Version of Record: https://www.sciencedirect.

com/science/article/pii/S0048969722023087
Manuscript_03a06f037811c7db9fd6b624289cf673

1 Assessment of PM2.5 using satellite lidar observations: Effect of

2 bio-mass burning emissions over India

3 Lakshmi N. B.a, Resmi E. Aa, and D. Padmalala

a
4 National Centre for Earth Science Studies, Thiruvananthapuram, India

5 Corresponding author: Lakshmi. N. B ([email protected])

7 Highlights:

8 • Wintertime PM2.5 is estimated using near-surface extinction retrievals from CALIOP.

9 • PM2.5 is modulated by agro-residue burning in the preceding November.

10 • PM2.5 is controlled by emissions rather than meteorological conditions.

11 • Influence of agro-residue burning in North India extends up to southern peninsula.

12

13

14

15

*
16 Corresponding author address: Lakshmi N. B., National Centre for Earth Science Studies

17 (NCESS), Thiruvananthapuram, India; E-mail: [email protected]

1
© 2022 published by Elsevier. This manuscript is made available under the Elsevier user license
https://www.elsevier.com/open-access/userlicense/1.0/
18 Abstract: The present study estimates the particulate matter with aerodynamic diameters less

19 than 2.5µm (PM2.5) over the Indian sub-continent using near-surface retrieval of aerosol

20 extinction coefficient (2007 - 2021) of Cloud-Aerosol Lidar with Orthogonal Polarization

21 (CALIOP) onboard Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation

22 (CALIPSO) satellite. Climatology of wintertime PM2.5 during the last 15 years shows the

23 highest concentration over the middle Indo-Gangetic Plain (IGP) and northwest India with a

24 3 to 4 fold increase in magnitude compared to peninsular India. Surface-level PM2.5 mass

25 concentration during winter (December to February) shows statistically significant positive

26 trends over the Indian subcontinent. It increases at a rate of ~3% over the IGP and arid

27 regions of northwest India, and ~4% over peninsular India during the last fifteen years (2006

28 - 2020). Interannual variability of average near-surface PM2.5 concentration over the Indian

29 sub-continent during the fog occurring season (December to February) shows a statistically

30 significant correlation with the post-harvest agro-residue burning over the western IGP

31 (Punjab and Haryana) during November. The wintertime near-surface PM2.5 concentration

32 shows a higher correlation with anthropogenic agro-residue burning activity compared to

33 meteorological parameters. The influence of agro-residue burning during November over

34 northern India extends up to peninsular India and might contribute to continental pollution

35 outflow and associated aerosol plumes persisting over the Northern Indian Ocean during the

36 winter season. Sustainable energy recovery solutions to the agro-residue burning need to be

37 implemented to effectively reduce the far-reaching implications of the post-monsoon burning

38 activity over the western IGP.

39 Keywords – PM2.5, CALIPSO, Lidar, Aerosol, India, Air quality, Agro-residue burning

40

41

2
42 1. Introduction

43 Monitoring of atmospheric particulate matter (PM) assumes great importance in the

44 context of air quality and pollution-induced health hazards. Regional aerosol loading over

45 India is alarmingly increasing (Moorthy et al., 2013), poising concern due to its potential

46 impacts on changing regional climate and human health. Studies have revealed the appalling

47 fact of the relationship between ambient PM2.5 (particulate matter with aerodynamic

48 diameters less than 2.5µm) exposure with premature mortality burden, reduced life

49 expectancy (Cohen et al., 2017; David et al., 2019), and risks of developing

50 cardiovascular/respiratory diseases (Dominici et al., 2006). Premature mortalities due to

51 PM2.5 exposure exceed one million per year in India (Cohen et al., 2017), with a significant

52 contribution (~46%) from the Indo-Gangetic Plains (IGP), which is one of the hotspots of

53 anthropogenically formed aerosol pollution over the globe (David et al., 2019) and is rapidly

54 increasing at a rate of 2.7% per annum with remarkably high occurrence over Northen India

55 (Jia et al., 2021). In addition, studies carried out during the Covid-19 pandemic have

56 established a positive link between PM2.5 emission and the number of positive cases in India

57 demonstrating the vulnerability of people residing in highly polluted regions to coronavirus

58 disease (Sahu et al., 2021).

59 Furthermore, wintertime PM2.5 pollution over India has been associated with adverse

60 weather conditions such as extreme haze and fog events over the IGP (Li et al., 2021).

61 Stagnant meteorological conditions during the winter season caused by weak winds and

62 shallow boundary layer due to low surface temperature reduce ventilation coefficient and

63 confine aerosols nearer to the surface across the IGP (Ojha et al., 2020). Under favorable

64 conditions, they hygroscopically grow to large particles, leading to severe haze-fog events

65 and lead to a drastic reduction in visibility disrupting road safety and air traffic. Aerosols

3
66 affect fog dynamics through their direct effect on radiation and by changing the microphysics

67 of fog formation through their hygroscopic properties. Studies have shown that aerosols play

68 a key role in the development of well-mixed, optically thick radiation fog that is difficult to

69 disrupt and stays for days causing severe visibility degradation (Boutle et al., 2018).

70 Modeling simulations revealed that aerosol-induced cooling at the surface and subsequent

71 feedbacks inhibit the growth of the atmospheric boundary layer, further reducing air quality

72 and visibility near the surface (Bharali et al., 2019).

73 Understanding the extent and trends of air pollution is essential for generating emission

74 inventories with improved accuracy for reliable forecasting and developing effective climate

75 mitigation policies to combat air pollution and related hazards. For exposure assessment

76 studies and evaluating the vulnerability of severe weather events, PM2.5 estimates at the

77 surface are very crucial. PM2.5 at the surface is generally estimated using in situ

78 measurements as well as remote sensing techniques. Though in situ methods provide real-

79 time observations with better accuracy, their observation capacity is limited over space and

80 time. Satellite-based remote sensing provides large-scale, long-term observations over

81 regional and global scales at the expense of accuracy and details, making it the best tool for

82 understanding the comprehensive representation and long-term trends in particulate matter.

83 Over the Indian region, there have been attempts made to derive spatially resolved high-

84 resolution PM concentrations from passive observations of satellite-based aerosol optical

85 depth (AOD) (Dey et al., 2012; Mhawish et al., 2020). Dey et al. (2012) used AOD

86 observations of Multiangle Imaging SpectroRadiometer (MISR) to estimate surface PM2.5

87 over the Indian sub-continent and Mhawish et al., (2020) coupled satellite observations of

88 AOD with meteorology and land-use variables in the framework of random forest model to

89 derive surface PM2.5 concentration. However, retrieving PM2.5 at the surface from columnar

90 AOD involves assumptions on aerosol vertical distribution, aerosol composition, and

4
 91 atmospheric humidity. The Indian region is characterized by the presence of elevated layers

92 of aerosols with high loading at free-tropospheric altitudes (Lakshmi et al., 2021). One of the

93 major sources of uncertainty in converting AOD to PM2.5 is the lack of adequate information

94 on aerosol vertical distribution. Generally, boundary layer height (BLH) is coupled with

95 AOD to retrieve PM2.5. However, the lack of measurements of BLH further hinders the

96 accuracy of the estimated PM2.5 at the surface. The Cloud-Aerosol Lidar with Orthogonal

97 Polarization (CALIOP) onboard the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite

98 Observation (CALIPSO), provides an excellent opportunity to derive PM2.5 concentration

99 from the direct measurements of aerosol backscatter near the surface, and long-term

100 observations of aerosol vertical distribution with better sensitivity over the land. Toth et al.,

101 (2019) demonstrated the capability of using direct extinction retrievals of CALIOP near to

102 the surface to retrieve PM2.5 mass concentration over the contiguous United States based on a

103 bulk mass scattering scheme. The present study estimates regional distribution and long-term

104 trend of PM2.5 over India based on direct near-surface observations of CALIOP coupled with

105 spatially resolved bulk properties estimated from reanalysis and observational datasets.

106 Because of the highly mixed aerosol system over south Asia, CALIOP's extinction retrievals

107 are highly uncertain as the lidar ratio selection algorithm does not estimate the lidar ratio

108 correctly (Lakshmi et al., 2017). The present study attempts to reduce this uncertainty by

109 normalizing CALIOP profiles of aerosol extinction coefficient with columnar AOD of

110 Moderate Resolution Imaging Spectroradiometer (MODIS) and attempts to quantify PM2.5

111 over India using vertically resolved, direct, near-surface extinction retrievals of AOD-

112 normalised CALIOP extinction coefficient. Based on the multiyear observations of CALIOP

113 and MODIS, we have examined the climatological mean state and long-term trends in PM2.5

114 concentration and investigate the different controlling factors that contribute to PM2.5 loading

115 during winter over the Indian region.

5
116 2. Observational datasets

117 Table 1 shows the details of satellite observations, ground-based, and reanalysis datasets used

118 in the study.

119 Table 1 - Details of the datasets used in the study.

Dataset Parameter Product and Spatial Period Reference

version resolution

CALIPSO Aerosol Level 2 Version 5 km / 60 m 2007 - 2021 Winker et

extinction 4.2 aerosol profile al., (2009)
coefficient at product (Horizontal/
532 nm
Vertical)

MODIS AOD at Collection 6.1 1⁰ x 1⁰ 2007 - 2021 Platnick et

550nm Level 3 Monthly al., (2017)
product
(MYD08_M3.6.1)

Fire count Collection 6.1 1 km x1 km 2006 - 2020 Giglio et

(MCD14DL) al., (2016)

MERRA-2 Mass Monthly 0.5 ° x 2007 - 2020 Gelaro et

concentration assimilation 0.625 ° al., (2017)
of aerosol product
species (M2TMNXAER)

Wind fields Monthly mean 0.5 ° x 2006 - 2021 MERRA

instantaneous 0.625 ° (2019)
product
(M2IMNPANA
v5.12.4)

AIRS Relative Level 3 1°x1° 2006 - 2021 AIRS

(Aqua) Humidity AIRS3STM (2019)
version 7
Surface 1°x1°
Temperature

6
 Central PM2.5 and Point 2017 - 2020 Gargava
Pollution PM10 observations and
Control Shukla.,
Board (2019)
(CPCB)

120 2.1 CALIOP

121 CALIOP is a two-wavelength polarization-sensitive lidar onboard the CALIPSO

122 satellite. It transmits laser (Nd: YAG) pulses at 532 nm and 1064 nm and measures range-

123 resolved backscattered radiation from the atmosphere, providing high-resolution vertical

124 profiles of aerosols and clouds over the globe, with a repeat cycle of 16 days. The present

125 study uses level 2 version 4.2 aerosol profile product during winter (December – February) of

126 2007 to 2021 (The year corresponds to the January and February months of the season).To

127 ensure the quality of the CALIOP data, different screening and quality checks are performed.

128 Cloudy pixels are screened using Cloud-Aerosol Discrimination (CAD) score. The CAD

129 values ranging from -70 to -100 are used to extract the cloud-free pixels with high confidence

130 (Liu et al., 2009). Range bins tagged as “clear air” (with no aerosols) are given an extinction

131 of 0.0 km−1. If the aerosol concentration is below the limit of the CALIOP detection limit, it

132 might be classified as a clear air sample, hence the average profiles represent the lower bound

133 of the aerosol extinction coefficient (Winker et al., 2013). However, the present study

134 normalizes average CALIOP profiles with average AOD of MODIS for each year which will

135 eventually reduce the uncertainties associated with the minimum detection as well as lidar

136 ratio. Extinction quality control flag (QC) of values 0 and 1 are used to obtain extinction

137 retrievals with high confidence. The QC flag of 0 gives the unconstrained retrievals where the

138 initial lidar ratio remained unchanged in iterations and 1 gives the constrained retrievals using

139 transmittance measurements. Based on the uncertainty estimates in the retrieved parameters,

140 the range bins with an extinction uncertainty flag of value 99.9 km-1 are excluded in the

7
141 present study. Details of the uncertainty estimation and its propagation in various aerosol

142 parameters are described in Young et al., (2013). CALIOP products are given as altitude with

143 reference to above mean sea level (AMSL). In the present analysis, each profile is corrected

144 for surface elevation and modified about altitude above ground level (AGL), using the

145 surface elevation statistics provided in the level 2 CALIOP product. The mean surface

146 elevation over a horizontal distance of 5 km spanned by each profile is used to convert the

147 profile altitude to AGL. The profiles are gridded vertically to a resolution of 100 m for the

148 convenience of approximating the near-surface aerosol layer from 100 to 500 m. Near-to-

149 surface CALIOP signals show negative values (surface anomaly) due to enhanced signal

150 attenuation (Winker et al., 2013). The present analysis excludes negative values and applies

151 valid range limits to the products (Aerosol extinction coefficient 0 - 1.25 km-1, backscattering

152 coefficient 0 - 0.05 sr-1 km-1) (Vaughan et al., 2018). More details on the screening

153 procedures are described in Lakshmi et al., (2021). CALIOP passes over the Indian region at

154 local time ~ 14 hrs during the day and ~21.30 hrs during the night. In the present study, both

155 night-time and daytime observations are used for retrieving the average PM2.5 mass

156 concentration during winter over the Indian region.

157 2.2 MODIS

158 MODIS onboard Aqua flew in formation with CALIPSO (~1 to 2 minutes ahead) as a

159 part of the A-Train and provided nearly collocated measurements from 2006 to 2018.

160 However, in September 2018, the CALIOP orbit was lowered from 705 km to 688 km and

161 joined Cloudsat forming the C-Train constellation. Although the ground track of the C-Train

162 differs from that of the A-Train, about every 20 days it crosses the ground track of the A-

163 Train, allowing simultaneous observations of A-Train and C-Train instruments. MODIS

164 makes observations with a ground swath of 2300 km, which provides near-global coverage

8
165 every day with 10 km resolution at nadir. Experimental datasets used for the study include

166 Collection 6.1 level 3 monthly product of AOD

167 (‘Deep_Blue_Aerosol_Optical_Depth_550_Land_Mean’) computed based on the Deep Blue

168 algorithms over land at 0.55 μm at a grid resolution of 1˚ x1˚ and fire count (MCD14DL)

169 observations. MODIS provides daily observations of aerosol properties over the globe in a

170 wide spectral range of 0.47 to 2.12 μm by measuring reflectance at the top of the atmosphere

171 that contains solar reflectance and terrestrial thermal emission. The deep blue band at 412 nm

172 provides well-differentiated signals from aerosols and surface reflection, hence providing

173 better estimates of aerosol properties over land. MODIS fire products are retrieved from

174 radiance measured at 4 and 11 μm at a spatial resolution of 1 km. MODIS (Aqua _terra)

175 Collection 6.1 standard active fire product (MCD14DL) has been used to analyze the fire

176 activity over the study region. Fire detection is carried out by identifying pixels with emission

177 intensity above a background thermal emission obtained from the surrounding pixels,

178 normalizing the variability of the surface temperature and reflection by sunlight. MODIS

179 detects fire activity overland by measuring brightness temperature at specific bands and have

180 been providing fire data at a spatial resolution of ~1 km over the globe for more than a

181 decade. To ensure the quality of the detected fire pixels, a nominal confidence threshold of

182 80% is used to screen the fire data.

183 2.3 MERRA

184 Modern-Era Retrospective Analysis for Research and Applications (MERRA),

185 version 2 includes assimilation of aerosol parameters from various ground-based and satellite

186 observations and model simulations of Goddard Earth Observing System-5 coupled with

187 Goddard Chemistry, Aerosol, Radiation, and Transport model (GOCART) aerosol-radiation

188 parameterization. Observation datasets include MODIS: Aqua and Terra Neural Net

9
189 Retrievals (NNR) of 550 nm AOD, Aerosol Robotic Network (AERONET), MISR (surface

190 albedo > 0.15), Advanced Very High-Resolution Radiometer (AVHRR). Spatial resolution is

191 0.5° ⨉ 0.625° with 72 layers up to 0.01 hPa. Simulated aerosol species are dust (DU), sea salt

192 (SS), sulfates (SO4), organic carbon, and black carbon. Validation and caveats of MERRA-2

193 aerosol assimilation products have been detailed in Randles et al., (2017) and Buchard et al.,

194 (2017). The present study uses MERRA-2, Aerosol Diagnostics version 5.12.4, time-

195 averaged, monthly mean, single-level, assimilation product (M2TMNXAER -

196 tavgM_2d_aer_Nx: 2d) to obtain species resolved aerosol composition near to surface.

197 2.4 AIRS- Aqua

198 Relative humidity and surface temperature observations near the surface are obtained

199 from the Atmospheric Infrared Sounder (AIRS) onboard the Aqua satellite. AIRS is a

200 hyperspectral infrared spectrometer providing global distribution of meteorological

201 parameters by measuring outgoing radiation at visible/near IR bands ranging from 0.4 to 1.0

202 µm and IR bands from 3.7 to 15.4 µm at a spatial resolution of 2.3 km and 13.5 km

203 respectively at nadir. Level 3 monthly Standard Physical Retrieval (AIRS-only) 1⁰ x 1⁰

204 version 7 (AIRS3STM) products are used in the present study. The monthly mean values are

205 arithmetic means of the daily observations, weighted by the number of samples per grid cell.

206 2.5 CPCB

207 As a part of the National Ambient Air Quality Monitoring Programme (NAMP) the

208 Central Pollution Control Board (CPCB) maintains a nationwide network to monitor air

209 quality regularly with 804 monitoring stations as of 31st December 2019, measuring sulfur

210 dioxide, nitrogen dioxide, and PM10. PM2.5 is observed at 317 stations covering 147 cities in

211 20 states and four Union Territories (311 over Residential/industrial/rural/other areas and six

212 over Ecologically sensitive areas) (Gargava and Shukla, 2019). The PM measurements are
10
213 carried out based on different methods - the gravimetric method, using Tapered Element

214 Oscillating Microbalance (TEOM), and the Beta-attenuation method (BAM) (Kamyotra and

215 Saha, 2013). The present study uses CPCB observations to estimate PM2.5 to PM10 ratio for

216 converting satellite-derived PM to PM2.5.

217 3. Methodology and validation

218 The following section discusses the methodology for retrieving and validating PM2.5 mass

219 concentration.

220 3.1 Retrieval of PM2.5 Mass concentration

221 The present study follows Toth et al., (2019) to retrieve surface-level particulate mass

222 concentration (PM) over the Indian region using, near-surface retrievals of aerosol extinction

223 coefficient (β) in km-1, mass scattering efficiency (ascat), mass absorption efficiency (aabs) in

224 m2g-1 estimated using Mie calculations based on aerosol composition, and hygroscopic

225 growth factor (frh) estimated from relative humidity observations, to retrieve total particulate

226 mass concentration, as given by the following equation

×
227 = ×
(1)

228 frh is the ratio of aerosol extinction coefficient in the ambient humidity to the corresponding

229 aerosol extinction coefficient in dry conditions and represents the humidity effect on aerosol

230 light scattering due to their hygroscopic growth. Mass concentration of PM(μgm-3) estimated

231 using equation 1, spans the entire size range of particles. It is converted into PM2.5 (as in

232 equation 2) using the ratio of PM2.5 to PM10 (ϕ) measured over point locations over the Indian

233 region.

234 . = PM × (2)

11
235 3.1.1 CALIOP – Aerosol extinction coefficient near to the surface

236 The analysis is carried out for the winter season (December to February) when

237 aerosols are confined within the PBL due to low convective conditions leading to a decline in

238 air quality and severe fog formation over the Indian region. Although CALIOP provides

239 accurate observations of aerosol vertical distribution, the magnitude of CALIOP retrieved

240 aerosol extinction coefficient has inherent uncertainties associated with the assumption of

241 lidar ratio (extinction to backscatter ratio) which is strongly associated with the composition

242 of aerosols present in the scattering volume (Kim et al., 2018; Omar et al., 2009; Winker et

243 al., 2009). The aerosol system over South Asia is highly heterogeneous due to the mixing of

244 transported dust aerosols from the arid regions in the west with local emissions of

245 anthropogenic aerosols, which makes the selection of lidar ratio challenging over this region

246 (Lakshmi et al., 2021; Nair et al., 2017; Satheesh and Srinivasan, 2002). The lidar ratio

247 selected by the CALIOP algorithm over South Asia significantly deviates from observations,

248 especially during the winter season (Lakshmi et al., 2017). Hence the present study uses

249 AOD observations of MODIS to normalize the vertical profiles of aerosol extinction

250 coefficient retrieved by the CALIOP to obtain more accurate estimates of aerosol extinction

251 coefficient near to surface over the Indian sub-continent. Level 2 CALIOP cloud-free aerosol

252 profiles (5 km resolution along the track) are gridded to 1o x 1o regional bins and averaged

253 over the winter season of each year from 2007 to 2021 (The year corresponds to the January

254 and February months of the season). Initially mean CALIOP aerosol extinction coefficient

255 profiles are normalized to unit AOD to obtain normalized fractional profiles year-wise and

256 then multiplied with the corresponding mean winter AOD of MODIS (Level 3 MODIS AOD

257 with 1o x 1o grid resolution). The AOD over India has been increasing and shows significant

258 interannual variability during recent decades (Babu et al., 2013; Kuttippurath and Raj, 2021).

259 We, therefore, have normalized the mean profile of the CALIOP aerosol extinction

12
260 coefficient using the MODIS average aerosol optical depth in the same year. Thus obtained

261 profiles of AOD-normalized aerosol extinction coefficient are averaged over the study period

262 from 2007 to 2021 to obtain a climatological mean state of aerosol extinction coefficient over

263 India. Toth et al. (2019) have analyzed the sensitivity of surface layer height to CALIOP

264 derived PM2.5 using ground-based observations over the Contiguous United States. The study

265 has observed that the mean bias was highest for the layer 0 to 100 m which is attributed to

266 the overlying attenuation of the signal at the surface (Toth et al., 2019). Hence the lower

267 bound of the near-surface layer is chosen to be 100 m in the present study. Over the Indian

268 sub-continent, the typical mixed layer height during winter is much lower than 1000 m. Over

269 Indo-Gangetic Plain, wintertime mixed layer height lies within the range of 100 to 900 meters

270 (Nair et al., 2007) and over the coastal regions, the lowest possible height of the mixed layer

271 is shown to be around 500 meters (Rajeev et al., 2016). The upper limit of 500m is chosen for

272 the near-surface aerosol layer assuming vertical homogeneity within the mixed layer. A

273 sensitivity analysis of the surface aerosol layer height has been carried out by comparing

274 CALIOP-derived PM2.5 concentration with in situ observations of PM2.5 over CPCB stations.

275 Spatially (1⁰ x 1⁰ grid) and temporally (same day) collocated in situ measurements and

276 satellite-derived PM2.5 are averaged for each station and compared with each other during the

277 winter of 2017 to 2020. Figs. 1a and 1b show the root mean square error and R2 values for

278 different combinations of near-surface aerosol layer height with spatially and temporally

279 collocated average PM2.5 measured over CPCB ground stations. An R2 value of 0.44 is

280 obtained for the near-surface layer of 100 to 500 m which is higher than other combinations

281 when the upper limit of near-surface aerosol layer height varied from 200 to 1000 m. In terms

282 of root mean square error, the lowest error is observed for 100 – 200 m layer height.

283 However, with a very low R2 value, it is excluded from being used to represent the surface

284 layer in the present study. Based on the above observations, we have chosen an altitude range

13
285 of 100 to 500 to best approximate the near-surface layer and Fig. 1c shows the climatological

286 mean state (2007 - 2021) of wintertime aerosol extinction within 100 to 500 m during the

287 winter season of 2007 to 2021.

288

289 Fig. 1. a) and b) represent the root mean square error and the R2 value of CALIOP-derived
290 PM2.5 for varying near-surface aerosol layer height and in situ measurements over CPCB
291 stations during the winter season of 2017 to 2020 c) spatial distribution of average aerosol
292 extinction coefficient within 100m – 500m AGL at 1⁰ x 1⁰ grid resolution during the winter
293 season of 2007 to 2021.

294 The spatial distribution of the near-surface aerosol extinction coefficient shows a highly

295 heterogeneous pattern over the Indian region. Being one of the global hotspots of aerosols,

14
296 IGP shows high values of near-surface aerosol extinction coefficient. Central and eastern

297 parts of IGP show higher loading compared to western IGP. Studies have identified an

298 “aerosol pool” region dominated by fine particles over central and eastern IGP (Kumar et al.,

299 2018) primarily caused by emissions from vehicular and industrial sources (Singh et al.,

300 2017) and sustained by stagnant meteorological conditions and low lying topography (Vinoj

301 and Pandey, 2022).

302 3.1.2 Mass extinction efficiency – ascat and aabs

303 The values of ascat and aabs are estimated using Mie-calculations by the OPAC

304 software (Hess et al., 1998; Koepke et al., 2015), based on near-surface aerosol composition

305 of the MERRA-2 reanalysis dataset. The OPAC estimates the optical properties of an external

306 mixture of aerosols based on the number density of each aerosol component defined by the

307 user. For the calculation of number density (Ni) of different aerosol components represented

308 by I, MERRA-2 average surface level mass concentration (Mi) during the winter of 2007 to

309 2021 is regridded to 1⁰ x 1⁰ spatial resolution and is divided by the mass normalized to one
∗
310 particle for each species as shown in equation 3.

311 = ∗ (3)

∗
312 Values of in μgm-3 for different species are adapted from Hess et al. (1998). The aerosol

313 composition of MERRA consists of the following components: dust, sea salt, organic carbon,

314 black carbon, and sulfate. It is used to represent a 5-component aerosol model in the OPAC

315 software that consists of soot (black carbon), mineral dust, sulfate, water-soluble particles,

316 and water-insoluble particles. Mi of black carbon, dust, and sulfate of MERRA-2 are directly

317 converted to corresponding number densities using equation 3. The Mi of water-soluble and

318 insoluble parts are assumed to be 40% and 60% of the total organic carbon of MERRA-2

15
319 based on the reported ratio of water-soluble to insoluble organic carbon over point locations

320 in India (Hegde et al., 2020; Miyazaki et al., 2009; Ram et al., 2012) and subsequently

321 converted into number densities. Equation 3 is applied to each grid point to obtain the spatial

322 map of the number densities of each of the five species which is then used in OPAC for dry

323 conditions (at a relative humidity of 0%) estimate ascat and aabs for that pixel. The spatial

324 distribution of ascat and aabs over the Indian region for the winter season (December –

325 February) is shown in Fig. 2.

326

327 Fig. 2. Spatial distribution of mean a) dry mass scattering efficiency (ascat) and b) dry mass
328 absorption efficiency (aabs) for the winter season of 2007 – 2021.

16
329 ascat shows high values around the Eastern part of the sub-continent which can be attributed to

330 the dominance of highly scattering sulfate aerosols as shown in supplementary Fig. S1. Aabs is

331 mainly determined by the relative dominance of black-carbon aerosols in the atmosphere and

332 the spatial distribution of aabs follows a similar pattern to black carbon percentage

333 contribution as shown in the supplementary Fig. S1. Arid regions in northwest India show

334 low values of both ascat as well as aabs due to the dominance of dust aerosols which have

335 lower mass absorption/scattering efficiency relative to other aerosol species (Hand and Malm,

336 2007).

337 3.1.3 Humidity correction: Hygroscopic growth factor

338 Based on the observations of hygroscopic properties of boundary layer aerosols

339 during the Indian Ocean Experiment (INDOEX), Maßling et al., (2003) parametrized the

340 hygroscopic growth of aerosols with relative humidity (RH) using the following two-

341 parameter fit equation

)
342 !"# = $ 1 − '( (4)

343 Where frh is the hygroscopic growth factor. The fit parameters a and γ for polluted air masses

344 with an initial dry size of 150 nm are 0.9756 and – 0.2831, respectively (Maßling et al.,

345 2003). RH is represented in fractions ranging from 0 to 1. RH at 1000 hPa, retrieved from

346 AIRS observations, is used to estimate frh over the Indian region and is shown in Fig. 3 for

347 night and day. Relative humidity shows significant diurnal variation over the tropics. Hence

348 present study carries out discrete humidity correction for night-time and day-time

349 observations.

17
350

351 Fig. 3. Spatial distribution of mean hygroscopic growth factor (frh) of aerosols during a)
352 night-time and b) day-time for the winter season (December to February) from 2007 to 2021.

353 The day-time frh shows a uniform pattern over the Indian landmass with values

354 ranging from 1.1 over inland to 1.4 over the coastal regions. However, night-time frh shows a

355 heterogeneous pattern with high values (1.7) over the southern part of the peninsula to low

356 values (1.2) over the arid regions of northwest India. High values of frh during the night can

357 be attributed to low temperature and high relative humidity that favors condensation and

358 growth of particles (Mandariya et al., 2020). Except for a few instances, observational

359 estimates of hygroscopic growth factors of aerosols are very scarce in Indian regions. A study

360 carried out at Ahmedabad, a semi-arid location in western India showed mean hygroscopic

361 growth factors of 1.08 ± 2, 1.09 ± 2, and 1.54 ± 1 at 40%, 65%, and 90% relative humidity

362 respectively during late winter (Sarangi et al., 2019). A study carried out over a polluted

363 urban location in central IGP during the winter season showed frh values of 1.27 ± 0.06 and

364 1.34 ± 0.07 for 100 and 150 nm particles, respectively (Mandariya et al., 2020).

365 3.1.4 PM2.5 to PM10 factor

18
366 Air quality data of CPCB has been used to estimate the spatial distribution of the ratio

367 of PM2.5 to PM10. Daily PM2.5/PM10 values have been evaluated and averaged over the winter

368 (December – February) of 2017 to 2019 and are shown in Fig. 4.

369

370 Fig. 4. The ratio of PM2.5 to PM10 over the Indian subcontinent during the winter (December
371 to February) of 2017 – 2019. Inset shows the frequency of occurrence of PM2.5 to PM10
372 factor.

373 The ratio of PM2.5/PM10 values mostly ranges from 0.3 to 0.7 over the Indian sub-

374 continent. The frequency distribution of PM2.5/PM10 peaks at 0.52 and the present study uses

375 this value for converting satellite-derived PM mass concentration to PM2.5 during winter over

376 the Indian sub-continent.

377 A flowchart of the methodology used in the present study for the retrieval of near-surface

378 PM2.5 mass concentration is given in Fig. 5.

19
379

380 Fig. 5. Flow chart of the methodology adopted for the present study

381 3.2 Validation

382 Fig. 6 shows the intercomparison of in situ measured PM2.5 concentration over stations under

383 CPCB (PM2.5_CPCB) with collocated CALIOP derived PM2.5 concentration (PM2.5_CALIOP) over

384 the stations during the winter of 2018 to 2020.

385

386 Fig. 6. a) Scatter plot of in situ measured PM2.5 over stations under CPCB with average

387 CALIOP-derived PM2.5 within 1⁰ x 1⁰ grid around the stations. Each point represents the

388 seasonal mean PM2.5 for each station location during the winter from 2018 to 2020. The color

20
389 of the symbol represents the year of the measurements and the inset shows the R2 values of

390 linear regression analysis for each year. The red line represents the linear regression fit and)

391 Locations of stations under CPCB used in the study.

392 Daily mean CALIOP profiles of aerosol extinction coefficient (5 km resolution along the

393 track) over 1⁰ x 1⁰ grid around CPCB stations are normalized using spatially (within 1⁰ x 1⁰

394 grid) and temporally (same day) collocated MODIS AOD. The AOD-normalized daily mean

395 CALIOP aerosol extinction coefficient profiles over a 1⁰ x 1⁰ grid around CPCB stations are

396 used to estimate the near-surface (100 – 500 m) aerosol extinction coefficient. Spatially

397 collocated (1⁰ x 1⁰ grid) mean frh (refer to section 3.1.3), ascat, and aabs (refer to section 3.1.2)

398 averaged for the winter of 2018 to 2020 are used in equation 1 to estimate total PM and is

399 converted into PM2.5 using a PM2.5/PM10 ratio of 0.52 (refer to section 3.1.4). Spatially (1⁰ x

400 1⁰ grid) and temporally (same day) collocated in situ measurements and satellite-derived

401 PM2.5 are averaged for each station and compared with each other using a scatter diagram as

402 shown in Fig. 6a. Validation analysis is done separately for each year from 2017 to 2020 and

403 is represented by color variation in Fig. 6a. Average collocated PM2.5 concentrations over a

404 total of 179 ground stations under CPCB are used for the validation analysis (Fig. 6a) and

405 their locations are shown in Fig. 6b. The CALIOP-derived PM2.5 and in situ measured PM2.5

406 correlate each other with an R2 of 0.42, a slope of 0.76, and a root mean square error of 51.4

407 μgm-3. For each year from 2017 to 2020, the R2 values range between 0.28 and 0.51, as

408 shown in the inset of Fig. 6a. The bias in CALIOP-derived PM2.5 can be attributed to the

409 difference in sampling time between CALIOP and ground-based stations during the course of

410 the day. Daily mean CPCB data represents a 24-hour average, while CALIOP samples at

411 local time ~ 14 hrs during the day and ~21.30 hrs at night over the Indian region. Biases in

412 the estimation of dry mass efficiencies hygroscopic growth factors, and near-surface

413 extinction can also contribute to the observed bias in PM2.5.

21
414 4. Results and Discussion

415 4.1 Regional distribution of PM2.5

416 Near-surface aerosol extinction coefficient (100 to 500 m) gridded at a spatial

417 resolution of 1⁰ x 1⁰ is averaged for the winter season from 2007 to 2021 and is used in

418 equation (1) along with spatially resolved bulk- properties to estimate the climatological

419 mean state of PM2.5 over the Indian region. Dry mass scattering and absorption efficiencies

420 were estimated from the climatological mean of species-resolved mass concentration of

421 MERRA reanalysis averaged over the study period. The hygroscopic growth factor is

422 estimated for each year and was averaged over the study period. The contour map of

423 estimated near-surface PM2.5 concentration is shown in Fig. 7.

22
424

425 Fig. 7. Spatial distribution of PM2.5 mass concentration at 1⁰ x 1⁰ grid resolution during the
426 winter season (December to February) of 2007 – 2021. Dotted lines represent the regions
427 under study.

428 Near-surface PM2.5 shows a large variability over the Indian subcontinent with low

429 concentration over peninsular India and high loading over the IGP and northwest India.

430 Average PM2.5 mass concentration over the IGP (142 ± 62 μgm-3) and northwest India (197 ±

431 42 μgm-3) show 3 to 4 fold increases compared to the relatively clear peninsular India (49 ±

432 17 μgm-3) (Areas are indicated by dotted lines in Fig. 7). The middle part of the IGP shows

433 the highest concentration of PM2.5 compared to the eastern and western parts with a 2-fold

434 increase. According to the National Ambient Air Quality Standards (NAAQS) (revised in

435 2009), the threshold concentration of PM2.5 in ambient air for the safety of human and

436 ecological health is 40 μgm-3

437 (https://cpcb.nic.in/uploads/National_Ambient_Air_Quality_Standards.pdf). The PM2.5 level

23
438 exceeds this standard over most parts of India except the southern part of the peninsula. PM2.5

439 exceeds national standards (NAAQS) over 70% of the monitoring stations under CPCB

440 located in Residential/Industrial/Rural areas and 100% of the stations at ecologically sensitive

441 areas (Gargava and Shukla, 2019). It is estimated that 60% of India's mean

442 PM2.5 concentration is emitted from anthropogenic sources which are domestic biomass

443 combustion, power plants, industrial coal combustion, and anthropogenic dust

444 (Venkataraman et al., 2018). The RF-based PM2.5 estimates showed high PM2.5 levels over the

445 middle and lower IGP, with the annual mean concentration exceeding 110 μgm-3. In winter,

446 the middle and lower IGP experienced very poor air quality, with mean PM2.5 concentrations

447 of >170 μgm-3 (Mhawish et al., 2020). An observational study exploring the relationship

448 between PM2.5 concentration, visibility, and BLH over China showed that aerosol-radiation-

449 boundary layer interaction and feedbacks cause a large, long-standing reduction in air quality

450 during fog–haze mixed events when the PM2.5 concentration is higher than 150 - 200 µg m−3

451 (Luan et al., 2018).

452 4.2 Long-term trend in PM2.5 over India

453 Fig. 8 shows the anomaly in average PM2.5 concentration (difference between mean PM2.5

454 during individual years and the average of PM2.5 over the years 2007 to 2021) over the IGP,

455 northwest India, and peninsular India during the winter (December to February) of the last

456 fifteen years (2007 to 2021). The linear regression model shows that PM2.5 concentration is

457 increasing over the study regions at a rate of 3.4% (7.3 μgm-3) over IGP, 3.3% (9.8 μgm-3)

458 over northwest India, and 4.6% (3.2 μgm-3) over peninsular India per year. The significance

459 of the observed trends is tested with the non-parametric Mann-Kendall test and they showed a

460 confidence level greater than 95%. A recent global study based on passive observations of

461 AOD and chemical transport simulations has shown a positive trend of ~2.44 ± 0.44 μgm-3yr-

24
 1
462 over India from 2005 to 2013, which is remarkably higher than the other parts of the world

463 (Hammer et al., 2020). AOD based observations of PM2.5 show an enhancement during the

464 present decade (2010–2019) compared with the last decade (2000–2010) (Dey et al., 2020).

465 Ground-based observations of AOD under the Aerosol Radiative Forcing over India

466 (ARFINET) network under the ISRO-Biosphere Geosphere Programme shows an increasing

467 trend at a consistent, high rate during the winter season (Babu et al., 2013). Station-based

468 trend analysis of PM2.5 is limited over India due to the lack of long-term in situ observations

469 (Pant et al., 2019). PM2.5 observations carried out under the National Air Quality Monitoring

470 Program (NAMP) managed by CPCB show no significant trend over the five years from

471 2015 to 2019 except for a slight decline in winter over northern India (Sharma and Mauzerall,

472 2022). A seasonal-trend decomposition analysis of annual mean PM2.5 at the US embassies in

473 five Indian megacities showed a decreasing trend of 2% - 8% from 2014 to 2019 (Singh et

474 al., 2021). Annual mean PM2.5 in the Capital city, Delhi, showed a decreasing trend from

475 2013 to 2019 with a higher occurrence of days with good air quality (Li et al., 2021).

476 Although local in situ observations over the urban regions show a decreasing trend in annual

477 mean PM2.5, satellite-derived trends over larger spatial extents show an increasing trend in

478 winter-time PM2.5 over the Indian region.

25
479

480 Fig. 8. The anomaly of average PM2.5 concentration near to surface during winter (December
481 - February) of 2007 to 2021 over a) IGP b) Northwest India (24⁰N - 29⁰N and 67⁰E -72⁰E)
482 and, c) Peninsular India (8⁰N - 20⁰N and 72⁰E - 86⁰E). The red dotted lines represent the fitted
483 linear regression curve. The year corresponds to the January and February months of the
484 season. The area bounded by the green line shows the study regions.

485 Interannual variability in PM2.5 concentration at the surface is determined by

486 variations in meteorological conditions as well as anthropogenic emissions of pollutants.

487 Surface temperature, relative humidity, wind speed, rainfall, and BLH modulate PM

488 concentration through dispersion, transportation, and accumulation of pollutants. Large-scale

489 agricultural crop-residue burning activities over the western part of IGP (Punjab and

490 Haryana) during the post-monsoon season emit a large quantity of aerosols into the

26
491 atmosphere and have been linked with air quality over the IGP during the following winter

492 (Jethva et al., 2019).

493 Fig. 9a shows the spatial distribution of fire counts per 10 km2 over the Indian region

494 averaged over the years 2006 to 2020 and it clearly shows the enhancement in fire activity

495 over the western IGP (Punjab and Haryana) during October and November.

496

497 Fig. 9. a) Average fire count per 10 km2 over India from September to December of 2006 to
498 2020 b) Correlation coefficient of wintertime PM2.5 concentration over the study regions
499 (Peninsular India, Northern India, and Indo-Gangetic Plain as shown by dotted lines in Fig. 7)
500 with fire count over the agro-residue burning region bounded by the green line (29˚N - 32˚N
501 and 74˚E - 77˚E) c) Anomaly in fire count per 10 km2 over the crop-residue burning region

27
502 during October and November over 2006 to 2020 d) Correlation coefficient between
503 wintertime PM2.5 concentration over the study regions and various influencing factors such
504 as fire count during November over the agro-residue burning area, temperature, relative
505 humidity, and wind speed over the study regions.

506 To examine the role of agro-residue burning fire emissions over western IGP in the

507 observed increase in PM2.5 concentration (Fig. 8), a linear regression analysis is carried out

508 between the total fire count detected over the agro-residue burning region, shown in Fig. 9a

509 bounded by the green line (29˚N - 32˚N and 74˚E - 77˚E), and the average near-surface PM2.5

510 concentration over the study regions (Peninsular India, Northern India, and Indo-Gangetic

511 Plain as shown by dotted lines in Fig. 7) during winter. The wintertime PM2.5 mass

512 concentration is averaged spatially over the study regions, for the years from 2007 to 2021

513 (The year corresponds to the January and February months of the season) to obtain yearly

514 time series and the correlation analysis is carried out between time series of PM2.5 with

515 similar time series of fire count over the agro-residue burning region during the preceding

516 October and November. Fig. 9b shows the Pearson’s correlation coefficient of average near-

517 surface PM2.5 concentration during winter (December - February) of 2007 to 2021 with total

518 fire count per 10 sq. km over the western IGP during the preceding October and November.

519 Wintertime PM2.5 over the study regions is positively correlated with fire activity during

520 November (95% confidence) and negatively correlated with fire activity during October

521 (85% confidence). As a result of the implementation of the Preservation of Subsoil Water Act

522 in 2009, a shift in fire emission patterns has been observed from the last week of October to

523 the first week of November due to the delayed cropping and burning activities (Sembhi et al.,

524 2020). Fig. 9c shows the trend in the fire count detected per 10 km2 over the agro-residue

525 burning region of western IGP as marked by the green box in Fig. 9a. While the total fire

526 count over the agro-residue burning region is increasing during November, it shows a

527 decreasing trend during October as shown in Fig. 9c. These contrasting trends in fire activity

528 during the last fifteen years contribute to the observed changes in the correlation coefficient

28
529 between fire activity and PM2.5 during October and November months. Studies using the

530 modeling and ground-based observations have suggested that delayed burning can lead to a

531 worsening of air quality over the source regions (Punjab and Haryana) and downwind over

532 Delhi, during the post-monsoon season due to favorable meteorological conditions

533 (Balwinder et al., 2019; Sawlani et al., 2019; Sembhi et al., 2020). Jethva et al., (2019)

534 attribute the increasing total vegetation index and rice production over Punjab to the increase

535 in the post-agricultural agro-residue burning from 2002 to 2016, which caused a 43%

536 increase in aerosol concentration over the IGP during October and November months. The

537 present study shows that agro-residue burning during November is significantly correlated

538 with PM2.5 concentration during the winter season (December to February).

539 Meteorological conditions during winter, such as weakened northwesterly winds,

540 cooler temperatures, shrunken boundary layer, higher atmospheric stability, and enhanced

541 relative humidity conditions create a stagnant atmosphere that prevents dispersion of

542 pollutants making it conducive to severe fog formation (Sawlani et al., 2019). To compare the

543 relative role of wintertime meteorology with anthropogenic fire activity in enhancing the air

544 pollution over the study regions, linear regression analysis is carried out between wintertime

545 PM2.5 concentrations with different meteorological parameters over the study regions (as

546 indicated by dotted lines in Fig. 7). Fig. 9d shows the Pearson’s correlation coefficient

547 between average wintertime PM2.5 concentration with average values of surface temperature,

548 relative humidity, and wind speed over the study regions and accumulated fire count per 10

549 sq. km over the agro-residue burning area over the western IGP during November during the

550 last 15 years. While a statistically significant (95% confidence) correlation with a coefficient

551 ranging from 0.5 to 0.6 is observed on an interannual scale between PM2.5 concentration over

552 the study regions with post-harvest fire activity, no significant correlation is observed with

553 different meteorological parameters. This indicates that the interannual variability of PM2.5

29
554 over the study regions is mainly controlled by the emission strength of the post-harvest

555 burning.

556 Fig. 10 shows the interannual variability of average PM2.5 concentration during winter over

557 the study regions with the anomaly in fire count over the agro-residue burning area over the

558 western IGP bounded by 29˚N - 32˚N and 74˚E - 77˚E during November over the years 2007

559 to 2021. The agro-residue burning during November is positively associated with wintertime

560 PM2.5 in India during most of the years except from 2015 to 2018 during which a deviation is

561 observed. Since agro-residue fire count does not show a positive correlation with PM2.5 from

562 2015 to 2018, meteorological conditions must have played a dominant role in controlling

563 PM2.5 concentration during that period. Interannual variability of wintertime pollution over

564 northern India has been also linked with natural climate variabilities such as El Niño and the

565 Antarctic Oscillation (AAO) (Gao et al., 2019). Nationwide lock-down as a part of the Covid-

566 19 pandemic in 2020 had brought down PM2.5 concentration to almost half (24 March – 31

567 May 2020) owing to the drastic reduction in particulate emission from anthropogenic sources

568 (Das et al., 2021). After the nationwide lockdown, anthropogenic activities were not restored

569 fully as unlocking was carried out in different phases from June 2020 to January 2021 in

570 various states (https://www.mha.gov.in/sites/default/files/MHAOrderDt_30052020.pdf).

571 During unlocking phases, restrictions were enforced on large gatherings, academic

572 institutions, public recreational activities, tourism, and nighttime travel. Despite the overall

573 reduction in anthropogenic activities, an anomalous increase in PM2.5 (110 μgm-3 from the

574 mean value) was observed during the following winter (December to February) in India as

575 shown in Fig. 10. A similar anomaly is observed in the post-monsoon fire count in

576 November 2020 as well which indicates the dominant contribution of post-monsoon crop-

577 residue burning in the observed enhancement in PM2.5 post-lockdown winter period.

30
578

579 Fig. 10. The anomaly of average near-surface PM2.5 concentration during winter (December -
580 February) of 2007 to 2021 (The year corresponds to the January and February months of the
581 season) and anomaly of fire count during November (preceding year) over the region of agro-
582 residue burning.

583 The interannual variability in fire activity over the western IGP during November is

584 significantly correlated with the variation of PM2.5 concentration over Northwest India as well

585 as peninsular India (95% confidence) (Fig. 10), which indicates the largely extended

586 influence of these emissions over the Indian sub-continent through long-range transport. To

587 examine the spatial extent of influence of post-monsoon fire emissions, the spatial

588 distribution of the correlation coefficient between agro-residue fire count and wintertime

589 AOD of MODIS has been generated. Linear regression analysis is carried out between the

31
590 total fire count of agro-residue burning during November every year from 2006 to 2020 over

591 the western IGP (region bounded by the green box in Fig. 9a) with corresponding average

592 wintertime AOD of MODIS and correlation coefficients have been estimated at 1⁰ x 1⁰ grid

593 resolution as shown by color variation in Fig. 11.

594

595 Fig. 11. Spatial distribution of correlation coefficient between MODIS total fire count (per 10
596 km2 over western IGP) due to agro-residue burning during November with a) MODIS mean
597 AOD during DJF b) MODIS mean AOD during December c) MODIS mean AOD during
598 January of the succeeding year d) MODIS mean AOD during February of the succeeding
599 year, at each 1⁰ x 1⁰ grid cell over the years 2006 to 2020. Black dots indicate pixels with
600 90% confidence, e) Windfield at 850 hPa with color indicating wind speed.

32
601 Interannual variability of agro-residue burning over western IGP (Punjab and

602 Haryana) shows a statistically significant correlation with wintertime AOD over IGP,

603 northwestern part of India, and the eastern part of peninsular India (Fig. 11a), indicating the

604 role of synoptic winds in dispersing the pollutants over large distances throughout the sub-

605 continent (Fig. 11e). During December, AOD over Northwest India shows a strong

606 interannual correlation with agro-residue burning (November). However, in January

607 influence of fire emissions extend to central India and the eastern part of peninsular India

608 with higher correlation coefficients. Long-term observations of fog events show that

609 peninsular India experiences significant occurrences of fog events during December and

610 January months (Sawaisarje et al., 2014), with a statistically significant increasing trend

611 (Srivastava et al., 2017). The long-range transport of bio-mass pollutants from the western

612 IGP towards southern India can contribute to increasing fog events over the peninsular

613 region.

614 Air quality over India had rapidly elevated when the nation has ensued a near “zero-

615 emission” condition during the nationwide lockdown due to the COVID-19 pandemic. This

616 indicates the rapid response of pollutants to controlled emissions. During a two-day-long

617 lockdown due to a national strike in India, black carbon mass concentration had decreased to

618 almost 22% of the average concentration that prevailed before the strike (Kompalli et al.,

619 2014). Restrictions implemented on anthropogenic emissions such as regulation of coal

620 boilers, use of clean fuels for residential purposes, and industrial restructuring, as a part of the

621 Five-year Clean Air Action Plan over China, have led to a reduction of 80% in annual mean

622 PM2.5 concentration in 2017 (Chen et al., 2019). These observations point to the sensitivity of

623 air quality to anthropogenic emissions and the vital role of effective regulations in preventing

624 air pollution and its adverse effects. India's capital city suffers extreme fog events during

625 winter and the implementation of vehicular restrictions with an odd-even scheme did not

33
626 yield the expected improvement in air quality (Chandra et al., 2018). Results from the present

627 study call for stringent emission regulations and the implementation of sustainable agronomic

628 practices to control the emissions from massive post-harvest agro-residue burning and its

629 effect on air quality over the country. As an alternative to disposal of crop-residue through

630 inefficient burning, recycling them through effective energy recovery techniques can benefit

631 the energy demand with minimum environmental impacts. The possibility of utilization of

632 agro-residue in the manufacturing of paper and sustainable construction materials can also be

633 looked into (Madurwar et al., 2013; Varghese et al., 2020).

634 In the projected global warming scenario, the increased land surface temperature can

635 result in a higher occurrence of forest fires and dust episodes, which can emit a large amount

636 of aerosols and further deteriorate the air quality, especially in the tropical atmosphere.

637 Simulations with the regional atmospheric chemistry models are essential to further

638 understand the interaction of anthropogenic air pollution with the regional climate. Data

639 science techniques coupled with a physics-based understanding of the interaction of

640 anthropogenic factors with natural processes will help to improve the fog forecasting systems

641 and deliver effective mitigation.

642 5. Summary

643 The present study estimates particulate matter of diameter less than 2.5µm (PM2.5)

644 over India, during the winter season, using near-surface observations of the space-borne lidar,

645 Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) during the last fifteen years

646 (2007 - 2021) which assumes importance in the context of degrading air quality over India

647 and potential health risks in the future. An increasing trend of ~3% is observed in near-

648 surface PM2.5 concentration over the Indo-Gangetic Plain (IGP) and northwest India, and

649 ~4% over peninsular India from winter of 2007 to 2021 (December 2006 - February 2021),

34
650 which assumes significance in the context of deteriorating air quality and increasing

651 occurrences of winter fog over India. Studies have established a linkage between fire

652 emissions from post-monsoon agro-residue burning over northern India with increased near-

653 surface PM2.5 concentration during the post-monsoon season over the source regions and

654 downwind regions. The present study establishes a statistically significant correlation

655 between fire activity during November with wintertime PM2.5 concentration near the surface

656 on the interannual scale. Interannual variability in near-surface PM2.5 in winter is controlled

657 by variability in agro-residue burning emission strength rather than meteorological

658 conditions. Even though most of the studies on the role of post-harvest agro-residue-burning

659 focus on the downwind regions over the IGP, our analysis shows that the influence of these

660 pollutants extends up to peninsular India through long-range transport and can contribute to

661 continental pollution outflow over to the Northern Indian Ocean. The present study highlights

662 the implications of anthropogenic post-harvest agro-residue burning in attributing to extreme

663 air pollution during the winter season. Sustainable solutions need to be implemented to

664 restrict anthropogenic fire emissions and provide sustainable solutions to contain the

665 excessive emission of PM2.5 and associated socio-economic impacts. Utilization of the energy

666 recovery potential of agro-residue by converting it into solid fuels, high-value chemicals, and

667 multiple consumer products, can provide sustainable solutions to the appropriate disposal of

668 massive amounts of agro-residue produced in India during the post-monsoon season.

669

670

671

672

35
673 Acknowledgment

674 The authors sincerely thank the Director of the National Centre for Earth Science

675 Studies (NCESS) for encouragement and support in carrying out the research work under the

676 Ministry of Earth Sciences, Government of India. The CALIPSO data are obtained from

677 LARC Atmospheric Sciences Data Centre (ASDC) through their website at

678 http://eosweb.larc.nasa.gov/. The MODIS aerosol optical depth data are provided by NASA

679 through https://modis.gsfc.nasa.gov/data/. We acknowledge the use of data from LANCE

680 FIRMS (https://earthdata.nasa.gov/data/near-real-time-data/firms) operated by NASA's Earth

681 Science Data and Information System (ESDIS) with funding provided by NASA

682 Headquarters. We also acknowledge the AIRS project for providing Level 3 monthly gridded

683 data (https://disc.gsfc.nasa.gov/). We acknowledge the Global Modelling and Assimilation

684 Office (GMAO), Greenbelt, MD, USA, Goddard Earth Sciences Data and Information

685 Services Centre (GES DISC) for providing the MERRA-2 reanalysis dataset

686 https://disc.gsfc.nasa.gov/datasets?project=MERRA-2. We thank Central Pollution Control

687 Board for providing surface PM2.5 and PM10 observations

688 (https://app.cpcbccr.com/ccr/#/caaqm-dashboard-all/caaqm-landing).

689 Reference

690 AIRS, 2019. Aqua/AIRS L3 Monthly Standard Physical Retrieval (AIRS-only) 1 degree x 1

691 degree V7.0, Greenbelt, MD, USA, Goddard Earth Sciences Data and Information

692 Services Center (GES DISC), Accessed: [Data Access Date],

693 10.5067/UBENJB9D3T2H. https://doi.org/10.5067/Aqua/AIRS/DATA321

694 Babu, S.S., Manoj, M.R., Moorthy, K.K., Gogoi, M.M., Nair, V.S., Kompalli, S.K., Satheesh,

695 S.K., Niranjan, K., Ramagopal, K., Bhuyan, P.K., Singh, D., 2013. Trends in aerosol

696 optical depth over Indian region: Potential causes and impact indicators. J. Geophys.

36
697 Res. Atmos. 118, 11,794-11,806. https://doi.org/10.1002/2013JD020507

698 Balwinder, S., McDonald, A.J., Srivastava, A.K., Gerard, B., 2019. Tradeoffs between

699 groundwater conservation and air pollution from agricultural fires in northwest India.

700 Nat. Sustain. 2, 580–583. https://doi.org/10.1038/s41893-019-0304-4

701 Bharali, C., Nair, V.S., Chutia, L., Babu, S.S., 2019. Modeling of the Effects of Wintertime

702 Aerosols on Boundary Layer Properties Over the Indo Gangetic Plain. J. Geophys. Res.

703 Atmos. 124, 4141–4157. https://doi.org/10.1029/2018JD029758

704 Boutle, I., Price, J., Kudzotsa, I., Kokkola, H., Romakkaniemi, S., 2018. Aerosol-fog

705 interaction and the transition to well-mixed radiation fog. Atmos. Chem. Phys. 18,

706 7827–7840. https://doi.org/10.5194/acp-18-7827-2018

707 Buchard, V., Randles, C.A., da Silva, A.M., Darmenov, A., Colarco, P.R., Govindaraju, R.,

708 Ferrare, R., Hair, J., Beyersdorf, A.J., Ziemba, L.D., Yu, H., 2017. The MERRA-2

709 aerosol reanalysis, 1980 onward. Part II: Evaluation and case studies. J. Clim. 30, 6851–

710 6872. https://doi.org/10.1175/JCLI-D-16-0613.1

711 Chandra, B.P., Sinha, V., Hakkim, H., Kumar, A., Pawar, H., Mishra, A.K., Sharma, G.,

712 Pallavi, Garg, S., Ghude, S.D., Chate, D.M., Pithani, P., Kulkarni, R., Jenamani, R.K.,

713 Rajeevan, M., 2018. Odd-even traffic rule implementation during winter 2016 in Delhi

714 did not reduce traffic emissions of VOCs, carbon dioxide, methane and carbon

715 monoxide. Curr. Sci. 114, 1318–1325. https://doi.org/10.18520/cs/v114/i06/1318-1325

716 Chen, Z., Chen, D., Kwan, M.P., Chen, B., Gao, B., Zhuang, Y., Li, R., Xu, B., 2019. The

717 control of anthropogenic emissions contributed to 80 % of the decrease in

718 PM2.5 concentrations in Beijing from 2013 to 2017. Atmos. Chem. Phys. 19, 13519–

719 13533. https://doi.org/10.5194/acp-19-13519-2019

37
720 Cohen, A.J., Brauer, M., Burnett, R., Anderson, H.R., Frostad, J., Estep, K., Balakrishnan, K.,

721 Brunekreef, B., Dandona, L., Dandona, R., Feigin, V., Freedman, G., Hubbell, B.,

722 Jobling, A., Kan, H., Knibbs, L., Liu, Y., Martin, R., Morawska, L., Pope, C.A., Shin,

723 H., Straif, K., Shaddick, G., Thomas, M., van Dingenen, R., van Donkelaar, A., Vos, T.,

724 Murray, C.J.L., Forouzanfar, M.H., 2017. Estimates and 25-year trends of the global

725 burden of disease attributable to ambient air pollution: an analysis of data from the

726 Global Burden of Diseases Study 2015. Lancet 389, 1907–1918.

727 https://doi.org/10.1016/S0140-6736(17)30505-6

728 Das, M., Das, A., Sarkar, R., Saha, S., Mandal, A., 2021. Examining the impact of lockdown

729 (due to COVID-19) on ambient aerosols (PM2.5): A study on Indo-Gangetic Plain (IGP)

730 Cities, India. Stoch. Environ. Res. Risk Assess. 35, 1301–1317.

731 https://doi.org/10.1007/s00477-020-01905-x

732 David, L.M., Ravishankara, A.R., Kodros, J.K., Pierce, J.R., Venkataraman, C., Sadavarte,

733 P., 2019. Premature Mortality Due to PM2.5 Over India: Effect of Atmospheric

734 Transport and Anthropogenic Emissions. GeoHealth 3, 2–10.

735 https://doi.org/10.1029/2018GH000169

736 Dey, S., Di Girolamo, L., van Donkelaar, A., Tripathi, S.N., Gupta, T., Mohan, M., 2012.

737 Variability of outdoor fine particulate (PM 2.5) concentration in the Indian

738 Subcontinent: A remote sensing approach. Remote Sens. Environ. 127, 153–161.

739 https://doi.org/10.1016/j.rse.2012.08.021

740 Dey, S., Purohit, B., Balyan, P., Dixit, K., Bali, K., Kumar, A., Imam, F., Chowdhury, S.,

741 Ganguly, D., Gargava, P., Shukla, V.K., 2020. A Satellite-Based High-Resolution (1-

742 km) Ambient PM2.5 Database for India over Two Decades (2000–2019): Applications

743 for Air Quality Management. Remote Sens. 2020, Vol. 12, Page 3872 12, 3872.

38
744 https://doi.org/10.3390/RS12233872

745 Dominici, F., Peng, R.D., Bell, M.L., Pham, L., McDermott, A., Zeger, S.L., Samet, J.M.,

746 2006. Fine particulate air pollution and hospital admission for cardiovascular and

747 respiratory diseases. J. Am. Med. Assoc. 295, 1127–1134.

748 https://doi.org/10.1001/jama.295.10.1127

749 Gao, M., Sherman, P., Song, S., Yu, Y., Wu, Z., McElroy, M.B., 2019. Seasonal prediction of

750 Indian wintertime aerosol pollution using the ocean memory effect. Sci. Adv. 5.

751 https://doi.org/10.1126/sciadv.aav4157

752 Gargava, P., Shukla, V.K., 2019. National Ambient Air quality status and trends 2019,

753 NAAQMS /45/2019-2020, Central Pollution Control Board, Ministry of Environment,

754 Forest and Climate change, Government of India.

755 Gelaro, R., McCarty, W., Suárez, M.J., Todling, R., Molod, A., Takacs, L., Randles, C.A.,

756 Darmenov, A., Bosilovich, M.G., Reichle, R., Wargan, K., Coy, L., Cullather, R.,

757 Draper, C., Akella, S., Buchard, V., Conaty, A., da Silva, A.M., Gu, W., Kim, G.K.,

758 Koster, R., Lucchesi, R., Merkova, D., Nielsen, J.E., Partyka, G., Pawson, S., Putman,

759 W., Rienecker, M., Schubert, S.D., Sienkiewicz, M., Zhao, B., 2017. The modern-era

760 retrospective analysis for research and applications, version 2 (MERRA-2). J. Clim. 30,

761 5419–5454. https://doi.org/10.1175/JCLI-D-16-0758.1

762 Giglio, L., Schroeder, W., Justice, C.O., 2016. The collection 6 MODIS active fire detection

763 algorithm and fire products. Remote Sens. Environ. 178, 31–41.

764 https://doi.org/10.1016/j.rse.2016.02.054

765 Hammer, M.S., Van Donkelaar, A., Li, C., Lyapustin, A., Sayer, A.M., Hsu, N.C., Levy,

766 R.C., Garay, M.J., Kalashnikova, O. V., Kahn, R.A., Brauer, M., Apte, J.S., Henze,

39
767 D.K., Zhang, L., Zhang, Q., Ford, B., Pierce, J.R., Martin, R. V., 2020. Global Estimates

768 and Long-Term Trends of Fine Particulate Matter Concentrations (1998-2018). Environ.

769 Sci. Technol. 54, 7879–7890. https://doi.org/10.1021/acs.est.0c01764

770 Hegde, P., Vyas, B.M., Aswini, A.R., Aryasree, S., Nair, P.R., 2020. Carbonaceous and

771 water-soluble inorganic aerosols over a semi-arid location in north west India: Seasonal

772 variations and source characteristics. J. Arid Environ. 172, 104018.

773 https://doi.org/10.1016/j.jaridenv.2019.104018

774 Hess, M., Koepke, P., Schult, I., 1998. Optical Properties of Aerosols and Clouds: The

775 Software Package OPAC. Bull. Am. Meteorol. Soc. 79, 831–844.

776 https://doi.org/10.1175/1520-0477(1998)079<0831:OPOAAC>2.0.CO;2

777 Jethva, H., Torres, O., Field, R.D., Lyapustin, A., Gautam, R., Kayetha, V., 2019. Connecting

778 Crop Productivity, Residue Fires, and Air Quality over Northern India. Sci. Rep. 9, 1–

779 11. https://doi.org/10.1038/s41598-019-52799-x

780 Jia, B., Gao, M., Zhang, X., Xiao, X., Zhang, S., Lam Yung, K.K., 2021. Rapid increase in

781 mortality attributable to PM2.5 exposure in India over 1998–2015. Chemosphere 269,

782 128715. https://doi.org/10.1016/j.chemosphere.2020.128715

783 Kamyotra, S.J.S., Saha, D.D., 2013. Guidelines for the Measurement of Ambient Air

784 Pollutants Volume-I, National Ambient Air Quality Series: NAAQMS/36/2012-13,

785 Central Pollution Control Board, Ministry of Environment & Forests, Government of

786 India.

787 Kim, M.H., Omar, A.H., Tackett, J.L., Vaughan, M.A., Winker, D.M., Trepte, C.R., Hu, Y.,

788 Liu, Z., Poole, L.R., Pitts, M.C., Kar, J., Magill, B.E., 2018. The CALIPSO version 4

789 automated aerosol classification and lidar ratio selection algorithm. Atmos. Meas. Tech.

40
790 11, 6107–6135. https://doi.org/10.5194/amt-11-6107-2018

791 Koepke, P., Gasteiger, J., Hess, M., 2015. Technical Note: Optical properties of desert

792 aerosol with non-spherical mineral particles: Data incorporated to OPAC. Atmos. Chem.

793 Phys. 15, 5947–5956. https://doi.org/10.5194/acp-15-5947-2015

794 Kompalli, S.K., Moorthy, K.K., Babu, S.S., 2014. Rapid response of atmospheric BC to

795 anthropogenic sources: Observational evidence. Atmos. Sci. Lett. 15, 166–171.

796 https://doi.org/10.1002/asl2.483

797 Kumar, M., Parmar, K.S., Kumar, D.B., Mhawish, A., Broday, D.M., Mall, R.K., Banerjee,

798 T., 2018. Long-term aerosol climatology over Indo-Gangetic Plain: Trend, prediction

799 and potential source fields. Atmos. Environ. 180, 37–50.

800 https://doi.org/10.1016/j.atmosenv.2018.02.027

801 Kuttippurath, J., Raj, S., 2021. Two decades of aerosol observations by AATSR, MISR,

802 MODIS and MERRA-2 over India and Indian Ocean. Remote Sens. Environ. 257,

803 112363. https://doi.org/10.1016/j.rse.2021.112363

804 Lakshmi, N.B., Nair, V.S., Babu, S.S., 2021. Assessment of the vertical distribution of

805 speciated aerosol absorption over South Asia using spaceborne LIDAR and ground-

806 based observations. Remote Sens. Environ. 253, 112164.

807 https://doi.org/10.1016/j.rse.2020.112164

808 Lakshmi, N.B., Nair, V.S., Suresh Babu, S., 2017. Vertical structure of aerosols and mineral

809 dust over the Bay of Bengal from multisatellite observations. J. Geophys. Res. Atmos.

810 122, 12,845-12,861. https://doi.org/10.1002/2017JD027643

811 Li, J., Hao, X., Liao, H., Hu, J., Chen, H., 2021. Meteorological Impact on Winter PM2.5

41
812 Pollution in Delhi: Present and Future Projection Under a Warming Climate. Geophys.

813 Res. Lett. 48, e2021GL093722. https://doi.org/10.1029/2021GL093722

814 Liu, Z., Vaughan, M., Winker, D., Kittaka, C., Getzewich, B., Kuehn, R., Omar, A., Powell,

815 K., Trepte, C., Hostetler, C., 2009. The CALIPSO lidar cloud and aerosol

816 discrimination: Version 2 algorithm and initial assessment of performance. J. Atmos.

817 Ocean. Technol. 26, 1198–1213. https://doi.org/10.1175/2009JTECHA1229.1

818 Luan, T., Guo, X., Guo, L., Zhang, T., 2018. Quantifying the relationship between PM2.5

819 concentration, visibility and planetary boundary layer height for long-lasting haze and

820 fog-haze mixed events in Beijing. Atmos. Chem. Phys. 18, 203–225.

821 https://doi.org/10.5194/acp-18-203-2018

822 Madurwar, M. V., Ralegaonkar, R. V., Mandavgane, S.A., 2013. Application of agro-waste

823 for sustainable construction materials: A review. Constr. Build. Mater. 38, 872–878.

824 https://doi.org/10.1016/j.conbuildmat.2012.09.011

825 Mandariya, A.K., Tripathi, S.N., Gupta, T., Mishra, G., 2020. Wintertime hygroscopic

826 growth factors (HGFs) of accumulation mode particles and their linkage to chemical

827 composition in a heavily polluted urban atmosphere of Kanpur at the Centre of IGP,

828 India: Impact of ambient relative humidity. Sci. Total Environ. 704.

829 https://doi.org/10.1016/j.scitotenv.2019.135363

830 Maßling, A., Wiedensohler, A., Busch, B., Neusüß, C., Quinn, P., Bates, T., Covert, D., 2003.

831 Hygroscopic properties of different aerosol types over the Atlantic and Indian Oceans.

832 Atmos. Chem. Phys. 3, 1377–1397. https://doi.org/10.5194/acp-3-1377-2003

833 Mhawish, A., Banerjee, T., Sorek-Hamer, M., Bilal, M., Lyapustin, A.I., Chatfield, R.,

834 Broday, D.M., 2020. Estimation of High-Resolution PM2.5over the Indo-Gangetic Plain

42
835 by Fusion of Satellite Data, Meteorology, and Land Use Variables. Environ. Sci.

836 Technol. 54, 7891–7900. https://doi.org/10.1021/acs.est.0c01769

837 Miyazaki, Y., Aggarwal, S.G., Singh, K., Gupta, P.K., Kawamura, K., 2009. Dicarboxylic

838 acids and water-soluble organic carbon in aerosols in New Delhi, India, in winter:

839 Characteristics and formation processes. J. Geophys. Res. Atmos. 114, 19206.

840 https://doi.org/10.1029/2009JD011790

841 Moorthy, K.K., Babu, S.S., Manoj, M.R., Satheesh, S.K., 2013. Buildup of aerosols over the

842 Indian Region. Geophys. Res. Lett. 40, 1011–1014. https://doi.org/10.1002/grl.50165

843 Nair, V.S., Babu, S.S., Manoj, M.R., Moorthy, K.K., Chin, M., 2017. Direct radiative effects

844 of aerosols over South Asia from observations and modeling. Clim. Dyn. 49, 1411–

845 1428. https://doi.org/10.1007/s00382-016-3384-0

846 Nair, V.S., Moorthy, K.K., Alappattu, D.P., Kunhikrishnan, P.K., George, S., Nair, P.R.,

847 Babu, S.S., Abish, B., Satheesh, S.K., Tripathi, S.N., Niranjan, K., Madhavan, B.L.,

848 Srikant, V., Dutt, C.B.S., Badarinath, K.V.S., Reddy, R.R., 2007. Wintertime aerosol

849 characteristics over the Indo-Gangetic Plain (IGP): Impacts of local boundary layer

850 processes and long-range transport. J. Geophys. Res. Atmos. 112.

851 https://doi.org/10.1029/2006JD008099

852 Ojha, N., Sharma, A., Kumar, M., Girach, I., Ansari, T.U., Sharma, S.K., Singh, N., Pozzer,

853 A., Gunthe, S.S., 2020. On the widespread enhancement in fine particulate matter across

854 the Indo-Gangetic Plain towards winter. Sci. Rep. 10. https://doi.org/10.1038/s41598-

855 020-62710-8

856 Omar, A.H., Winker, D.M., Kittaka, C., Vaughan, M.A., Liu, Z., Hu, Y., Trepte, C.R.,

857 Rogers, R.R., Ferrare, R.A., Lee, K.P., Kuehn, R.E., Hostetler, C.A., 2009. The

43
858 CALIPSO automated aerosol classification and lidar ratio selection algorithm. J. Atmos.

859 Ocean. Technol. 26, 1994–2014. https://doi.org/10.1175/2009JTECHA1231.1

860 Platnick, S., King, M.D., Meyer, K.G., Wind, G., Amarasinghe, N., Marchant, B., Arnold,

861 G.T., Zhang, Z., Hubanks, P.A., Ridgway, B., 2017. MODIS Atmosphere L3 Monthly

862 Product. NASA MODIS Adaptive Processing System, Goddard Space Flight Center,

863 USA.

864 Rajeev, K., Mishra, M.K., Sunilkumar, S. V., Sijikumar, S., 2016. Dual polarization

865 micropulse lidar observations of the diurnal evolution of atmospheric boundary layer

866 over a tropical coastal station, in: Lidar Remote Sensing for Environmental Monitoring

867 XV. p. 98790V. https://doi.org/10.1117/12.2228049

868 Ram, K., Sarin, M.M., Tripathi, S.N., 2012. Temporal trends in atmospheric PM 2.5, PM 10,

869 elemental carbon, organic carbon, water-soluble organic carbon, and optical properties:

870 Impact of biomass burning emissions in the Indo-Gangetic Plain. Environ. Sci. Technol.

871 46, 686–695. https://doi.org/10.1021/es202857w

872 Randles, C.A., da Silva, A.M., Buchard, V., Colarco, P.R., Darmenov, A., Govindaraju, R.,

873 Smirnov, A., Holben, B., Ferrare, R., Hair, J., Shinozuka, Y., Flynn, C.J., 2017. The

874 MERRA-2 aerosol reanalysis, 1980 onward. Part I: System description and data

875 assimilation evaluation. J. Clim. 30, 6823–6850. https://doi.org/10.1175/JCLI-D-16-

876 0609.1

877 Sahu, S.K., Mangaraj, P., Beig, G., Tyagi, B., Tikle, S., Vinoj, V., 2021. Establishing a link

878 between fine particulate matter (PM2.5) zones and COVID -19 over India based on

879 anthropogenic emission sources and air quality data. Urban Clim. 38, 100883.

880 https://doi.org/10.1016/j.uclim.2021.100883

44
881 Sarangi, B., Ramachandran, S., Rajesh, T.A., Dhaker, V.K., 2019. Black carbon linked

882 aerosol hygroscopic growth: Size and mixing state are crucial. Atmos. Environ. 200,

883 110–118. https://doi.org/10.1016/j.atmosenv.2018.12.001

884 Satheesh, S.K., Srinivasan, J., 2002. Enhanced aerosol loading over Arabian Sea during the

885 pre-monsoon season: Natural or anthropogenic? Geophys. Res. Lett. 29, 21–1.

886 https://doi.org/10.1029/2002GL015687

887 Sawaisarje, G.K., Khare, P., Shirke, C.Y., Deepakumar, S., Narkhede, N.M., 2014. Study of

888 winter fog over Indian subcontinent: Climatological perspectives. Mausam 65, 19–28.

889 https://doi.org/10.54302/MAUSAM.V65I1.858

890 Sawlani, R., Agnihotri, R., Sharma, C., Patra, P.K., Dimri, A.P., Ram, K., Verma, R.L., 2019.

891 The severe Delhi SMOG of 2016: A case of delayed crop residue burning, coincident

892 firecracker emissions, and atypical meteorology. Atmos. Pollut. Res. 10, 868–879.

893 https://doi.org/10.1016/j.apr.2018.12.015

894 Sembhi, H., Wooster, M., Zhang, T., Sharma, S., Singh, N., Agarwal, S., Boesch, H., Gupta,

895 S., Misra, A., Tripathi, S.N., Mor, S., Khaiwal, R., 2020. Post-monsoon air quality

896 degradation across Northern India: Assessing the impact of policy-related shifts in

897 timing and amount of crop residue burnt. Environ. Res. Lett. 15.

898 https://doi.org/10.1088/1748-9326/aba714

899 Singh, N., Murari, V., Kumar, M., Barman, S.C., Banerjee, T., 2017. Fine particulates over

900 South Asia: Review and meta-analysis of PM2.5 source apportionment through receptor

901 model. Environ. Pollut. 223, 121–136. https://doi.org/10.1016/j.envpol.2016.12.071

902 Srivastava, S.K., Sharma, A.R., Sachdeva, K., 2017. An observation-based climatology and

903 forecasts of winter fog in Ghaziabad, India. Weather 72, 16–22.

45
904 https://doi.org/10.1002/wea.2743

905 Toth, T.D., Zhang, J., Reid, J.S., Vaughan, M.A., 2019. A bulk-mass-modeling-based method

906 for retrieving particulate matter pollution using CALIOP observations. Atmos. Meas.

907 Tech. 12, 1739–1754. https://doi.org/10.5194/amt-12-1739-2019

908 Varghese, L.M., Agrawal, S., Nagpal, R., Mishra, O.P., Bhardwaj, N.K., Mahajan, R., 2020.

909 Eco-friendly pulping of wheat straw using crude xylano-pectinolytic concoction for

910 manufacturing good quality paper. Environ. Sci. Pollut. Res. 27, 34574–34582.

911 https://doi.org/10.1007/s11356-020-10119-1

912 Vaughan, M., Pitts, M., Trepte, C., Winker, D., Detweiler, P., Garnier, A., Getzewich, B.,

913 Hunt, W., Lambeth, J., Lee, K.P., Lucker, P., Murray, T., Rodier, S., Tremas, T.,

914 Bazureau, A., Pelon, J., 2018. Cloud – Aerosol LIDAR Infrared Pathfinder Satellite

915 Observations Data Management System Data Products Catalog, Release 4.20, Document

916 No: PC-SCI-503 (https://www-

917 calipso.larc.nasa.gov/products/CALIPSO_DPC_Rev4x20.pdf).

918 Venkataraman, C., Brauer, M., Tibrewal, K., Sadavarte, P., Ma, Q., Cohen, A.,

919 Chaliyakunnel, S., Frostad, J., Klimont, Z., Martin, R. V., Millet, D.B., Philip, S.,

920 Walker, K., Wang, S., 2018. Source influence on emission pathways and ambient PM2.5

921 pollution over India (2015-2050). Atmos. Chem. Phys. 18, 8017–8039.

922 https://doi.org/10.5194/acp-18-8017-2018

923 Vinoj, V., Pandey, S.K., 2022. Role of meteorology in atmospheric aerosols and air pollution

924 over South Asia. Asian Atmos. Pollut. 97–110. https://doi.org/10.1016/b978-0-12-

925 816693-2.00018-4

926 Winker, D.M., Tackett, J.L., Getzewich, B.J., Liu, Z., Vaughan, M.A., Rogers, R.R., 2013.

46
927 The global 3-D distribution of tropospheric aerosols as characterized by CALIOP.

928 Atmos. Chem. Phys. 13, 3345–3361. https://doi.org/10.5194/acp-13-3345-2013

929 Winker, D.M., Vaughan, M.A., Omar, A., Hu, Y., Powell, K.A., Liu, Z., Hunt, W.H., Young,

930 S.A., 2009. Overview of the CALIPSO mission and CALIOP data processing

931 algorithms. J. Atmos. Ocean. Technol. 26, 2310–2323.

932 https://doi.org/10.1175/2009JTECHA1281.1

933 Young, S.A., Vaughan, M.A., Kuehn, R.E., Winker, D.M., 2013. The retrieval of profiles of

934 particulate extinction from Cloud-Aerosol Lidar and Infrared Pathfinder Satellite

935 Observations (CALIPSO) data: Uncertainty and error sensitivity analyses. J. Atmos.

936 Ocean. Technol. 30, 395–428. https://doi.org/10.1175/JTECH-D-12-00046.1

937

938

939

940

47