Source Impact Analysis Using Char-EC/Soot-EC Ratios in the Central Indo-Gangetic Plain (IGP) of India

This study measured the carbonaceous aerosols in the atmospheric fine particulate matter (fPM; dia. < 2.5 μm, n = 102 samples) above an urban location (Prayagraj) on the central IndoGangetic Plain (IGP) for 1 year (December 2016–November 2017). During the study period, the fPM mass concentrations exhibited very high variability, ranging from 22 to 367 (avg. ± SD: 149 ± 87) μg m–3, and approximately 40% of the annual average consisted of carbonaceous aerosols (organic carbon [OC], water-soluble OC [WSOC], and elemental carbon [EC]). Furthermore, several diagnostic ratios (EC/fPM, OC/EC, char-EC/soot-EC, and WSOC/OC) indicated that the characteristics of the carbonaceous aerosols significantly differed between seasons, with emission sources, seasonal meteorology, and atmospheric chemistry driving the variation in abundance. Specifically, our trajectory analysis revealed an association between northwesterly air masses originating in northwestern India and higher concentrations of fPM and carbonaceous aerosols at the study site, which mainly occurred during the post-monsoon season and winter. However, anthropogenic emissions from local sources dominated the fPM and carbonaceous aerosols in summer and the monsoon season, during which we also observed the influence of air masses of mixed (continental/marine) origin. To identify the sources of the carbonaceous aerosols and evaluate their contributions, we analyzed the char-EC/soot-EC ratio in combination with the OC/EC ratio. Cross plots of these ratios identified biomass burning and fossil-fuel combustion as the largest sources during winter and the post-monsoon season, and summer and the monsoon season, respectively. Our results demonstrate the advantages of utilizing the char-EC/soot-EC ratio rather than the OC/EC ratio as a tracer of these two sources in the IGP region.


A C C E P T E D M A N U S C R I P T
a sterilized glass centrifuge tube. Sonication was done by placing sample tubes in an ultrasonic   161 Secondary OC (SOC) was estimated using the EC-tracer method (Cabada et al., 2004;162 Srivastava et al., 2018;Turpin and Huntzicker, 1995). This method assumes that the minimum 163 OC/EC ratio involves OC contribution only from the primary emission source (Castro et al., 1999;164 Pavuluri et al., 2011). Accordingly, the minimum OC/EC ratio can be used for the estimation of 165 primary OC (POC) and SOC. In this study, SOC was calculated for each season separately using 166 the following equations:

173
Temporal variability of fPM mass concentration 174 Temporal variability of fPM (dia. < 2.5µm), mass concentration is shown in Fig 3a. The higher during winter and post-monsoon, followed by that in summer, and lowest in the monsoon 180 season (Fig. 3a).

A C C E P T E D M A N U S C R I P T
9 different sites of Indo-Gangetic Plain (IGP) (Kulshrestha et al., 1998;Ram and Sarin, 2010;193 Satasangi et al., 2012;Singh et al., 2017). The variability in the abundance pattern of ambient 194 aerosols was found to be highly dependent on emission sources, seasonal meteorology, and 195 atmospheric chemistry, which has been discussed in the following sections.

204
In other words, the OC and EC annually averaged mass fraction (%) in fPM were 18% and 10%,

Concentration weighted trajectories (CWT) and cluster analysis
Integrated Trajectory (HYSPLIT) model with an input of GDAS Meteorological Data at an 216 atmospheric height of 500 meters above mean sea level (Draxler and Rolph, 2003 the seasonally segregated trajectories were clustered with measured fPM mass concentration based 222 on a pre-existing algorithm of the nearest angle-distance function (Sirois and Bottenheim, 1995).

223
Moreover, we have also carried out the CWT analysis, which is also a receptor model used widely and contributed equally (50%). Cluster #1 represents the influence of long-range transport from 237 northwest regions extending from Afghanistan and Pakistan (Fig. 4a). However, cluster #2 is

A C C E P T E D M A N U S C R I P T
11 representative of regional transport from upwind IGP (Fig. 4a). Summer months have witnessed 239 the influence of air masses associated with three major clusters. Cluster #1 comprised 36% of 240 AMBTs from the western side of the study site and originated over gulf countries and traversed 241 through the Arabian Sea (Fig. 4b). Both cluster #2 and 3 comprised equally (32% of AMBTs) and 242 followed a tract of southeastern (the Bay of Bengal and eastern states) and northwestern upwind 243 region, respectively (Fig. 4b).
The monsoon period has experienced by and large the influence of 244 south-westerly air masses that originated over the northern Indian Ocean (Fig. 4c). Nearly 54% of 245 air-masses traverse through the Arabian Sea, whereas the remaining 46% of the air masses have a 246 signature from the marine atmospheric boundary layer (MABL) over the Bay of Bengal (Fig. 4c).

247
The lowest concentrations have been observed during the monsoon period attributable to deeper 248 boundary layer height, high wind-speed and reduced anthropogenic activities (particularly BB). with easterly wind-system as predicted by cluster #2 (Fig. 4d). The higher concentrations of fPM 255 in this study during post-monsoon and winter seasons and the prevailed northwesterly wind system 256 suggest that local sources in conjunction with long-range transport are responsible for degrading 257 air quality over the study region.  Fig. 5f). The data set of each 274 season were subjected to statistical one-way ANOVA analysis in order to assess the significant 275 difference in the seasonally averaged mass concentration/ratio of analyzed parameters (Table S1).

276
The ANOVA analysis has been carried out using the SPSS (version-20) software. boundary layer height), and atmospheric chemistry (e.g. secondary aerosols formation).

A C C E P T E D M A N U S C R I P T
13 Previous studies have reported distinct ratios of OC/EC and char-EC/soot-EC for different 285 source identification, as represented in Table 3. To identify probable sources of carbonaceous 286 aerosols at the study site in each season, we have constructed a cross plot employing the mean 287 value of existing datasets of OC/EC and char-EC/soot-EC ratios for various sources viz. biomass 288 burning, biofuel burning, coal combustion, and vehicular emissions (Fig. 6). The mean value of 289 OC/EC and char-EC/soot-EC (taken from existing literature as shown in Table 3) for biomass 290 burning was reported to be 6.0 ± 4.6 and 4.8 ± 2.2, for biofuel burning 3.3 ± 0.9 and 5.4 ± 2.8, for was conducted for each season (Fig. S1). The ratio analysis of the winter season has shown that 298 there is an influence of both biomass and fossil fuel emissions (Fig. 6a). However, a strong 299 correlation of K + with char/soot-EC (r= 0.85) as compared to OC/EC (r = 0.22) with higher mean 300 K + (7.1 µg m -3 ) value clearly indicated the dominance of biomass burning sources in winter months 301 (Fig. S1a). Biomass burning was also observed as a major source in post-monsoon season owing 302 to the strong correlation of char/soot-EC ratios with K + (r = 0.92) with higher average levels of K + 303 (5.3 µg m -3 ) (Fig. S1d). However, no correlation of K + with OC/EC (r = -60) can be attributed to ratio is quite opposite to that of SOC/OC ratio (Fig. 7). Both POC and SOC fractions were found 323 to exhibit significantly (p < 0.05) high seasonal differences. The mass fraction of SOC is always 324 larger than the contribution by POC, the exception being the observation during wintertime when 325 POC/OC ratio is higher than the SOC/OC ratio. Summing up, a distinctly different and opposite 326 seasonal trend for primary and secondary fractions of OA over central IGP has been found in this 327 study (Fig. 7).

A C C E P T E D M A N U S C R I P T
16 We sincerely thank the financial support received from CSIR, DST-PURSE and CIF Facility in 354 School of Environmental Sciences, Jawaharlal Nehru University, New Delhi to conduct this 355 research work.

A C C E P T E D M A N U S C R I P T
36 657 Fig. S1. Regression analysis of OC/EC and char/soot-EC ratios with K + in each season.