Natural Hazards

, Volume 71, Issue 1, pp 881–894

Impact of anthropogenic activity and cyclonic storm on black carbon during winter at a tropical urban city, Pune


  • M. P. Raju
    • Indian Institute of Tropical Meteorology
    • Indian Institute of Tropical Meteorology
  • P. S. P. Rao
    • Indian Institute of Tropical Meteorology
  • S. Tiwari
    • Indian Institute of Tropical Meteorology
  • P. C. S. Devara
    • Indian Institute of Tropical Meteorology
Original Paper

DOI: 10.1007/s11069-013-0937-y

Cite this article as:
Raju, M.P., Safai, P.D., Rao, P.S.P. et al. Nat Hazards (2014) 71: 881. doi:10.1007/s11069-013-0937-y


Black carbon (BC) aerosols are emitted into the atmosphere as a byproduct of different combustion processes and are reported to be a very strong absorber of solar radiation. In this paper, we present results on BC aerosols over Pune, a tropical urban city in south west India during Diwali festival in the month of November 2010. Daily mean BC showed about 5 % increase on Diwali day compared with preceding and succeeding period with concentrations reaching as high as about 21 μg/m3 in the morning on Diwali day, mainly due to the influence of extensive fireworks. However, the strong winds accompanied by occasional rainfall due to severe cyclonic storm “Jal” formed in the Bay of Bengal on the same day dampened this effect and reduced BC to about 2 μg/m3 within 6 h. There was only 5 % increase in mean BC concentration on Diwali day during 2010 as compared to the average increase of about 17 % during preceding 4 years on Diwali day, mainly due to the impact of weather conditions induced by Jal.


BC aerosolsDiwali festivalDiurnal variationLong-range transportSevere cyclonic storm “Jal”

1 Introduction

Black carbon (BC), the light absorbing fraction of carbonaceous aerosols, mainly originates from combustion sources, has complex climatic implications involving atmospheric heating and surface cooling (Ramanathan et al. 2001). The main sources of BC in the atmosphere are forest fires and anthropogenic combustion processes such as power generation, vehicular emission, residential heating and biomass burning for agricultural or domestic purposes. The total amount of BC released into the atmosphere from all sources is estimated to be around 6–30 × 1012 g/year (Crutzen and Andreae 1990). Freshly emitted BC is mostly hydrophobic and eventually becomes hydrophilic by oxidation or coating with sulfate and organics (Liousse et al. 1993). It has become evident that BC aerosols, emitted from combustion, can be carried far away to thousands of kilometers from source regions. Typical concentrations of BC range from 3 to 10 μg/m3 in urban areas, 0.1–1 μg/m3 in rural areas and 0.001–0.1 μg/m3 in remote regions. The highest concentrations (>10 μg/m3) occur over densely populated continental regions, and the mass fraction of BC in urban aerosols is typically a few percent (Heintzenberg and Winkler 1984). Due to its high absorbing potential, BC is reported to be the second largest contributor to global warming, after green house gas CO2 (Jacobson 2002).

Normally, BC is inert in ambient atmosphere as it is insoluble in polar and non-polar solvents and is stable in air or oxygen at temperatures up to approximately 350–400 °C. However, due to its great porosity, it has the ability to adsorb other species from the vapor phase, especially organics. BC particles are usually small enough (≤1 μm size) to be readily inhaled and further get deposited in the lungs or other airways. They may act as vehicles for the transport and localized deposition of harmful compounds to the human pulmonary system. Epidemiological studies provide sufficient evidence of the association of cardiopulmonary morbidity and mortality with BC exposure. The review of the toxicological studies suggested that BC may not be directly a major toxic component of fine particulate matter, but it may operate as a universal carrier of a wide variety of chemicals of varying toxicity to the lungs, the body’s major defense cells and possibly the systemic blood circulation (WHO 2012).

Several studies on BC aerosols over the Indian region have been reported in the recent years (Babu and Moorthy 2002; Badarinath et al. 2009; Pant et al. 2006; Tiwari et al. 2012; Ram et al. 2010; Ramachandran and Rajesh 2007; Safai et al. 2007, 2008, 2010; Dumka et al. 2010; Das et al. 2009; Raghavendra Kumar et al. 2011; Beegum et al. 2009; Raju et al. 2011). However, studies on BC emissions during intense combustion activities like burning of crackers, especially during Diwali festival are very few (Babu and Moorthy 2001; Tiwari et al. 2009). Diwali/Deepawali, also known as the “Festival of Lights” is a major Indian festival. It usually falls in the post-monsoon/winter period in the months of October/November, and is one of the most popular and widely celebrated festivals in India. During this festival, extensive display of fireworks is observed. High concentrations of anthropogenic particulate and gaseous components are injected into the atmosphere due to burning of crackers and fireworks. Fireworks mainly contain chemicals such as arsenic, sulfur, manganese, sodium oxalate, aluminum and iron dust powder, potassium perchlorate, strontium nitrate, barium nitrate and charcoal (Mclain 1980; Wang et al. 2007). In addition, burning of fireworks releases pollutants such as sulfur dioxide (SO2), carbon dioxide (CO2), carbon monoxide (CO), suspended particles (including particles below 10 μm in diameter, i.e., PM10) and several metals such as aluminum, manganese and cadmium, which are associated with serious health hazards (Hirai et al. 2000). Bach et al. (1975) reported an increase in total suspended particulate matter (TSPM) on an average of 300 % at 14 locations and by 700 % in the lung penetrating size ranges at one location due to fireworks on New Year’s Eve. Increase in particle number is witnessed in the accumulation mode range (100 nm) during the millennium fireworks in Leipzig, Germany (Wehner et al. 2000). Liu et al. (1997) reported the chemical composition and particle size of typical firework mixtures. BC increased by a factor of three (Babu and Moorthy 2001), and SO2, NO2, PM10 and TSP showed an increase by 2–10 times (Ravindra et al. 2003) during Diwali. Several interesting results are reported in literature related to Diwali fireworks over different parts of India (Ravindra et al. 2003; Agarwal et al. 2006; Barman et al. 2008; Kakoli and Gupta 2007; Singh et al. 2003). Babu and Moorthy (2001) observed a large increase in carbon particles after Diwali festival in Thiruvananthapuram. Tiwari et al. (2012) have reported the impact of firecracker burning on aerosol and gaseous pollutants in Delhi. Mass ratios of BC and aromatic organic carbon are reported to be highest on the day of Diwali and a day before Diwali event, proving the festival to be a major carbon intensive event (Agrawal et al. 2011). However, all the above-mentioned studies report excess BC concentrations during Diwali, whereas the present study reports both enhancement and reduction in BC at Pune during Diwali (except on Diwali day) due to the effect of a cyclonic storm over this region during the Diwali period of 2010.

2 Location of the experimental site

The city of Pune is situated in the leeward side of the Western Ghats. It is about 100 km to the east of Arabian Sea. BC mass concentrations were monitored during Diwali festival from November 1–10, 2010 at Pashan, a semi-urban region in Pune city. Anthropogenic activities, such as brick making in the nearby surroundings, are also one of the causes of pollution. The current population of Pune urban agglomerate is over 4 million. Pune receives annual rainfall at an average of 76 cm. Almost 80 % of the annual rainfall takes place during monsoon (June–September). During post-monsoon/winter, winds are normally from east-north-east crossing across the northeast and central India, and dispersion of particulate pollutants is comparatively less due to the low ventilation coefficients.

3 Methodology of BC sampling

BC observations during the period of 10 days (November 1–10, 2010) are studied at Pune (18°32′N, 73°51′E, 559 m AMSL). Also for the sake of comparison, past BC observations from the same location during Diwali period have been used. Continuous observations of BC were carried out using an aethalometer (Magee Sci., Inc., USA, Model AE-42). The principle of aethalometer is to measure the attenuation of light beam at 880 nm wavelength (Hansen et al. 1984). In this method, atmospheric air is pumped through an inlet with a flow rate set for 3 LPM, which impinges on a quartz microfiber strip. The light beam from a high-intensity LED lamp is then transmitted through the sample deposit on the filter strip. The measurement of the attenuation (ATN) of light beam is linearly proportional to the amount of BC deposited on filter strip. The time base for each observation was set at 5 min. The filter-based absorption technique used in aethalometer is reported to have shown good comparison with the other methods used for monitoring of BC particles (Allen et al. 1999; Babich et al. 2000). Operational principal and detailed description of the instrument can be found at

It is observed that the BC concentrations obtained using an aethalometer apparently rise after the filter tape advances, and therefore, the relationship between ATN change and BC concentration is not linear, and as the ATN increases, the measured BC concentration becomes underestimated (Weingartner et al. 2003; Arnott et al. 2005). This loading effect should be taken into account when using empirical correction algorithms. In order to account for this effect, we have used the correction algorithm presented by Virkkula et al. (2007) and also employed by Park et al. (2010) and Raju et al. (2011).

4 Results and discussions

4.1 Temporal variations in BC during Diwali festival and effects of severe cyclone “Jal”

Concentrations of BC are normally observed to be in increasing order from October to January (post-monsoon and winter) over Pune due to the prevailing meteorological conditions such as low ventilation and more atmospheric stability (Safai et al. 2007). As the Diwali festival generally falls during this period (October–November), the concentrations of BC that are already in the increasing trend are further enhanced to show very high values during the period of fireworks. Therefore, during Diwali, BC concentrations are bound to be high and so are observed in each year (Table 1). However, in 2010, even though there was wide spread and intermittent burning of fire crackers at the observational site during November 2–8, 2010 giving rise to high emissions of combustion-derived aerosols, BC concentrations during this period did not show substantial increase except from the evening of November 4, 2010 (preceding Diwali day) through the morning hours of the Diwali day (November 5, 2010). There was only 5 % increase in daily mean BC concentration on Diwali day as compared to 15, 9, 11 and 34 % increases in 2005, 2006, 2007 and 2009, respectively. In Table 1, data for 2008 are not shown as the average for Diwali day (28 October 2008) was not representative of the mean for all the 24 h (data from 0700 to 1700 hours were not obtained as the instrument did not work during that period). It can be seen from Table 1 that the daily mean BC concentration on Diwali day was 5 % more than preceding and succeeding days in 2010, whereas in earlier years, this increase was observed to be around 17 %. This is mainly attributed to the impact of weather conditions induced by a severe cyclonic system (named as “Jal”). This was the fourth severe cyclonic storm of the 2010 in North Indian Ocean cyclone season. Jal developed from a low pressure area in the South China Sea that organized into a tropical depression. The active date-wise description of development and dissipation of Jal is summarized as:
Table 1

Statistical details of BC mass concentration (μg/m3) before, during and after Diwali day and during Diwali Month at Pune during 2010 as compared with earlier years


BC before Diwali

BC on Diwali

BC after Diwali

BC in Diwali Month


4.83 ± 2.05

6.15 ± 2.88

5.85 ± 3.14

7.86 ± 2.55


2.80 ± 1.85

3.14 ± 1.62

2.94 ± 2.14

3.57 ± 1.38


4.35 ± 2.88

4.95 ± 2.56

4.47 ± 2.33

5.11 ± 2.18


1.10 ± 0.83

2.86 ± 1.56

2.70 ± 1.94

3.38 ± 1.33


7.30 ± 3.31

7.45 ± 4.30

6.81 ± 2.48

8.25 ± 2.90

On November 1, an area of low pressure associated with a monsoon trough formed in the central South China Sea. On November 2, the low pressure area became a strong tropical disturbance wave. During the next day, i.e., on November 3, 2010, the Japan Meteorological Agency (JMA) reported that the system had intensified into a tropical depression which moved westward and concentrated into a depression at 0600 UTC of November 4, 2010. The system intensified into deep depression at 0300 UTC of November 5, 2010 as shown in Fig. 1 and on the same day at 0600 UTC, the system intensified into a cyclonic storm named “Jal” over the same area. Moving continuously toward west-northwesterly or northwesterly direction, the system further intensified into a severe cyclonic storm at 0000UTC of November 6, 2010 over southwest Bay of Bengal (BoB) and adjoining southeast Bay. On November 7, 2010, when it was slightly away from the Tamil Nadu coast of India, the system weakened into a cyclonic storm as it lost its energy by giving high amount of precipitation. From November 1 to 8, 2010, wind speeds around 35 mph were observed with maximum of ~50 mph. Figure 2 shows cloud cover over India on November 5, 2010. More details on formation and development of Jal and synoptic conditions associated with it have been described by Chandrasekar and Balaji (2012) and Srinivas et al. (2013).
Fig. 1

Track of severe cyclonic system Jal as observed from Kalpana satellite
Fig. 2

Satellite picture showing cloud cover over India on November 5, 2010

To investigate the role of long-range transport of air masses from BoB, 4-day air-mass back trajectories were computed arriving at an altitude of 1,000 m (red line) and 2,000 m (blue line) AGL over Pune for the period of November 2–5, 2010 when both Diwali and Jal episodes were active, using the Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) model (Draxler and Rolph 2003) in Fig. 3. Even though the trajectory analyses are limited in absolute accuracy (Stohl et al. 1995), they are quite useful for determining the potential pathways of air masses. As shown in Fig. 3, air masses at 1 and 2 km altitudes were from southeast coming from BoB region. Thus, the effect of Jal during Diwali over Pune during first week of November is clearly seen from the air-mass trajectory analysis.
Fig. 3

Air-mass back trajectories over Pune during November 2–5, 2010

As seen from Fig. 4, daily mean BC concentration from November 1 to 11, 2010 depicts the effect of both Jal and Diwali festival over Pune. BC data for 10 November were excluded as the daily mean BC on this day was very high (11.9 μg/m3) due to some local anthropogenic burning near the sampling site, especially during early morning hours (0000–0900 hours). BC concentration showed gradual decrease from 1 to 4 November; however, on 5 November (Diwali day) in spite of existence of Jal effect, BC showed increased concentration mainly due to burning of fire crackers. However, later on once the activity of burning of fire crackers was reduced, again BC concentration showed reduction from 6 to 9 November. The concentration of BC regained its excepted value, which is generally observed in transition from post-monsoon to winter season from 11 November onward. This implies toward possible lifetime of BC particles of about 8 days over Pune, which is in comparison with that reported by Babu and Moorthy (2001) for Trivandrum during Diwali festival in October 2000. Being in the submicron size range, chemically inert and hydrophobic in nature (in its nascent form) BC has long atmospheric lifetime (Babu and Moorthy 2001; Andronache 2004). Kohler et al. 2001 reported that the main mechanism for removal of BC is wet removal. The average lifetime of BC in the lower atmosphere is generally from 7 to 10 days during dry weather conditions and 3–5 days during the wet seasons (Reddy and Venkataraman 1999).
Fig. 4

Daily variation in mean BC concentration during 1–11 November during 2010 as compared with that during 5 year period of 2005–2009 at Pune (verticalbars indicate SD. Solid lines indicate the mean for all these days, except 10 November)

4.2 Diurnal variation in BC during Diwali

It can be seen from Fig. 5a that the impact of emissions from burning of crackers started right from the night of November 4, 2010 (preceding to Diwali day) and persisted through the morning hours up to 0700 hours on 5 November (Diwali day). Generally, BC concentration is low during midnight to early morning (0000–0600 hours); however, during this period (evening hours of 4 November to morning hours of 5 November), the variation was entirely different than the normal BC variation. Concentration of BC started increasing from 1500 hours on 4 November and reached up to about 17.7 ± 2.82 μg/m3 at 0000 hours (midnight on 5 November) and then gradually decreased up to about 8.6 ± 0.38 μg/m3 at 0500 hours on 5 November. Then, again due to fresh crackers burning activity during early morning hours on 5 November, BC concentration again reached to 20.7 ± 5.10 μg/m3 at 0700 hours on 5 November. There was an impact of cyclonic system Jal during this entire period of about 9 days (November 1–9, 2010). Even on the Diwali day (5 November) that witnessed vigorous activity of fireworks around the sampling site, BC concentration dropped from >20 μg/m3 at 0700 hours to 4.8 μg/m3 at 1000 hours, i.e., within 4 h, there was about 75 % decrease in BC concentration. This decreasing trend continued further up to 1300 hours (2.3 μg/m3), which is attributed mainly due to the scavenging effect of rains (Pune received about 20 mm rainfall on November 5, 2010 during 0900–1000 hours).
Fig. 5

a Hourly variation in BC during evening hours on 4 November through morning hours on November 5, 2010 (vertical bars indicate SD and dashed lines indicate rainy period). b Hourly variation in BC before, during and after Diwali day in 2010

Thus, Diwali day during 2010 experienced impact of burning of fire crackers as well as wet deposition due to rains that resulted from severe cyclonic storm Jal over BoB. This is clearly demonstrated in Fig. 5b, which shows hourly BC variation on Diwali day (5 November) in comparison with that observed before Diwali day (1–4 November) and after Diwali day (6–9 November). BC concentrations were significantly high on Diwali day in the morning hours due to extensive burning of fire crackers but on the same day due to combined effect of high winds and rains; concentrations of BC sharply decreased after morning peak and were persistently low till the end of the day, though BC concentrations showed increase after 1500 hours and reached to another peak in the late night hours (the second peak on Diwali day was comparatively low by about 2 μg/m3 than that observed on preceding and succeeding days).

However, the important feature observed on the Diwali day in 2010 was that in spite of the impact of cyclonic storm associated with more wind speeds (mean 4.7 knots on Diwali day with values even reaching up to 11 knots during morning hours as compared to 3.3 knots on days before Diwali and 3.4 knots on days after Diwali) and heavy occurrence of rainfall (22 mm on Diwali day compared to 1 and 4 mm during preceding and succeeding days), the mean BC concentration was observed to be more (about 5 % more) on Diwali day than that during preceding and succeeding days. This can be attributed mainly to the extensive firecrackers burning activity on Diwali day, which even showed some impact on atmospheric temperature that was about 0.2 °C more on Diwali day than preceding and succeeding days (BC showed good correlation with temperature on the Diwali day, i.e., r = 0.52 at about 80 % confidence level). Figure 6 shows the hourly variation in BC along with that of wind speed (WS), relative humidity (RH) and rain fall (RF) on Diwali day in 2010. It is observed that BC showed inverse correlation with all these parameters (r = −0.55 with RH, −0.35 with WS and −0.22 with RF at about 85 % confidence level).
Fig. 6

Mean hourly variation in BC along with that of WS, RH and RF on Diwali day during 2010 at Pune

4.3 Comparative study of BC with earlier observations at Pune during Diwali

Figure 7 shows the mean hourly BC variation on Diwali day and that for October–November months (normal period in which Diwali festival falls) of 2010 compared with the same for a 4 year period of 2005–2007 and 2009 at Pune. More increase in BC mass concentration on Diwali day of 2010 than during earlier years is indicative of more fireworks burning activity during 2010. In fact, the mean BC concentration was 4.28 ± 1.79 and 4.98 ± 2.05 μg/m3 for Diwali day and for October and November months, respectively, for the 4 year period, whereas in 2010, the mean BC concentration was 7.45 ± 4.30 and 6.00 ± 2.19 μg/m3 on Diwali day and during October–November months, respectively. As seen from Fig. 7, BC concentration was more during most of the hours on the Diwali day of 2010, except during 8–13 h when high winds accompanied with rains were experienced due to Jal. Also, the hourly mean BC concentration observed as morning peak (20.7 μg/m3) on Diwali day of 2010 was more than 3 times higher than the hourly mean BC observed as morning peak (7.8 μg/m3) on Diwali day during the preceding years and was two times higher than the mean BC (10.2 μg/m3) observed as morning peak during the background morning peak BC concentration in during October–November months at Pune. Also, as seen from Fig. 4, the mean BC for period of 1–11 November during 2010 (7.31 ± 0.67 μg/m3) was more than that observed for earlier 4 years (5.06 ± 0.40 μg/m3) for the similar period. These features clearly infer to more impact of fireworks on Diwali day of 2010 in spite of the Jal effect.
Fig. 7

The mean hourly BC variation on Diwali day 2010 and that for months of October–November 2010 compared with the same for earlier four year period at Pune

Figure 8 shows the mean hourly variation in BC on Diwali days compared with the mean BC during the month in which this festival was celebrated in the respective year. Variation in BC concentration was almost similar on Diwali days compared with the mean BC during that particular month. However, the occurrence and magnitude of BC peaks were different, which is attributed to the impact of episodical source (burning of crackers) activity on the Diwali days. As seen from Table 1, mean monthly concentration of BC (the month in which Diwali festival occurred) was always more as compared with that on the Diwali day during the respective year. This clearly indicates that though the burning of fire crackers during Diwali festival is a major source of BC during that episode, its impact was superseded by the natural seasonal BC variation, which shows the gradual increase in BC from end of monsoon in September to end of winter in February. Similar observations have been reported by Babu and Moorthy (2001) at Trivandrum, a coastal station in South India; when the sudden increase in BC during Diwali festival showed gradual decrease in nocturnal BC peak; however, the daytime minimum BC did not recover to the values prior to Diwali event and even showed weak increasing trend, which the authors attributed to the normal increasing trend of background BC during October–November months.
Fig. 8

Mean hourly variation in BC on Diwali days and the same in the respective month in each year during 2005–2010 at Pune

One would expect less BC concentrations during Diwali 2010 compared to earlier year Diwali days due to Jal effect; however, the mean BC during 2010 was observed to be more in before, during and after Diwali as compared to earlier years. In fact, as seen from Table 1, mean BC during Diwali month was also more in 2010 as compared to earlier years. This indicates that not only the increase in fireworks activity in 2010 (which could not be quantified due to lack of data on fireworks burning activity for these all years) but also the increase in vehicular emissions due to increase in number of vehicles in Pune city could have contributed to the mean BC concentration during this period. Incomplete combustion in vehicular emissions is reported to be a major local source for BC, and the number of two wheelers is ever increasing in Pune (Safai et al. 2013).

5 Conclusions

Concentrations of BC aerosols were studied over Pune during the Diwali festival in 2010. Normally, BC concentrations are observed to be high during Diwali or on increasing trend during this period because of transition from post-monsoon to winter season. The daily mean BC concentration on Diwali day was 5 % more than preceding and succeeding days in 2010, whereas in earlier years, this increase was observed to be around 17 %. The dilution of BC concentration during the Diwali week in 2010 was mainly attributed to the high winds and occasional rains due to the severe cyclonic system Jal. In spite of the unfavorable meteorological conditions due to the cyclonic storm, the effect of cracker-burning was observed to overcome it, especially during the period when the fireworks activity was intensive in the night preceding Diwali day through the morning on the Diwali day. Thus, the Diwali day in 2010 witnessed effects of both source and sink mechanism on BC over Pune.


Authors are thankful to the Director, Indian Institute of Tropical Meteorology, Pune for encouragement to undertake this work. Authors are also thankful to the ISRO-GBP/ARFI, Department of Space, Government of India for providing financial support.

Copyright information

© Springer Science+Business Media Dordrecht 2013