport, small–scale industries and domestic cooking. The cold and dry weather associated with the northwesterly coming from northwest Himalayas is characteristic of the winter season over central IGB (Mishra and Shibata 2012).

Sample collection and analysis

Aerosol samples were collected at two sites, the first site is a roadside (Agra) and another site is Indian Institute of Tropical Meteorology Pune, New Delhi, on the rooftop of the building (15 m above the ground level) during winter season (2011–2012) (Fig. 1).

Fig. 1
figure 1

Google map indicating sampling site over Agra (AGR) and Delhi (DEL)

Full size image

Aerosol samples were collected for 24 h on quartz filter papers by a medium volume sampler (model: APM 550, Envirotech, New Delhi) at flow rate, 16.6 l min−1). There is more chance in varying the PM2.5 mass and its carbonaceous species in winter due to vehicular activities and local sources like wood, cow dung, and crop residue burning for space heating. The quality control in monitoring was made to check the daily flow rate calculation to make sure that the fluctuation in flow rate was within the range. The cascaded impactor was used to classify aerosols depending on their sizes of 10–2.5 μm (PM10–2.5) and less than 2.5 μm (PM2.5). The impactor filter was changed after 48 h of sampling or when the filter gets clogged. Periodic cleaning of the sampler was done to make the sampler dust free so that the dust on the sampler may not be counted with the mass concentration of the sample. These precautions were help in monitoring for OC concentration to overcome for positive (adsorption) or negative (volatilization) artifacts of the carbon concentration in carbonaceous aerosols at the time of sampling (Turpin et al. 2000; Mader and Pankow 2001). Backup filters have been used to evaluate and to correct for the positive sampling artifact. Artifacts were also prevented by deploying by sampling setups such as diffusion denuders / particle trap impactor to remove gaseous OC from the sample air-stream and the use of filters in series, impregnated or not by different adsorbents (Mader et al. 2003). It was suggested that the particle trap impactor which provides a sharp size cutoff with maximum collection efficiency, moreover, the efficiency was constant at high particle loadings. These properties of particles trap impactors make them suitable for sampling of carbonaceous aerosols.

Before exposure, the quartz fiber filter papers were pre-baked at 450 °C for 2 h to eliminate the organic species. The filter papers were conditioned in desiccators at 20–30 °C and relative humidity of 20–35 % in humidity controlled room for 24 h and weighed on electronic balance (A&D, model GR-202, Japan). After the samples collection, they were kept in desiccators for 24 h and weighed. The conditioned and weighed PM2.5 filters were placed in cassettes and placed inside polyethylene zip-lock bags and taken to the field for sampling to avoid contamination of the filters on the way. Laboratory blank filters were also collected to reduce gravimetric bias due to filter handling during and after sampling. Filters were handled only with tweezers coated with Teflon tape to reduce the possibility of contamination. After weighing, the samples were wrapped in aluminum foil and sealed in polyethylene zip-lock bags and stored in desiccators until the time of analysis of OC-EC to prevent the degradation of organic compounds due to photo-oxidation. The exposure filters were analyzed within a month period of sampling. Blank sample was also taken on the sampler and all data were reported against the field blank value.

The particle concentrations were determined gravimetrically by the difference in their weights before and after the sampling. Then those exposed filter papers were analyzed for the concentrations of OC and EC by semi-continuous thermal/optical carbon analyzer (Sunset Laboratory Inc. Model 4L) using NIOSH 5040 (National Institute for Occupational Safety and Health) protocol (Birch and Cary 1996). An aliquot of sample filter (2.1 cm2) is stepwise heated in a furnace up to 840 °C in a non-oxidizing atmosphere (100 % He); furnace is then cooled to 550 °C and filter is stepwise heated to 850 °C in an oxidizing atmosphere (90 % He, 10 % O2), during each temperature step, evolved carbon is converted to methane and detected by a flame ionization detector. A calibration is performed at the end of each analysis by introducing a known amount of methane gas into the oven and measuring its constant response. The carbon evolved before the split line is considered as OC, which evolved after the split line is quantified as EC (Tiwari et al. 2013a). The measurements of BC mass concentrations were carried out by seven-wavelength Aethalometer (Model AE-31, Magee Scientific Company, Berkley, CA, USA) with temporal resolution of 5 min. The BC was estimated from the light attenuation at 880 nm using a specific mass absorption cross-section of 16.6 m2g−1(Hansen 2005; Tiwari et al. 2013a). The data of BC was retrieved on high-resolution data of 5-min interval and made hourly data for interpretation. The statistical analysis such as correlation and t-test of data were performed using Microsoft Excel and SPSS software.

Meteorological status over AGR and DEL locations

The meteorological parameters such as temperature (Tem.), wind direction (WD), wind speed (WS), and relative humidity (RH) were monitored at AGR and DEL stations by automatic weather monitor (Envirotech WM251). Average values of Tem., WS, WD, and RH were 16.7 °C, 2.5 m s−1, 174°, and 62.1 % at AGR while at DEL it was 15.89 °C, 9.50 m s−1, and 67.28 % for Tem., WS and RH respectively. Figures 2 and 3 show the day to day variability of meteorological parameters at AGR and DEL, respectively. The relative humidity was higher more than 46 % for most of the study period; it is due to temperature along with the climatic conditions mostly dry. These meteorological parameters may play crucial role in atmospheric organic and elemental carbon, chemistry, and photochemistry of mineral dust particles. Moreover, these meteorological parameters can change the atmospheric conditions, which resulted in variations of PM mass and its carbonaceous aerosol concentrations level in distinction season in northern part of India. The meteorological conditions may also act as an important factor in the distribution of carbonaceous aerosol from long range transportation (detail discussed in “OC/EC ratios and sources of carbonaceous aerosols” section.

Fig. 2
figure 2

Day to day variability of meteorological conditions (wind speed (WS), wind direction (WD), atmospheric temperature (Tem.), and relative humidity (RH) at AGR

Full size image
Fig. 3
figure 3

Day to day variability of meteorological conditions (wind speed (WS), atmospheric temperature (AT), and relative humidity (RH) at DEL

Full size image

Results and discussion

Mass concentration levels of fine aerosols particles (PM2.5) over AGR and DEL

Figure 4 presents the mass concentration of PM2.5 measured during sampling period in winter of 2011–2012 (November to February) from Agra and Delhi, respectively. The average PM2.5 concentration was 165.42 ± 119.46 μg m−3 at Agra while 211.67 ± 41.94 μg m−3 at Delhi. It is substantially higher than the annual standard stipulated by central pollution control board called national ambient air quality standard (NAAQS) (http://www.cpcb.nic.in/National-Ambient-Air-Quality-Standards.php) for fine (40 μg m−3) and by the United States Environmental Protection Agency (USEPA) (PM2.5 = 15 μg m−3) standards (http://www.epa.gov/air/criteria.html). The average concentration of PM2.5 was 27 % higher at DEL than AGR in entire period of study. The results are similar and sometimes much higher than the earlier reported values of fine particles over IGP region such as at Agra (PM2.5 = 90.16 ± 7.21 μg m−3, Pipal et al. 2011), Delhi (PM2.5 = 148.4 ± 67 μg m−3, Dey et al. 2012; PM2.5 = 97 ± 56 μg m−3 and 122.30 ± 90.70 μg m−3, Tiwari et al. 2010, 2013b; PM2.5 = 123 ± 87 μg m−3, Guttikunda and Calori 2013), Lucknow (PM2.5 = 101.05 ± 22.50 μg m−3, Pandey et al. 2012), and Kanpur (PM2.5 = 95 μg m−3, Sharma and Maloo 2005). Apart from these recent studies, some studies also reported similar values obtained for PM2.5 during the winter (190 ± 109 μg m−3) at Gual Pahari, a station close to Agra and different to New Delhi. However, the level of PM2.5 in Agra and Delhi is higher than Beijing, China (115 μg m−3), as reported by He et al. (2001). Study done by Watson et al. (2002) has reported the PM2.5 concentration of 138 μg m−3 during wintertime at the California’s San Joaquin Valley, USA, which is lower than that of Agra and Delhi. Simultaneously, black carbon (BC) was also measured at Delhi and was 10.6 μg m−3 which is approximately 6 % of fine particles and their day to day variability is depicted in Fig. 5. Similar contribution was reported by Hussain et al. (2007) ~5–15 % at urban air of Lahore; ~7 to 15 % at Kanpur by Tripathi et al. (2005), ~4 to 15 % at Delhi by Ganguly et al. (2006), 7 % at Hyderabad by Latha and Badrinath (2005), whereas lower in Pune ~2.3 % by Safai et al.(2007). Monthly BC was also separated and was in the order of December (17.6 μg m−3) > January (11.7 μg m−3) > November (8.3 μg m−3) > February (5.1 μg m−3). The concentration of BC in December is higher due to meteorological effect (Tiwari and Singh 2013.

Fig. 4
figure 4

Day to day variability of mass PM2.5 and carbonaceous species at DEL (A) and AGR (B)

Full size image
Fig. 5
figure 5

Day to day variability of black carbon over DEL during study period

Full size image

Table 1 shows the monthly mean mass concentrations of PM2.5 at both sites along with carbonaceous aerosols. Month wise, higher concentration was observed in December (285.87 μg m−3) at Agra (AGR) and in November (241.92 μg m−3 ) at Delhi (DEL) while lower concentration was in February (109.08 μg m−3, AGR; 219.50 μg m−3, DEL) at both sites. The monthly characteristics of PM2.5 trend observed is as December > January > November > February at AGR whereas at DEL, the trend was as November > January > December > February. A large variation in mass concentrations PM2.5 was observed due to different emissions sources at both places. Over all significantly higher at the DEL may be due to combined effect of climatic conditions and anthropogenic emissions such as vehicular exhaust, waste incineration, coal and biomass and bio-fuel combustion, and re-suspended soil dust. These highest concentrations of fine particles during winter period were due to low wind speed, low mixing height, as well as low temperature. During stable and cold conditions, the pollutant could not disperse and accumulation of pollutants. Whereas the concentrations of PM2.5 was also significantly higher than NAAQS standard at AGR which is influenced by traffic emission thus may be due to vehicular activities, soil dust and fugitive and re-suspension of road dust. Apart from these sources the accumulation of particles especially in fine size due to thermal inversion is the major cause for these high concentrations of PM2.5 over north central region during the winter period, and the source for these particles could mostly be from the local emission sources such as combustion of biomass, biofuel, and fossil fuel. Awasthi et al. (2011) reported that higher PM2.5 over northern part of India was due to biomass burning which was transported from western part of India.

Table 1 Monthly mean mass concentrations of PM2.5, carbonaceous species (OC and EC), OA, and TCA at AGR and DEL
Full size table

Carbonaceous aerosols

Figure 4 also shows the day to day variability of OC, EC, and TC along with mass of PM2.5, during study period over AGR and DEL. Average concentration of OC, EC, and TC were 69.96 ± 34.42, 9.53 ± 7.27, and 79.01 ± 38.98 μg m−3 at AGR, while 50.11 ± 11.93, 10.67 ± 3.56, and 60.78 ± 14.56 μg m−3 at DEL, respectively. Higher OC was observed in AGR and DEL it is due to more biomass burning during winter. Monthly OC was higher in December at AGR and DEL and the lower was in February at both sites. In the case of EC, it follows the similar trend as OC at both sites. The EC was observed higher than BC over DEL during study period. Generally, the EC may be lower as compared to BC (Hussain et al. 2007), but at present study, it was taken by offline techniques with limited time period so we got lower value of BC over. The higher concentrations of OC and EC in November are due to the long-range transport of primary pollutants as well as local emissions such as festival effect called Deewali and burning of fossil fuel along with meteorological impact during December and January (Tiwari et al. 2013b). OC concentrations ranked in the order of December > January > November > February whereas EC concentrations were in order December > November > February > January at AGR while at DEL, it followed January > November > December > February for OC and the increasing order for EC, i.e., November > December > January > February. High contributions of OC and EC in the month of November and December are attributed to more incidences of domestic biomass burning and impact of late afternoon traffic loads. In addition to this, during festival period, people play with fire crackers and produce huge amount of smoke over Delhi (Attri et al. 2001) during winter this leads to formation of secondary aerosols by condensation processes fostered by the low temperature and the low boundary layer height. In fact, condensation of semi-volatile gaseous precursors is enhanced by low temperatures, whereas temperature inversions limit their dilution and allow for supersaturating conditions with the exceedance of the equilibrium vapor pressure. Other processes that can contribute to cold-season OC levels are the adsorption of semi-volatile OCs onto existing solid particles and the dissolution of soluble gases that can undergo reactions in particles (Pandis et al. 1992).

The total carbon aerosol (TCA) was determined as the sum of organic aerosol (OA) and EC. OA is calculated as OC multiply by factor 1.6 (Feng et al. 2009; Satsangi et al. 2012; Turpin and Lim 2001). The earlier studies done on carbonaceous aerosols reported that the conversion factor 1.6 for urban aerosols and 2.1 for non-urban aerosols (Chen and Yu 2007; Satsangi et al. 2012). The thousands of individual organic compounds constitute the OC fraction, the composition of which varies widely in both space and time. This prevents a bottom up approach to derive at OA, i.e., obtaining OA by summing up individual organic groups or functional groups (Welthagen et al. 2003).

TCA was higher at AGR (114.41 μg m−3) than DEL (99.58 μg m−3). The percentage contribution of TCA in PM was at AGR (58.67 %) than DEL (41.67 %). The monthly characteristic of TCA is as follows: December > January > November > February at Agra, while January > November > December > February at Delhi (Fig. 6). In a recent study at Kanpur, Ram and Sarin (2010) have reported about one-third contributions of carbonaceous aerosols to the PM10 mass during the winter period. They further reported a three- to fourfold increase in the concentration of OC and suggested biomass burning (wood fuel and agricultural waste) emission as a dominant source over the station. It is also due to long-range transportation of aerosol during winter, the enhancement of aerosol concentrations due to massive biomass burning over western part of India especially Haryana and Punjab states (Kaupp and McLachlan 1999; Awasthi et al. 2011). Apart from these, the IGP and semi-arid tract region also contribute to air pollutants over the northern part of India. It is important to measure carbonaceous species in particles in such areas in this region to understand transport and transformation characteristics of anthropogenic air pollutants and their effects on various aspects of regional and global environment changes. Carbonaceous species in particles may play an important role in global climate change by affecting radiative forcing (Crutzen and Andreae 1990; IPCC 1996).

Fig. 6
figure 6

Comparison of TCA and OC/EC ratios between AGR and DEL

Full size image

OC/EC ratios and sources of carbonaceous aerosols

Table 2 shows the OC/EC during the present study over AGR and DEL along with other studies carried out in India and other parts of the world. The ratios of OC to EC concentrations (OC/EC) is used to assess emission and transportation characteristics of carbonaceous aerosols (OC and EC), secondary organic aerosol formation and different removal rates by deposition of OC and EC (Cachier et al. 1996). Ambient OC to EC ratios greater than those characteristic of the primary emissions for a given area are used as an indicator for the SOA formation. The OC to EC ratio is also influenced by the measurement methods because OC and EC values greatly depend on the sampling and analytical methods (Turpin et al. 2000; Jacobson et al. 2000). If the OC/EC ratio value is found to be greater than 2, it suggests the formation of secondary organic aerosols apart from primary emission sources. Therefore, the OC/EC has been used to indicate the presence of primary and secondary organic aerosols (Gray et al. 1986). Some studies have reported that the OC/EC ratios values from different sources such as diesel- and gasoline-powered vehicles if the OC/EC = 1–4.2 (Schauer et al. 2001), residential coal smoke has OC/EC = 2.5–10.5 (Chen et al. 2006), and biomass burning with OC/EC = 7.7 (Zhang et al. 2007). The OC/EC ratios are 13.75 and 5.43 at AGR and DEL, respectively, this can be said on the basis of reported ratios that sources may like is from vehicular exhaust, residential coal smoke, and biomass and diesel burning. The ratio was similar in all months over DEL however large variations were seen at AGR (Fig. 6) with higher during February 2012 and lower in November 2011. The relatively higher ratio values during winter may be due to the change in the relative strength of their sources. As mentioned in “Carbonaceous aerosols” section, the higher OC at AGR may be more biomass burning as compared to DEL because it is surrounded by rural environment. The OC and EC values reported at Agra of this study are higher than those of all other studies (Satsangi et al. 2012, Agra; Pachauri et al. 2013, Agra; Tiwari et al. 2013b, Delhi; Ram and Sarin 2010, Kanpur; Rengarajan et al. 2007, Hisar; Ram et al. 2010b, Allahabad; Ram et al. 2008, Mt. Abu; Rastogi and Sarin 2009, Ahmadabad; Venkataraman et al. 2002, Mumbai; Salam et al. 2003, Dhaka; Duan et al. 2007, Hongkong; Lin and Tai 2001, Taiwan; Smith et al. 1996, Lahore; Feng et al. 2006, Beijing; Salma et al. 2004, Budapest; Viana et al. 2007, Europe; Feng et al. 2009, Shanghai; Fan et al. 2003, Canada; Lewtas et al. 2001, USA; and Lonati et al. 2007, Italy. At DEL, the ratios were similar at other locations mentioned earlier references. In comparison with DEL and AGR, very low ratio were observed in Dhaka (Salam et al. 2003), Hongkong (Duan et al. 2007), Taiwan (Lin and Tai 2001), Beijing (Feng et al. 2006), Budapest (Salma et al. 2004), and Europe (Viana et al. 2007). In order to know, the transport pathways of the carbonaceous species in the atmospheric aerosols over the northern part of India, monthly 7-days air mass backward trajectory analysis was carried out at height of 500 m 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 (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. The back trajectories are very important to identify the origin of source regions and the transport pathways of aerosols to reach the measurement site and also to investigate the aerosol properties and types (Bian et al. 2011).

Table 2 The OC, EC, and OC/EC of the present study along with other studies done in the World
Full size table

Figures 7 and 8 show that monthly 7-day air mass backward trajectory clusters analysis at Agra and Delhi during study period which supports the transport of aerosols from the long range transportation from Thar Desert and other part of the world over northern Indian region. During the months of December, January, and February air masses appear to be transported from the Neighboring countries such as Pakistan, Afghanistan, Iran, and Turkmenistan, and South Indian part and India ocean at low altitudes passes over Thar Desert before entering into Agra and IGB (Tiwari et al. 2012). While in the month of November, it was seen from north Indian region and local sources. Marine environment (Arabian Sea as well as Bay-of- Bengal) and carry mostly sea salt aerosols while traversing through the continental mainland before they reach the measurement site (Reddy et al. 2011). Whereas at Delhi, it was seen from Middle Eastern and Arabian Sea covering south Indian part and Thar Desert for the months of January and February, while during the month of November and December, the air masses over DEL passes mostly in Indian sub-continental region as well as ocean. Ramachandran and Rajesh (2007) reported the impact of long-range transport and meteorological conditions in higher carbonaceous aerosol. The air masses from different source regions lead to the formation of different aerosol types (Reddy et al. 2011).

Fig. 7
figure 7

Monthly 7-day air mass backward trajectory clusters analysis at Agra during Nov 2011, Dec 2011, Jan 2012, and Feb 2012 (vertical lines show standard deviation, representing latitudinal air mass spreading)

Full size image
Fig. 8
figure 8

Monthly seven day air mass backward trajectory clusters analysis at Agra during Nov 2011, Dec 2011, Jan 2012, and Feb 2012 (vertical lines show standard deviation, representing latitudinal air mass spreading)

Full size image

Assessment of SOC, POC, and SOA in AGR and DEL

The secondary organic carbon (SOC) fraction in aerosol is either estimated using the EC tracer method (Cabada et al. 2004) or by adding up all the oxidation products found in aerosols (Schauer et al. 1996). The EC-tracer method has been widely used in many studies (Docherty et al. 2008; Lin et al. 2009) which are very suitable for the regions where the source apportionment studies are still lacking such as India, since it only requires ambient OC and EC concentrations. In the EC-tracer method, ambient OC to EC ratios greater than those characteristic of the primary emissions for a given area are used as an indicator for the SOA formation. The estimation of secondary organic aerosol (SOA) is also important because it plays an important role in the formation of haze, visibility degradation, climate, and health. It is very difficult to directly separate the secondary organic carbon (SOC) and primary organic carbon (POC) from total OC and estimation of SOA. Turpin and Huntzicker (1995) have suggested a mathematical equation for quantification of SOC using EC as the tracer for POC. The mathematical equation for the measurement of SOC is given below:

$$ \mathrm{SOC}=\mathrm{TOC}-\mathrm{EC}\times {\left(\mathrm{OC}/\mathrm{EC}\right)}_{\mathrm{primary}} $$

Where (OC/EC) primary is the ratio for primary sources. The EC tracer was also applied in many studies, they were estimated (OC/EC)primary from atmospheric concentrations assuming that either the lowest observed OC/EC ratio represents the pure primary aerosol (Castro et al. 1999), or the data for the days when SOA formation could obviously be excluded were averaged (Strader et al. 1999).Chow et al. (1996) proposed a ratio value (2.0) between OC and EC to estimate SOC, however, the primary ratio of OC/EC is usually not available because it is affected by many factors such as the type of emission sources as well as its temporal and spatial variation, atmospheric temperature, etc. Castro et al. (1999) discussed in details about (OC/EC) primary which represented by the observed minimum ratio [(OC/EC) min]. In addition, in the EC tracer method, it is supposed that the primary OC and EC have the same source, and representing as OC/EC ratio for the primary aerosol ((OC/EC)primary), EC is primary in origin. The obligations can be met in the PM2.5 size fraction, since OC in the coarse fraction originates from other sources such as pollen, spores, plant debris, tyre rubber, and paved road dust than EC. OC can sometimes be solely primary in origin under favorable meteorological conditions and close to the main primary source of OC. Kleeman et al. (2000) was measured a ratio 1.1 for pure primary aerosol from diesel engines. It has to be emphasized that the calculation approaches yield semi-quantitative information only because of the large uncertainties associated with them. Apart from this, the good correlation between OC and EC indicates similar sources, it may be transported or localized. Due to these reasons, it is assumed that the primary OC and EC have the same source and considered as OC/EC ratio for the primary aerosol ((OC/EC)primary). If the OC/EC is not significant, since EC are mostly emitted from anthropogenic activities, in that case, OC and EC are both emitted from different sources and or not transported. In this, (OC/EC)primary emission rates vary by sources, and, therefore, the ratio is influenced by meteorology, time fluctuations in emissions, and local sources such as vehicles, fireplaces, cooking operations (Seinfeld and Pandis 1998).

The monthly SOC and POC were estimated on the basis of above equation and depicted in Table 3 for AGR and DEL. The average concentration of SOC and POC at AGR were 48.16 and 26.52 μg m−3 while at DEL it was 38.78 and 27.55 μg m−3, respectively. In the case of POC, similar concentrations were observed at both places but in the case of SOC, it was 24 % higher over AGR as compared to DEL it is due to the high concentration of OC over AGR. Secondary organic aerosol (SOA) was also determined and was 42 % higher at AGR than DEL which indicates that the secondary aerosol formation at AGR are more, it is due to rural environment and higher concentrations of coarse mode particles. The SOA contribution in PM2.5 was also estimated and was ~32 and 12 % at AGR and DEL, respectively. The correlation coefficient analysis was performed to measure the degree of linear relationship between two variables. For source identification, the relationship and t test between PM2.5 and OC/EC species were studied (Table 4). The correlation coefficient between OC and PM2.5 is positively significant at AGR (0.78) and DEL (0.60) while in the case of EC to PM2.5, it is less significant at both places which indicate the major contribution of OC in PM mass as compared to EC. The correlation between OC and EC at both places were significantly correlated (0.47, AGR and 0.53, DEL) it is due to the common sources emission for OC and EC over study region. The t test carried out indicates that the there is significant differences in PM2.5, OC, and EC concentrations between both sites.

Table 3 Level of SOC, POC, and SOA in AGR and DEL during study period
Full size table
Table 4 Correlation coefficient between PM2.5 and carbonaceous species at AGR and DEL
Full size table

Conclusions

Carbonaceous species in particles may play an important role in global climate change by affecting radiative forcing. Therefore, it is important to easure carbonaceous species in particles at such areas in the region to understand transport and transformation characteristics of anthropogenic air pollutants and their effects on various aspects of regional and global environment changes. Due to lack of in situ measurements of atmospheric carbonaceous aerosols over northern part of India, continuous monitoring of fine particles and its carbonaceous constituents were monitored at two places (AGR and DEL) in northern part of India during winter season of 2011–12. The atmospheric aerosol (PM2.5) were collected at AGR and DEL using medium volume air sampler (offline measurement) and characterized for its carbonaceous aerosols [organic carbon (OC) and elemental carbon (EC)] along with BC (online) and their results were presented. The average PM2.5 concentration was 165.42 ± 119.46 μg m−3 at Agra while at Delhi it was 211.67 ± 41.94 μg m−3. It is substantially higher than NAAQS and USEPAS. Simultaneously, Black Carbon (BC) was 10.6 μg m−3 which is approximately 6 % of fine particles. Monthly BC was also separated, it was in the order of December (17.6 μg m−3) > January (11.7 μg m−3) > November (8.3 μg m−3) > February (5.1 μg m−3). The average concentration of PM was 27 % higher at DEL than AGR of the total PM concentrations. The average concentration of OC and EC were 69.96 ± 34.42 and 9.53 ± 7.27 μg m−3, respectively. Total carbon (TC) was 79.01 ± 38.98 μg m−3 at Agra, while it was 50.11 ± 11.93 (OC), 10.67 ± 3.56 μg m−3 (EC), and 60.78 ± 14.56 μg m−3 (TC) at Delhi. The OC/EC ratios were 13.75 and 5.45 at Agra and Delhi, respectively. The higher OC/EC ratio at AGR indicates that the formation of secondary organic aerosol and from variable primary sources. Significant correlation between PM2.5 and its carbonaceous species were observed indicating similarity in sources at both sides. The average concentration of SOC and POC at AGR were 48.16 and 26.52 μg m−3 while at DEL, it was 38.78 and 27.55 μg m−3, respectively. In the case of POC, similar concentrations were observed at both places but in the case of SOC, it was about 24 % higher over AGR as compared to DEL due to the high concentration of OC over AGR. Secondary organic aerosol (SOA) was 42 % higher at AGR than DEL which indicates that the secondary aerosol formation at AGR is more; it may be due to the rural environment. The SOA contribution in PM2.5 was also estimated and was ~32 and 12 % at AGR and DEL, respectively. Back trajectories was calculated and indicates that the long-range pathways of aerosols over north part of India from the Pakistan, Afghanistan, Iran, Turkmenistan, Middle Eastern, Indian Ocean, and Arabian Sea covering south Indian part and passes over Thar desert before entering into northern Indian and IGB. Being highly loaded of fine particles along with carbonaceous aerosol, it is suggested to take necessary and immediate action in mitigation of the emission of carbonaceous aerosol in the northern part of India.