Current Issue

Black carbon (BC) is one of the primary components of airborne particulate matter (PM) (http://www.ccacoalition.org/en/files/blackcarbonsasiafinalreport52212pdf; accessed on 04.04.2016). Black carbon, which is presumed to be mainly elemental carbon (EC) (Babich et al., 2000), is generated by combustion of carbon containing materials (Hansen et al., 1988) and is mostly ultrafine in nature (Gong et al., 2016). Atmospheric BC concentration is generally low in remote areas e.g. in South Pole (Hansen et al., 1998; Andrae et al., 1995) and Manua Loa (Bodhaine, 1995) but higher concentrations (1-30 μg m^-3) are observed near its sources in urban areas and regions witnessing biomass burning (Ruellan et al., 1999).

Firecrackers are known to be a major source of particulates and their bursting causes substantial air pollution during popular public festivals, namely the ‘Festival of Lights’, known as ‘Diwali’ in India and the ‘Lantern Festival’ in China (Barman et al., 2008; Wang et al., 2007; Ravindra et al., 2003). Firecracker bursting results in emissions of smoke laden with heavy load of fine particulates (Barman et al., 2008; Ravindra et al., 2003) that are reported to carry various metals, anions and cations (Wang et al., 2007; Kulshrestha et al., 2004). On a Diwali day in Lucknow in India, 24-h average concentration of PM₁₀ was found to be higher by almost 2.5 and 5.7 times than pre-Diwali and a normal day, respectively, and on Diwali night, the 12 h mean level of PM₁₀ was 4 times higher than its respective daytime concentration (Barman et al., 2008). Thakur et al. (2010) recorded higher particulate concentration (6.44 times for suspended particulate matter (SPM), 7.16 times for PM₁₀, 5.35 times for PM_2.5) during Diwali over a typical winter day in India and particulate concentrations exceeded its respective 24 hour residential air quality standards by several times (11.6 times for SPM, 22.3 times for PM₁₀, and 34.3 times for PM_2.5). During fireworks on Lantern Day in China, PM₁₀ levels showed an elevation of up to 183% over the previous day while PM_2.5 and PM₁₀ were more than six and four times higher than a normal day (Wang et al., 2007). During a firework episode in Milan, Italy, PM₁₀ emitted by fireworks was found to constitute about 50% of total PM₁₀ mass in air (Vecchi et al., 2008). Peshin et al. (2017) reported that average surface PM₁₀ and PM_2.5 concentration levels showed an average increase of 387 and 278% in 2013; 123 and 145% in 2014; 42 and 43% in 2015 on Diwali day over a normal day at a site in India. During Chinese New Year in 2016, PM_2.5 were found to be significantly higher than Grade I National Standard of 35 μg m^-3 and contributions of secondary water soluble inorganic ions (WSII) to total WSIIs were lower in rural sites than urban ones in Central and East China (Zhang et al., 2017).

Firecracker bursting is also responsible for BC emissions (Ramachandran and Rajesh, 2007; Suresh Babu and Moorthy, 2001). As per various reports, proportion of BC or EC in ambient particulate matter varies widely (Lim et al., 2011; Fromme et al., 2008). Lim et al. (2011) have reported that average mass fractions of elemental carbon (EC) were 10.4% in PM₁, 9.8% in PM_2.5 and 6.0% in PM₁₀ at a site in Jeju Island during a 1-year study. The Indian Ocean Experiment (INDOEX) measurements have reported a 14% share of BC in fine aerosol dry mass in marine boundary layer (Lelieveld et al., 2001).

Studying escalation in ambient particulate and BC concentrations and evaluating the alterations in BC/PM₁ ratios could be crucial to understand and characterize atmospheric perturbation during firecracker bursting episodes. We report the results of a study on temporal variation of particulates (PM₁, PM_2.5 and PM₁₀) and BC in ambient air within a residential area observed before, during and after firecracker bursting during the Festival of Lights (Diwali) in Nagpur City, India. Local background data on ambient air for the same parameters was collected for three-and-a-half months leading to the Diwali day in October. The extent of short-term escalation in particulate matter and BC and changing BC/PM₁ were studied and evaluated. Possible use of BC/PM₁ ratios as a diagnostic ratio for firecracker-polluted air is discussed. Instigating scientific debate, paving for future studies and development of firecracker-smoke related diagnostic ratios for future air pollution studies underline the importance of this study.

The air quality monitoring activity was conducted in 2012 at a site located inside a residential colony in the city of Nagpur in India. Particulate matter (PM₁, PM_2.5 and PM₁₀) and BC concentrations in ambient air were studied continuously for three-and-half months to trace their temporal variation, finally leading to the Diwali night. The 4-day monitoring activity during Diwali celebration was divided into a few sampling periods designated as ‘Background’ (period before firecracker bursting started), ‘Pre-Diwali Night’ (period during sporadic firecracker bursting activities the night before Diwali), ‘Interim Background’ (period after the sporadic firecracker activities stopped up to Diwali evening), ‘Diwali Night’ (period of actual firecracker bursting during Diwali night) and ‘Post Diwali Background’ (few hours period after firecracker bursting ended on Diwali night) to track short-term perturbations in ambient particulates. Ambient particulate concentration was monitored by a portable aerosol spectrometer (Model 1.109, Grimm Aerosol Technik GMBH & Co., Germany) installed on a raised platform of approximately 4.7 meter height at a distance of approximately 10 meters from the principal firecracker bursting site. Particulate mass concentration (μg m^-3) in ambient air was measured in real time in three different size channels viz. 0.25-1.0, 0.25-2.5 and 0.25-10 μm at 1-minute time interval (time-base) during Diwali celebrations (for 4 days) while pre-Diwali background data for three-and-a-half month period was captured at three-hour intervals. A shorter sampling time-base for Diwali celebrations was selected to capture short-lived particulate generation events during firecracker bursting activities. The PM₁ mass concentration values were used for calculation of BC/PM₁.

A factor called ‘C factor’ which is a ratio of the actual mass of particulates deposited on the filter paper fixed within the instrument and mass of particulates reported by the instrument, was calculated. The C factor is used as a correction factor that is multiplied with the instrument reported particulate concentration values to get the corrected particulate concentrations in ambient air. The instrument reports particulate mass concentrations based on its calibration with monodisperse polystyrene latex particles and therefore multiplication of ‘C factor’ with instrument generated mass concentration values is essential to arrive at the actual ambient particulate mass concentrations.

Measurement of real time ambient BC mass concentration was made by an Aethalometer (Model AE-42, Magee Scientific Inc., USA), configured for a time base of 1-minute during the 4-day Diwali celebrations while the pre-Diwali background data of three-and-a-half months was collected at three-hour intervals. The shorter time-base was selected for the same reason as cited earlier in case of aerosol spectrometer. The monitoring activity during Diwali celebrations was divided into the same sampling periods discussed in one of the earlier sections. The sampling inlet of Aethalometer was fitted with a size-selective cyclone with 1-μm cut off diameter (BGI Inc.) to allow the measurement of BC in PM₁ only and to get realistic measures of BC/PM₁. The Aethalometer was installed on the same platform along with the spectrometer to simultaneously sample ambient air at the same point. The proportion of BC in PM₁ (i.e. BC/PM₁) was calculated from the mass of BC reported by the Aethalometer fitted with 1 μm size selective inlet and simultaneous mass concentration of PM₁ reported by the spectrometer, as shown below:

BC/PM₁=[(Average BC conc. in μg m^-3 reported by Aethalometer fitted with a cyclone with 1 μm cut off diameter)/(Average PM₁ conc. in μg m^-3 reported by Spectrometer)]

The observed data were subjected to descriptive as well as advanced statistical analysis. Two different regression analyses with prediction and confidence intervals were performed between ambient PM₁ and BC mass concentrations that were observed during (i) the entire monitoring duration and (ii) without mass concentrations of the values recorded during firecracker bursting period on Diwali night, to understand whether firecracker bursting led to disproportionate emissions of BC and PM₁ and disturbed their business as usual relationship. Pearson correlation analysis were also performed on the same data to evaluate if the business as usual correlation between PM₁ and BC was disturbed by the firecracker bursting or not. Mann-Whitney Rank Sum Test, which is used to compare two sample means that come from the same population and used to test whether two sample means are equal or not, was performed on ambient PM₁ and BC mass concentration values observed during firecracker bursting period on Diwali night and no-firecracker activities prior to Diwali, to examine if these groups of data under two different circumstances were statistically different from each other or not. The Mann-Whitney Rank Sum Test has been used earlier in air pollution studies for comparison of sample data sets (Röllin et al., 2004; Di Giorgio et al., 1996). The above statistical analyses and tests were performed by Sigmaplot (Version 13, SysStat) and Statistica (Version 13, Dell). Further, to examine statistically significant difference amongst various values in the data sets, Duncan’s Multiple Range Test (DMRT) was conducted by MSTAT-C statistical program (Michigan State University, USA).

Significantly, this study highlighted the alteration in BC/PM₁ ratio, apart from the perturbations in ambient particulate matter and BC concentrations during firecracker bursting activity. The following important observations were made: (i) Firecracker bursting led to much greater increase in fine particulate concentration over coarser particulates in ambient air; concentrations of PM_2.5 and PM₁₀ were clearly dominated by PM₁ (ii) Firecrackers generated substantial amount of BC also but it was much lesser in magnitude than PM₁ (iii) BC/PM₁ ratio declined substantially from the background levels, by about 3 times, during the firecracker bursting episode but returned to background values within a few hours after the termination of firecracker bursting episode.

We hypothesize that such altered BC/PM₁ ratios during firecracker bursting could indicate atmospheric pollution by firecracker smoke and hence may be attempted as diagnostic ratios for such type of air pollution. Diagnostic BC/PM₁ ratios for firecracker-polluted air should preferably be arrived at after sufficient number of air quality studies on firecracker bursting and examination of similar ratios. This study may open up a scientific debate on the possible drawbacks, uncertainties and applicability of the proposed diagnostic ratio. Similar regional studies under different activities (e.g. firecracker bursting, traffic pollution, biomass burning, volcanic activity etc.) could lead to streamlining of signature diagnostic BC/PM₁ ratios, each specific to a particular activity.