1 Introduction

The World Health Organization (WHO) acknowledged the COVID-19 pandemic as a worldwide health disaster affecting society, the economy, and the environment. The COVID-19 virus can remain viable for up to 3 days on plastic and stainless steel and 3 hours on aerosols (Stanam et al. 2020). Black Carbon (BC) is the finer constituent of atmospheric aerosol. It acts as the carrier of viruses and wide-ranging chemicals of varying toxicity to the human body besides its role in the climate forcing (WHO et al. 2012; Wang et al. 2014). It is linked to the atmospheric environment through different physicochemical and biological processes. It travels a long distance in ambient air quickly due to its smaller size (Ramanathan and Carmichael 2008). Epidemiological studies have shown the association of cardiopulmonary mortality and morbidity with exposure to BC, causing a public health problem, including premature deaths of vulnerable groups (Jansen et al. 2005). Some studies have reported the typical health effects of BC in equivalent amounts of daily passively-smoked cigarettes (Van der Zee et al. 2016; Singh and Gokhale 2021). Hence, reducing human exposure to BC would mitigate its impacts on their health.

Climate change affects global diversity (Beyer et al. 2021). Bats are the probable zoonotic cause of both SARS-CoV-1 and SAR-CoV-2 (recognized as COVID-19 pandemic), which result in severe critical respiratory syndrome from wildlife (agricultural animals) to humans (Banerjee et al. 2019; Beyer et al. 2021; Xiao et al. 2020), indicating a link between BC and Coronavirus. The number of CoVs numbers existing in a region is strongly linked with native bat species, and is influenced by climatic situations that drive species' geographical dispersals (Beyer et al. 2021). The early studies on the COVID-19 epidemic impacting airborne and water quality are reported in the recent literature (Ahmed et al. 2020; Chauhan and Singh 2020; Chowdhuri et al. 2020; Kumari and Toshniwal 2020; Mahato et al. 2020; Ranjan et al. 2020; Goel et al. 2021; Gogoi et al. 2021; Pathak et al. 2021; Saxena and Raj 2021; Sharma et al. 2020; Westhaus et al. 2021; Zhang et al. 2021), while BC and BrC with radiation effects are not studied yet. BC is the next most climate driving agent present in the atmosphere after CO2 (Ramanathan and Carmichael 2008). Climate change might cause the evolution and spread of the two SARS coronaviruses (CoVs) (Banerjee et al. 2019).

BC has significant negative influences on climate and human health (Singh and Dey 2012; Bond and Bergstrom 2006). It has a considerable warming potential with the rate of warming changes with location and elevation (Ramanathan and Carmichael 2008). The mountainous and hilly topography experience additional quick variations in temperature than surroundings at minor peaks (Goldreich et al. 1986). Almost 70% of the region in northeast India is under a hilly ecosystem where alpine vegetation, snow-capped high mountains (1–5 km a.s.l.), and rocky surface dominate the physical landscape (Gogoi et al. 2009; Pathak et al. 2021). Thus, hilly regions with considerable human settlements may substantially impact the BC and its contribution to light absorption. BC is a primary product of incomplete burning of biomass and fossil fuel. Traffic contributes to about 33–90% BC emission in areas with rich vehicular activities and congestion (Schneider et al. 2008; USEPA 2012). About 8% comes from the open burning of biomass (Reid et al. 2004). The traffic-originated BC particles are blacker than the particles from biomass combustion, hence having higher warming effects (Han et al. 2007; Ramanathan and Carmichael 2008). Diesel engines produce a relatively high amount of BC (Evans et al. 2017). It is a primary pollutant, so its concentration in the atmosphere is directly linked to anthropogenic activities.

The current COVID-19 pandemic forced partial or complete lockdown leading to the sudden closure of anthropogenic activities. The lockdowns were lifted gradually, and the actions were successively increased. The COVID-19 epidemic has triggered nearly 3 million deaths worldwide, and over 145 million confirmed cases (https://covid19.who.int/). To contain the spread of COVID-19 in India, a countrywide lockdown was forced in different periods starting March 25 to May 31, 2020. During the lockdown, the total or partial shutdown of anthropogenic activities (stone quarries, road construction, open burning of biomass, hill cutting, vehicular movement, etc.) improved the air quality concerning PM2.5, PM10 (particulate matter <2.5 and <10 μm), BC, NO2, SO2, CO, NH3, and O3 pollutants in Indian megacities, e.g., New Delhi (National capital region), Kolkata (Eastern India), Mumbai (Southern India) (Chowdhuri et al. 2020; Kumari and Toshniwal 2020; Mahato et al. 2020; Goel et al. 2021) and PM2.5 worldwide (Chauhan and Singh 2020). The lockdown also reduced India's aerosol optical depth (AOD); nearly 45% of the AOD dropped during lockdown periods (Ranjan et al. 2020). In a study by Mahato et al. (2020), PM10 and PM2.5 concentrations decreased by about 60% and 40%, respectively, throughout the initial lockdown (LD1: March 25–April 14) in the National capital region (Delhi) as compared to 2019. Chowdhuri et al. (2020) reported a significant decrease in lightning flash by about 49% in summer because of the lockdown in Kolkata, India. Pathak et al. (2021) assessed the impacts of lockdown over the upper Brahmaputra river basin (Dibrugarh) of northeast India and observed that the air quality in respect of O3, NO2, NOx, CO, SO2, PM10, PM2.5, PM1, and BC improved moderately during lockdowns as compared to the same period of 2015–2019. Gogoi et al. (2021) reported a decrease in BC by about 60% and about 30% over Indo-Gangetic plains (IGPs) and northeastern India, respectively, during LD2. It also found that the BC was high over the IGPs and northeast India compared to the rest of India. The COVID-19 amid lockdown also reduced the air impurity levels (by 30–50%) in the neighbouring region (southwestern China) (Chen et al. 2020) and other areas in the world (Chauhan and Singh 2020).

Guwahati is the largest and fastest-growing city in northeast India. It is situated in a small to the steep-sloped hilly region of the lower Brahmaputra river basin, where fossil fuel (road traffic and small industries) and biomass burning (wood burning and forest fires) are the leading sources of BC (Chakrabarty et al. 2012; Barman and Gokhale 2019; Singh and Gokhale 2021). Barman and Gokhale (2019) showed that in Guwahati, traffic and biomass contribute 85–92% and 9–15%, respectively, to the total BC. While in a study of source apportionment to increase the accuracy of contributions by using site-specific properties, the contributions were found to be 70% and 30% from traffic and biomass during the winter season in this region (Singh and Gokhale 2021). This study has been carried out to determine the impacts of lockdown in different phases on BC880 and BrC originated in this hilly region of northeast India and its effects on light absorptions. Given the significant biomass burning in the area in the winter season, the results have been analyzed to see how total or partial reduction of fossil fuel consumption has influenced the trend of BC880. A seven discrete wavelength Aethalometer utilizing dual spot technology was used for BC measurements, and source apportionment was done as per the AE33 method (Sandradewi et al. 2008). The research results in this region's unique geography and environment would contribute to the recent literature on COVID-19 associated air pollution studies.

2 Methodology

The real-time BC concentrations were measured in a rural-industrial region of North Guwahati at Air and Noise Pollution Laboratory of the Indian Institute of Technology Guwahati (IITG, 26.1878°N, 91.6916°E, elevation: 56 m asl) from 12th February to 3rd July 2020 (figure 1). A seven-channel Aethalometer (Model Number: AE-33, Magee Scientific Inc.) was set up on the top floor (nearly 12 m high), operated at two liters per minute with a waterproof PM2.5 inlet attached on the top of the inlet tube at a one-min interval. A TFE coated glass fiber filter tape roll (30 mm × 30 ft., Part No. 8050) was used. There are cement and brick kilns industries in the surrounding area along the trafficked national highways, NH27 at about 2 km to the west and NH427 at about 250 m to the south. A wireless weather station was (model: Davis Vantage Pro2, California, USA) installed in the same building to collect the meteorological parameters. It was located on the top floor to confirm that no hindrance weakens the measurements. Data were recorded at the 5-min interval, including atmospheric temperature (AT: °C), precipitation or rainfall (RF: mm), surface-reaching solar radiation (SR: W m−2), relative humidity (RH: %), wind speed (WS: m s−1), wind direction (WD: °), and barometric pressure (BP: mbar). Davis Vantage Pro2 weather station is capable of measuring the pre-tested values of meteorological parameters with maximum accuracy as: AT (0.1°C precision ± 0.5°C), RH (1%RH precision ± 3%RH), RF (214 cm2 collector with 0.2 mm bucket), WS (0.4 ms-1 steps ± 5%), WD (16-point compass), BP (internal console ± 1.0 mbar), and SR (detects SR at wavelengths of 300–1100 nm ± 5%) which is also capable of transferring the recorded data to the console unit placed up to 300 m distance (Jenkins 2014). The sunshine hour is essential in estimating the surface-reaching solar radiation. In this study, seven sunshine hours per day (09:00–16:00) were considered. To avoid interruption in the power supply, a car battery (12 V, three nos.) was attached to an AC to DC signal inverter, which powered the instruments for 15 min during the electricity interruption period.

Figure 1
figure 1

The location of IIT Guwahati (XY, solid red star) in Assam, northeast India.

Full size image

The COVID-19 lockdown stopped all anthropogenic activities that improved India's air quality (Mahato et al. 2020; Pathak et al. 2021; Zhang et al. 2021). Northeast India has several anthropogenic sources, including one coal-fired power plant (Bongaigaon Thermal Power Station), 110 large scale and 102 medium-scale (https://industries.assam.gov.in/portlets/industry-in-assam) industries (refinery, bottling plant, fertilizer, metal alloy, cement, paper, and board mill, etc.) in diverse topographical areas of this region, significantly influence the ambient air quality. The number of on-road automobiles in Assam has increased from 0.53 million in 2001 to 2.84 million in 2018 (Economic survey of Assam 2017–2018). Previous studies reported on-road traffic as a dominant source of BC in the Brahmaputra River valley region (Barman and Gokhale 2019; Singh and Gokhale 2021). The vehicles plying on the roads of Guwahati are all mixed in the various proportion of BS-II, BS-III, and BS-IV compliant. Besides, a 972-km Indian railway network across the northeastern states is diesel power-driven, which significantly contributes to the atmospheric BC.

To control the COVID-19 epidemic, a lockdown in northeast India was announced by the Indian Government. The lockdown in four phases and the limitations in several activities started from 24th March to 31st May 2020. Each stage had a different relaxation scale in several essential activities such as public transportation and essential industries' opening. Hence, the sampling was also done in six different stages: pre-lockdown (February 12–March 21), lockdown 1 (March 25–April 14), lockdown 2 (April 15–May 3), lockdown 3 (May 4–17), lockdown 4 (May 18–31), and post-lockdown (June 7–July 3 2020). The lockdown period was divided into different phases based on the measures taken and the standard operating procedures issued by the Indian Government to extend the lockdowns from time to time, depending upon the situation of the pandemic. During this period (February–July), the mean daily minimum and maximum temperatures varied from 14.80 to 21.65°C (mean: 19.39 ± 2.78°C) and 26.69 to 35.06°C (mean: 28.19 ± 2.98°C), correspondingly, with total precipitation of 101 mm. The mean RH was more in LD4 (86 ± 2%) and lower in LD1 (78 ± 7%), while the total of sunny daylight hours was almost similar. The mean WSs in LD2 (2.38 ± 0.61 m s−1) were more than twice the pre-LD (0.99 ± 0.36 m s−1) period.

The Aethalometer (AE33) model measures light reduction at seven discrete wavelengths (950, 880, 660, 590, 520, 470, 370 nm) using the in-built dual spot technology (Drinovec et al. 2015; Goel et al. 2021) and the hourly-average BC concentrations were estimated at 880 nm (BC880). The absorption angstrom exponent (AAE) was calculated from the coefficient of aerosol absorption (denoted as babs). It is a product of aerosol BC mass concentration and mass absorption cross-section (MAC) (Dumka et al. 2018). The MAC values of 18.47, 14.54, 7.77, and 7.19 m2 g−1, correspondingly, were used for 370, 470, 880, 950 nm wavelengths (Drinovec et al. 2015). The AAE370-880 values were estimated from the resulting babs (λ) for the wavelength pair of 370 and 880 nm by equations (1 and 2) (Sandradewi et al. 2008).

$${b}_{\mathrm{abs}(\uplambda )}={BC}_{\lambda } \times {MAC}_{\lambda },$$
(1)
$$AAE=-\frac{\mathrm{ln}({b}_{\mathrm{abs}(1)}/{b}_{\mathrm{abs}(2)})}{\mathrm{ln}\left({\lambda }_{1}/{\lambda }_{2}\right)},$$
(2)

where babs(1) and babs(2) are the coefficients of aerosol absorption (babs) at two distinct wavelengths.

The two-component mixing (AE33) model was used for the source apportionment of BC from two primary sources associated with biomass and fossil fuel burning, taking AAE of fossil fuel as 1.0 and biomass as 1.7 (Singh and Gokhale 2021). If AAEFF as 1.0 and AAEBB as 2.0 are applied according to the previous studies (Barman and Gokhale 2019; Tiwari et al. 2015), then biomass and FF burning involved are calculated consistently with the six distinct phases of lockdowns, as presented in figure S1. Several studies have reported the spectral band 370–880 nm wavelength for the segregation of BC sources (Dumka et al. 2018; Chen et al. 2020; Goel et al. 2021). The source apportionment of BC was done with the support of equations (3–9), with the babs estimated at 370 and 880 nm. The babs were evaluated by the BC mass concentrations at 370 and 880 nm using equation (1). The Aethalometer model (AE33) assumes that BC originated simply from biomass and fossil fuel burning sources, as equation (3).

$$\mathrm{BC}={\mathrm{BC}}_{\mathrm{FF}}+{\mathrm{BC}}_{\mathrm{BB}}.$$
(3)

The babs for the wavelength of 370 and 880 nm are separated as indicated by equations (4 and 5) (Sandradewi et al. 2008).

$${b}_{\mathrm{abs}\_370}={b}_{\mathrm{abs}\_\mathrm{ FF}\_370}+{b}_{\mathrm{abs}\_\mathrm{ BB}\_370},$$
(4)
$${b}_{\mathrm{abs}\_880}={b}_{\mathrm{abs}\_\mathrm{ FF}\_880}+{b}_{\mathrm{abs}\_\mathrm{ BB}\_880}.$$
(5)

Equations (4–7) were used to calculate FF and BB's light absorption at 370 and 880 nm wavelength. The AAEFF and AAEBB values as 1.0 and 1.7 were initially used in equations (6 and 7).

$$\frac{{b}_{\mathrm{abs}\_\mathrm{FF}\_370}}{{b}_{\mathrm{abs}\_\mathrm{FF}\_880}}={\left(\frac{370}{880}\right)}^{-{\mathrm{AAE}}_{\mathrm{FF}}},$$
(6)
$$\frac{{b}_{\mathrm{abs}\_\mathrm{BB}\_370}}{{b}_{\mathrm{abs}\_\mathrm{BB}\_880}}={\left(\frac{370}{880}\right)}^{-{\mathrm{AAE}}_{\mathrm{BB}}}.$$
(7)

The BCFF and BCBB mass concentrations (at 880 nm) were estimated by equations (8 and 9).

$${\mathrm{BC}}_{\mathrm{FF}\_880}= \frac{{b}_{\mathrm{abs}\_\mathrm{FF}\_880}}{{\mathrm{MAC}}_{(880\;\mathrm{ nm})}},$$
(8)
$${\mathrm{BC}}_{\mathrm{BB}\_880}= \frac{{b}_{\mathrm{abs}\_\mathrm{BB}\_880}}{{\mathrm{MAC}}_{(880\mathrm{\; nm})}}.$$
(9)

Considering that BrC has insignificant absorption (AAEBC = 1) at longer wavelengths (e.g., 950 and 880 nm), the residual absorption by BrC (babs_BrC) was calculated by deducting the absorption by BC (babs_BC) from the overall babs at the shorter wavelengths (660, 590, 520, 470, 370 nm) (Liu et al. 2015; Chen et al. 2020).

$${b}_{\mathrm{abs}\_\mathrm{BC}}\left(\uplambda \right)={b}_{\mathrm{abs}(880)}\times {\left(\frac{\uplambda }{880}\right)}^{-{\mathrm{AAE}}_{\mathrm{BC}}},$$
(10)
$${b}_{\mathrm{abs}\_\mathrm{BrC}}\left(\lambda \right)= {b}_{\mathrm{abs}}\left(\lambda \right)- {\mathrm{b}}_{\mathrm{abs}\_\mathrm{BC}}\left(\uplambda \right).$$
(11)

The R-statistical code (R core Team 2021) was used for data analysis and visualization with packages such as ‘openair’, ‘ggplot2’, ‘ggpubr’. All the packages can be retrieved through the Comprehensive R Archive Network (source link: https://cran.r-project.org/). The probable source provinces adding to the total BC880 concentration at the monitoring site were recognized using the Bivariate polar plots. Multivariate linear correlation analysis was used to reveal the statistically significant parameters.

3 Results

Descriptive information of total BC880, BCBB_880, BCFF_880, babs_BrC, spectral dependence from the AE-33 measurements, and local meteorological conditions during various lockdowns are listed in table 1. The time-based and daily variations in the BC880 mass concentrations are presented in figure 2(a, b), respectively. The BC880 concentration ranged from 1.57 to 15.60 μg m−3, with a mean concentration of 5.75 ± 4.24 μg m−3. The maximum concentration was in the Pre-LD and the minimum in LD4. The average BC880 concentrations during Pre-LD, LD1, LD2, LD3, LD4, and Post-LD were 9.29, 4.70, 3.41, 3.69, 2.94, and 6.56 μg m−3, individually. The BC880 concentration during the Pre-LD period was comparable to the values reported in Delhi (Goel et al. 2021). However, during LD1, LD2, LD3, LD4, and Post-LD, the BC880 concentration reduced by ~49%, ~63%, ~60%, ~68%, and ~29%, respectively. The anthropogenic activities such as public transportation and essential industries were relaxed in succeeding lockdowns (stage-wise). The stage-wise easing implemented in different lockdowns dominated the BC880 concentration from fossil fuel-associated emissions. Though, after LD3, the BC880 concentrations declined due to rain events and the abundant WS. The rainy and windy May and June decreased the BC880 and PM concentrations (table 1). BC880 concentration's evening and morning peaks also significantly dropped (62% and 75%) to low (nearly 2 μg m−3) during the LD period. After LD4, the evening and morning peaks again started recurring due to relaxation given in anthropogenic activities. However, in the Pre-LD and Post-LD periods, the strength of both evening and morning peaks was high, up to ~14 μg m−3, because of heavy vehicular activities and emissions.

Table 1 Summary of the parameters. The values indicate the daily-averaged mean value ± standard deviation. The surface-reaching solar radiation shows the seven sunshine hours averaged value.
Full size table
Figure 2
figure 2

(a) Daily-averaged hourly BC880 mass concentration – each colour represents different phases of lockdown. (b) Box plot of the diurnal variation of BC880 mass concentration during the study period.

Full size image

AAE370-880 values were 1.01–1.66 (mean: 1.25 ± 0.08) (figure 3b), which is lesser than the AAE over Delhi during lockdown (1.32 ± 0.16) (Goel et al. 2021) and during wintertime (1.29 ± 0.08) (Dumka et al. 2018), and was more than (Tiwari et al. 2015) for December 2011–March 2012 (1.09 ± 0.11). The average AAE370–880 during Pre-LD, LD1, LD2, LD3, LD4, and Post-LD were 1.19, 1.38, 1.33, 1.26, 1.21, and 1.20, respectively. The higher AAE370–880 throughout LD1, LD2, and LD3 show the notable contributions from domestic cooking and fire events, and small AAE370–880 during Pre-LD, LD4, and Post-LD indicate the most contribution from fossil fuel burning causes. The AAE470–950 ranged from 1.05 to 1.59 (mean: 1.28 ± 0.06) (figure 3a). The average AAE470–950 values during Pre-LD, LD1, LD2, LD3, LD4, and Post-LD were 1.23, 1.38, 1.33, 1.28, 1.26, and 1.23.

Figure 3
figure 3

Diurnal variation of AAE470–950 (left), and AAE370–880 (right).

Full size image

The segregation of two sources was done via 370–880 nm pair and estimated AAEFF and AAEBB of 1 and 1.7, respectively. Figure 4(a and b) shows the results. The daily mean BCFF_880 concentration was wide-ranging from 1.21 to 11.30 μg m−3, with a mean concentration of 4.13 ± 2.65 μg m−3. The average BCFF_880 mass concentrations throughout Pre-LD, LD1, LD2, LD3, LD4, Post-LD were 7.28, 2.39, 1.94, 2.37, 2.14, 5.03 μg m−3, respectively. The BCFF_880 mass concentration was maximum during Pre-LD and Post-LD. There were no constraints on industrial, vehicular, or additional anthropogenic activities (figure 4a). The BCFF_880 was reduced to 67% in LD1 and 74% in LD2. During the lockdowns, the movement of pooled vehicles was restricted except those for essential services such as doctors, police, and other essential commodities. The daily averaged BCBB_880 concentration wide-ranging from 0.37 to 4.30 μg m−3, with a mean concentration of 1.63 ± 0.82 μg m−3. The average BCBB_880 concentrations during Pre-LD, LD1, LD2, LD3, LD4, Post-LD were 2.00, 2.32, 1.47, 1.33, 0.80, and 1.53, respectively. The BCBB_880 concentration was great during LD1 and LD2 due to the forest fire events in the neighbouring region. During LD4, a sudden decrease in BCBB_880 concentrations could be attributed to rain events and an increased proportion of vehicular emission, change in WD, and high WS (table 1). Diurnal variation in BCFF_880 and BCBB_880 concentrations are shown in figure 5(a and b). High BCFF_880 mass concentration (12–18 μg m−3) was detected during the morning rush hour (9:00–10:00 LST) during the Pre-LD period. From LD1 to LD4, morning and evening peaks (3–6 μg m−3) were also witnessed, which could be due to the hospitals, Narayana Super Specialty and GNRC, located within 1 km radius.

Figure 4
figure 4

(a) Concentration of BC880, BCFF_880, and BCBB_880, and (b) contribution of the two sources to the overall BC880 throughout lockdowns.

Full size image
Figure 5
figure 5

Diurnal variation of BCBB_880 (left), and BCFF_880 (right).

Full size image

The influence of biomass and fossil fuel burning on the overall BC880 during Pre-LD, LD1, LD2, LD3, LD4, and post-LD are shown in figure 4(b). The fraction of FF was high during Pre-LD, LD4, and Post-LD, i.e., 78%, 74%, and 76%, respectively. The large fraction of FF during Pre-LD is due to no limitation on anthropogenic activities such as industrial, vehicular, railway, etc. In Post-LD, the fraction of FF was similar to Pre-LD due to the easing given during this period. In Post-LD, the Indian railway was not functioning, which subsequently directed to the growth of unattended vehicles. Moreover, being the gateway to northeast India, every vehicle passes through Guwahati city to other regions. From LD1 to Post-LD, an increasing FF fraction was observed due to easing in consecutive lockdown and unlock. The fire events might influence BB fraction, which was confirmed by the MODIS satellite retrieved fire data. In northeast India, forest fires are significant biomass burning sources, as revealed in figure S4(a–f). The fire spot data (combined product from Terra and Aqua MCD 14 ML) was collected from the Fire Information for Resources Management System (https://firms.modaps.eosdis.nasa.gov/). Though, a distinct drift was observed in the fraction of BB, which was highest during LD1 and lowest in Post-LD, the fraction of BB was large during LD1 and LD2 due to decreased vehicular emission in LD1 and increased fire events during LD2. Monthly and phase-wise changes in BC880, BCFF_880, and BCBB_880 concentration and their changes during various phases are shown in figure 6. The quantitative analysis of phase-wise anomaly was organized to comprehend BC's spatial dissimilarity, sources, and radiation effects compared to the study period mean (figure S2).

Figure 6
figure 6

Monthly and phase-wise (lockdown) changes in BC880 concentrations and contribution of FF and BB in total BC880.

Full size image

Bivariate polar concentration plots of BC880, BCFF_880, BCBB_880, and AAE370-880, show the conjugate effect of WS, WD, and potential source locations affecting the receptor (figure 7). The mean synoptic shallow wind fields were from the southeast (figure S3). The almost stagnant atmosphere with weak winds (<2 m s−1) and high pressure (765 bar) enabled subsiding and massive aerosol (BC) build-up near the surface. The maximum AAE370-880 values are linked with lower wind speeds from the southwest. The bivariate plots disclose the domination of the local emission sources as the ultimate BC880, BCFF_880, BCBB_880 concentrations were related with the low WS from the eastern direction towards the monitoring station. The east region corresponds to the heavily occupied Guwahati suburbs and neighbouring rural agricultural areas, having high traffic and biomass burning. South-easterly wind inflow directions were dominant over this region, with a mean WS of 1.68 m s−1. During the sampling period, the daily mean temperature varied from 18.88° to 29.17°C (mean: 23.31 ± 2.66). It is described in figure 8 that the BC880 mass concentrations were diminished due to rainy events. Relative humidity correlates substantially with BC880, while T, WS, RF, and SR were negatively associated (figure 9). Rainfall played a significant role in dipping BCBB_880 concentration in LD3 to post-LD as more significant fire events occurred in March and April (figure S4a–f). Figures 8 and 9 indicate the considerable impact of meteorological parameters on BC880 concentration in this region before and during the lockdown period.

Figure 7
figure 7

The bivariate polar plot indicates wind fields' effect on the measured AAE370–880, BCBB_880, BCFF_880, and BC880 concentrations.

Full size image
Figure 8
figure 8

Temporal variations in BCFF_880, BCBB_880, and BC880 mass concentration with the rainfall, temperature, and relative humidity. Rainfall denotes the daily averaged values.

Full size image
Figure 9
figure 9

Correlation plot between different parameters. Warm colors show a positive correlation, and cool colors show a negative correlation. For zero correlation, the shape becomes a circle.

Full size image

Variation in radiation effect during lockdowns is shown in figure 10. The average babs (370 nm) was 210.28 ± 88.25 Mm−1, with 122.63 ± 28.39 Mm−1, 85.60 ± 17.12 Mm−1, 88.72 ± 39.23 Mm−1, 66.57 ± 23.78 Mm−1, 184.88 ± 111.73 Mm−1 Pre-LD, at LD1, at LD2, at LD3, at LD4, and Post-LD, separately (table 1). The babs_370 was higher than in Chongqing, China (113 ± 31.4 Mm−1) during the COVID-19 epidemic (Chen et al. 2020). The average babs (at 880 nm) was 75.40 ± 31.98 Mm−1, 36.54 ± 7.41 Mm−1, 26.46 ± 4.34 Mm−1, 28.70 ± 11.65 Mm−1, 22.81 ± 7.26 Mm−1, and 65.64 ± 41.60 Mm−1 during Pre-LD, at LD1, LD2, LD3, LD4 and Post-LD, respectively (table 1). Also, lockdown actions reduced babs_880 and babs_370 considerably to 72% and 69%, respectively, till LD4. Equally, BC and BrC were the effective sunlight absorber at several wavelengths, while BC was dominant from 370 to 880 nm. The average babs_BrC (370 nm) was 30.94 ± 15.05, 35.72 ± 12.06, 22.67 ± 7.09, 20.45 ± 11.90, 12.30 ± 6.80, 28.76 ± 15.67, 26.15 ± 14.04 Mm−1, respectively, during different phases (table 1). The attribution to BrC and BC at various wavelengths between 370 and 880 nm is observed in figure 10. BrC accounted for ~15% of the overall babs at short wavelength (370 nm) earlier LD1, increased up to ~30% during consecutive lockdowns, and then decreased to 16% during Post-LD. The portion of babs_BrC_370 was considerably lower than in Chongqing (during lockdown) and Guangzhou (in winter), China (Qin et al. 2018; Chen et al. 2020).

Figure 10
figure 10

Radiation effect of BC and BrC for the 370–950 nm in the different lockdowns.

Full size image

4 Discussion

The effect of lockdowns and their related levels of BC880 has been assessed in this study to conclude the link between the concentration of BC880 and emission sources. Therefore, various phases of lockdowns with pre-lockdown and post-lockdown have been selected purposively. After the announcement of lockdown 1, anthropogenic and industrial activities were completely shut down all over India. Though there was a massive decline in energy use, PM2.5, PM10, CO, NO2 during the initial lockdown (March 16 to April 14, 2020) were decreased around 43, 31, 10, and 18%, respectively, in India from 2017 to 2020 (Sharma et al. 2020). Another study conducted in Delhi during March 3–April 14, 2020 by considering 34 monitoring stations showed an essential reduction in PM10 (60%), PM2.5 (39%), CO (30.35%) and NO2 (52.68%), and national air quality index improved by 40–50% (Mahato et al. 2020). There was a 17% increase in ozone in Delhi (Sharma et al. 2020); a possible reason might be due to lower NOx and volatile organic compounds (VOCs) in urban locations due to a sudden decrease in vehicular activities. The life-threatening lockdowns had affected the economy globally but helped improve air quality during the period, which started to deteriorate again when the lockdown was successively lifted (Chauhan and Singh 2020; Chowdhuri et al. 2020; Kumari and Toshniwal 2020; Mahato et al. 2020; Sharma et al. 2020; Goel et al. 2021). A study conducted by Barman and Gokhale (2019) in the City Center (Guwahati) and in the residential area having much less commercial and vehicular activities than at the City Center has found the BC concentration values of 17.14 and 6.42 μg m−3 during pre-monsoon (2017), respectively, which were too high as compared with the value found in this study (i.e., 3.68 μg m−3). Further, in winter (January, February), the values are even higher due to unfavourable meteorology. Therefore, the significant decrease in the values recorded in this study during the lockdown period has been the result of no anthropogenic activities and not the effect of seasonal transformation. This study showed that BC880 concentration was dropped up to 68%, and BrC accounted for 25% of the babs at 370 nm during the lockdowns.

In the study region, the hilly topography may have also promoted the accumulation of black carbon due to poor dispersion. At the same time, the reduction was not as effective as the other studies reported in India and other parts of the world (Chen et al. 2020; Sharma et al. 2020; Saxena and Raj 2021; Zhang et al. 2021). Moreover, other parts of the country experienced a significant reduction in the AOD during distinct stages of lockdown. At the same time, the changes in AOD in northeastern India were minimal (Ranjan et al. 2020). This study revealed that the IGPs (e.g., Uttarakhand) had a more significant affirmative influence of lockdown, where the AOD level diminished to ~43% throughout the lockdowns.

BC's deposition and transport significantly affected this region due to rain events and calm wind conditions. The BC880 concentrations might also have been affected by the terrain of this region and may also further have influenced the probability of fog occurrence (Hůnová et al. 2021). Moreover, a study by Goel et al. (2021) in Delhi reported a reduction of 78% in BC880 mass concentration during the lockdowns, and it was 68% in the present study. Hence, region-specific and restricted emission control policies should be applied to air quality management for sustainable development. This study focuses on black carbon aerosols, their sources, and the light absorption properties of BC and BrC. It also provides the assessment to determine the effect of an epidemic with its control during the lockdown periods in the region. While many studies carried out in India and other areas during the pandemic period were mainly focused on the reduction of air pollutants (PM2.5, PM10, SO2, NO2, O3, and BC) during a few lockdowns (Chauhan and Singh 2020; Chen et al. 2020; Mahato et al. 2020; Goel et al. 2021; Sharma et al. 2020; Saxena and Raj 2021), this study provides the knowledge on the impact on BC880 emission sources and the ambient BC880 concentrations. The sudden decrease in the BC880 concentrations during the various lockdown phases showed that the control at the origins of anthropogenic activities could improve air quality. And therefore, the imposing lockdown helped lessen coronavirus spread and improved the region's air quality significantly.

5 Conclusion

The longtime lockdown during the COVID-19 pandemic showed that direct control at the sources could reduce air pollution significantly. The BC880 mass concentration decreased by 49%, 63%, 60%, 68%, and 29% successively in the various phases of lockdowns. An improved method of source apportionment showed a significant variation in the contribution of BC880 sources; for example, the share of fossil fuel sources in the total BC880 was the lowest in LD1, which continues to increase due to successive easing in the shutdown of activities. The BCFF_880 fraction was found to be reduced from 0.78 to 0.51. During March and April, the natural forest fire events increased the fraction of biomass burning in BC880 during the first two lockdowns. At several bands, the aerosol absorption coefficient declined by 69%, and AAE was reduced by 64% during lockdowns. BrC played a significant share in light absorption, accounting for 25% of babs at 370 nm. Thus, this study revealed a substantial reduction in BCBB_880, BCFF_880, total BC880 over BRV, and significant changes in aerosol absorption coefficients due to the proportion of black and brown carbons. The results may be helpful as baseline data to study the impacts of sources on the contribution to atmospheric BC and mitigations strategies.