Aerosol Radiactive

Atmospheric Environment 298 (2023) 119609
Contents lists available at ScienceDirect
Atmospheric Environment
journal homepage: www.elsevier.com/locate/atmosenv
Aerosol radiative feedback enhances particulate pollution over India: A

process understanding
Arushi Sharma a, Chandra Venkataraman a, b, *, Kaushik Muduchuru a, Vikas Singh c,
Amit Kesarkar c, Sudipta Ghosh d, Sagnik Dey d
a
Interdisciplinary Program in Climate Studies, Indian Institute of Technology Bombay, Mumbai, Maharashtra, India
b
Department of Chemical Engineering, Indian Institute of Technology Bombay, Mumbai, Maharashtra, India
c
National Atmospheric Research Laboratory, Gadanki, Andhra Pradesh, India
d
Centre for Atmospheric Sciences, Indian Institute of Technology Delhi, New Delhi, India
H I G H L I G H T S G R A P H I C A L A B S T R A C T
• Online coupled WRF-Chem simulations

was analysed to investigate aerosol
direct effect.
• Direct effects reduced radiation, tem
perature and PBL height, annually and
seasonally.
• PM2.5 levels increased due to ADE where
primary aerosol concentrations were
dominant.
• The effects reduced O3 in most of the
regions, due to high levels of NO2 and
SO2.
• PM2.5 levels due to ADE has implications
on estimation of premature mortality.
A R T I C L E I N F O A B S T R A C T
Keywords: In this study, we investigate the role of aerosol direct radiative feedback on regional meteorological changes and
WRF-Chem subsequent distribution of air pollutant with the focus on particulate matter PM2.5 and ozone (O3) over India.
Air quality simulations WRF-Chem simulations have been applied for Baseline case with both aerosol direct and indirect effects and
Meteorology
another simulation including only aerosol indirect effects. The aerosol direct radiative feedback on the meteo
Atmospheric chemistry
Primary and secondary species
rology and air quality were investigated by taking the differences between these two simulations. A compre
hensive model evaluation showed the model had the capacity to reproduce the observations and well captured
the temporal and spatial variation in meteorological and aerosol fields. We found that the reduction in incoming
shortwave solar radiation, temperature at 2 m and planetary boundary layer height annually and across all the
season due to aerosol radiation interaction (ARI). The concentration of PM2.5 and gaseous precursors like SO2
and NO2 showed enhancements, while O3 concentration mostly showed a reduction. Spatial and annual average
PM2.5 concentration was increased by +6%, SO2 concentration was increased by +1.4%, NO2 concentration was
increased by +3.4%, and O3 concentration was reduced by − 1.4%. Largest regional increases of 40% in par
ticulate pollution induced by ARI occurred over north-western Indian region. The increase in PM2.5 concentra
tion was highly related to stabilisation induced through meteorological variables and increases in primary
aerosol concentration. The decrease in O3 concentration was highly related increases in NO2 and SO2
* Corresponding author. Particle and Aerosol Research Lab, Room 225, Department of Chemical Engineering, Indian Institute of Technology Bombay, Powai,
Mumbai, 400076, Maharashtra, India.
E-mail address: [email protected] (C. Venkataraman).
https://doi.org/10.1016/j.atmosenv.2023.119609
Received 2 June 2022; Received in revised form 18 January 2023; Accepted 22 January 2023
Available online 30 January 2023
1352-2310/© 2023 Elsevier Ltd. All rights reserved.
A. Sharma et al. Atmospheric Environment 298 (2023) 119609
concentration. In the most polluted regions of India, ARI significantly enhance surface-level particulate pollution
and related premature mortality, of particular concern considering likely future increasing trends. This study
highlights the importance of inclusion and better representations of ARI and for implementing effective miti
gation plans.
1. Introduction et al., 2011). This diversity results in seasonally varying

physico-chemical and optical properties (Hu et al., 2005), with dust
Atmospheric aerosols exert radiative forcing by scattering and domination in summer (Gautam et al., 2009) and biomass fuel, but
absorbing solar radiation, termed as the aerosol direct effect (ADE). agricultural residue and biomass fuel burning domination in the
Absorbing aerosols like black carbon (BC) and brown carbon (BrC) have post-monsoon and winter seasons (David et al., 2018; Sadavarte et al.,
a medium to high mass absorption efficiency in the shortwave spectrum 2016). The effects of aerosol radiative feedback on air quality in India
(Jacobson, 2001; Liu et al., 2015), while dust absorbs preferentially in are not yet studied. Therefore, we investigate the ADE feedback to PM2.5
the longwave spectrum (Heinold et al., 2008). In contrast, scattering and its constituents, as well as precursor gases and ozone, via changes in
aerosols like sulphate (SO4) and organic aerosol (OA) scatter radiation, meteorology and atmospheric chemistry.
but could enhance absorption when present as a coating or a shell on a In this study, we conduct a pair of simulations with the Weather
black carbon core (Schnaiter et al., 2003; Tasoglou et al., 2017). Typi Research and Forecasting (WRF) model coupled with Chemistry (WRF-
cally, in world regions with a dominance of scattering aerosols such as Chem) for a full year over India (2015), with and without aerosol ra
sulphates, ADE induces a decrease in solar radiation reaching the sur diation interaction (ARI) coupling to meteorology, to evaluate ARI ef
face, with concomitant cooling of surface temperature (Atwater, 1970; fects on particulate pollution levels and their implication for premature
Gao et al., 2015; Sokolik and Toon, 1996; Yu et al., 2006). An abundance mortality. Section 2 describes the model setup along with the method
of BC and dust, which induce diabatic heating, could lead to a warming ology for emissions input and dust tuning, the sensitivity experiment
of surface temperatures in regions of China (Chen et al., 2020) and India design, and discusses the details of observational data used to validate
(Mondal et al., 2020). Further, solar absorption by BC can heat the at the model. In section 3.1, we evaluate the model-simulated meteoro
mosphere and reduce cloud formation (Gao et al., 2016; Twomey, logical variables and aerosol loading with observations. In section 3.2,
1991). Together, these effects can alter meteorological fields like hu the effects of ARI on meteorology and PM2.5 and its components are
midity, lapse rate, precipitation, wind speed and planetary boundary demonstrated. Processes that control the distribution of major compo
layer mixing (Forkel et al., 2012; Gao et al., 2015; Xing et al., 2017). nents of PM2.5 are investigated. In section 3.3, we discuss that exposure
Reduction in downwelling radiation from aerosol absorption can change to PM2.5 can cause adverse health effects leading to premature mortality.
the abundances of atmospheric oxidants like hydroxyl radical (OH), Section 4 concludes the important takeaways of this study.
hydrogen peroxide (H2O2), ozone (O3) via photochemical processes,
which affect the production of secondary SO4 and organic aerosol (SOA). 2. Methodology and data
Under conditions of significant aerosol loading, such ADE-induced al
terations in meteorological fields and atmospheric chemistry can affect In this study, we conducted an online-coupled WRF-Chem simulation
air quality. for 2015, covering the South Asia Cordex Domain, to study the impact of
Atmospheric meteorology-chemistry models have been used to ARI on particulate pollution levels, composition and its implications for
evaluate links among meteorology, emissions, atmospheric processes estimated premature mortality.
and ambient concentrations of air pollutants (Nguyen, 2014; Daly and Three simulations were made, one with WRF and two with WRF-
Zannetti, 2007). Recent studies using coupled meteorology-chemistry Chem (Table S4). The two WRF-Chem experiments were made for the
models have employed either multi-decadal simulations (Xing et al., entire year of 2015, with one month of spin-up: the first simulation
2015) or pairs of sensitivity simulations, with controls on radiative included aerosol direct, semi-direct and indirect effects (BASE), while
feedback to meteorology and chemistry (Li et al., 2017; Nguyen et al., the second simulation included only aerosol indirect effects through
2019a, 2019b; Wang et al., 2016a), to evaluate ADE influence on air cloud interaction and excluded aerosol direct and semi-direct effects
quality. The modelling literature has widely considered East Asia (NoARI). The BASE run is expected to simulate a ‘real-world’ response
(Nguyen et al., 2019a, 2019b; Wang et al., 2014, 2016a; Xing et al., considering both the changes in radiation and cloud properties caused
2015, 2017), North America and Europe (Forkel et al., 2012, 2015a; by their interaction with aerosols and hence is used to validate the model
Xing et al., 2015). Many studies point to ADE influences leading to in simulation output, such as meteorological variables, aerosol concen
creases in both PM2.5 and ozone concentrations from ~2 to ~20% (Hong tration and aerosol properties with observation. The difference between
et al., 2017a; Xing et al., 2015), while others indicate opposing changes these BASE and NoARI simulations will help us understand the impact of
with increases in PM2.5, but decreases in ozone concentrations (Nguyen direct, semi-direct and rapid adjustment effects (also called aerosol ra
et al., 2019a, 2019b; Wang et al., 2016a). Underlying mechanisms diation interaction (ARI)) on meteorological variables and air pollutant
include reducing surface-reaching solar radiation, increasing atmo levels. The model setup discussed below includes physics, chemistry,
spheric stabilisation, and consequent reduction in planetary boundary boundary conditions and meteorological data assimilation.
layer height and ventilation (Hong et al., 2017a; Nguyen et al., 2019a,
2019b; Wang et al., 2016a; Xing et al., 2015). Larger impacts of ADE on
2.1. WRF-chem model configuration and meteorological data assimilation
PM2.5 concentrations enhancements (up to ~20%) are found during
haze episodes (Wang et al., 2016a, 2019a; Xing et al., 2015).
The WRF-Chem model version 3.9.1 was used with a horizontal
India experiences some of the highest aerosol loading worldwide
resolution of 27 km and 48 vertical layers from the surface to 50 hPa.
(Johnson, 2020), with significant health impacts (Pandey et al., 2021).
The model domain covered the CORDEX South Asia region as shown in
The regional aerosol is a mixture of dust transported from desert regions
Fig. S1. The initial and lateral boundary conditions for the meteoro
in West Asia and Africa (Sijikumar et al., 2016), significant anthropo
logical fields were obtained from the European Centre for Medium-
genic aerosol from fossil fuel and biomass burning (Sadavarte et al.,
Range Weather Forecasts (ECMWF) reanalysis – Interim (ERA-Interim)
2016; Srivastava and Naja, 2021; Venkataraman et al., 2005, 2018), as
dataset (Dee et al., 2011; Simmons et al., 2014). The chemical boundary
well as secondary species including sulphate arising from SO2 emissions
conditions were extracted from the Model for Ozone and Related
from power plants and industries (Guttikunda et al., 2014; Henriksson
chemical Tracers (MOZART) (Emmons et al., 2010). The details of the
2
physics options chosen for the meteorological part are listed in Table S1. Global Ozone Chemistry Aerosol Radiation and Transport (GOCART)
The chemical gas-phase mechanism included the second-generation scheme (Ginoux et al., 2001). It is a function of particle size, air density,
Regional Acid Deposition Model (RADM2) (Stockwell et al., 1990);), a and surface moisture. The AFWA GOCART dust scheme divides dust
condensed gas-phase photooxidation mechanism with 63 chemical particles into five predetermined bins based on their effective aerosol
species and 136 gas-phase reactions. The applied aerosol module for size. These size bins are B1 (Dp = 1.46 μm), B2 (Dp = 2.8 μm), B3 (Dp =
secondary inorganic (SIA) and organic aerosols (SOA) was Modal 4.8 μm), B4 (Dp = 9 μm) and B5 (Dp = 16 μm). We mapped the first three
Aerosol Dynamics Model for Europe/Secondary Organic Aerosol Model dust-size bins of GOCART (B1, B2 and B3) to the MADE three-lognormal
(MADE/SORGAM) (Ackermann et al., 1998; Schell et al., 2001), a modal distribution to minimise the underestimation of simulated PM2.5 and
scheme which uses a three-lognormal approach for aerosol size distri aerosol optical depth (AOD). Multiple test simulations with different
bution: the Aitken mode (<0.1 μm diameter), the accumulation mode dust size mapping were performed for April because this month has high
(0.1–2 μm diameter), and the coarse mode (>2 μm diameter). In each dust concentrations over the Indian region (Gautam et al., 2009).
mode particles were assumed to have the same chemical composition Table S3 describes the final dust mapping chosen for this study, i.e. 10%
(internally mixed), while externally mixed among different modes (Zhao of B1 is mapped to Aitken mode, 90% of B1 is mapped to the Accumu
et al., 2010). The three-dimensional (3D) variational (VAR) assimilation lation mode, and the aggregation of 100% of B2, 100% of B3 and 50% of
technique (Barker et al., 2004) was cyclically applied every 12 h to the B4 is mapped to coarse mode. Using this mapping, the normalised mean
simulated meteorological fields while keeping the chemical fields bias (NMB) of AOD was improved to − 29% compared to other test
continuous across the 12-h segments using the Weather Research and simulations, and the contribution of dust to PM2.5 concentration was
Forecasting (WRF) model data assimilation system (WRFDA) module. ~40% which was in accordance with the expert’s judgement for April
The 12 h of meteorology assimilation in initial conditions allowed the month.
model to adequately evolve and captured the aerosol-feedback effects
while simulating meteorological conditions closer to observations. Many 2.2. Premature mortality calculation
studies have shown that the data assimilation has improved the pre
diction of surface PM2.5 and O3 concentrations (Sabari and Reynoso, We used the methodology described in GBD, 2018 (Murray et al.,
2020). Assimilation datasets used here were obtained from National 2020) to calculate the premature mortality due to exposure of PM2.5
Centers for Environmental Prediction (NCEP, https://rda.ucar.edu/d concentrations which included several methodological updates and new
atasets/ds337.0/), which has data collection from land and marine input data for estimation of deaths and DALYs. We performed this
surface, radiosonde, pibal and aircraft reports, US radar-derived winds, calculation twice, both for the BASE and NoARI simulations. The dif
the Special Sensor Microwave Imager (SSM/I) oceanic winds and the ferences between these suggest the total attributable deaths to only ARI
Total Cloud Water (TCW) retrievals, and satellite wind data. effect. For this study, we estimated mortality for six different causes: for
all ages in case of Chronic Obstructive Pulmonary Disease (COPD), Lung
2.1.1. Emissions input cancer (LC), Type II Diabetes (T2_DM), Ischemic Heart Disease (IHD),
For this work, we combined different anthropogenic emission in and stroke and children less than 5 years of age in Lower Respiratory
ventories to include the diverse sources of emissions and to better pre tract Infection (LRI). We obtained 2015 gridded population data from
dict the evolution of aerosol constituents. The regional specific emission the NASA Socioeconomic Data and Application Center (http://sedac.
inventory i.e. the Speciated multipOllutant Generator for India (SMOG- ciesin.columbia.edu/), based on the 2011 India census. To calculate
India (Pandey and Venkataraman, 2014; Sadavarte and Venkataraman, the age specific gridded population, we used the percentage distribution
2014), was obtained over the Indian region and merged with global of every state from Carter, 2015 (https://censusindia.gov.in/vital_sta
emissions obtained from an open-source Community Emissions Data tistics/SRS_Reports_2015.html). The cause-specific mortality rates
System (CEDS) (Hoesly et al., 2017). We also considered the Global fire (MR) for the age-cause-sex group over India in 2015 were obtained from
emissions database (GFED) (Giglio et al., 2013) and trash burning in the IHME website (https://vizhub.healthdata.org/gbd-compare/india).
ventory (Wiedinmyer et al., 2014). It is prepared for the Indian Network We calculated the population attributable fraction (PAF) and premature
Project on Carbonaceous Aerosol Emissions, Source Apportionment and mortality described in Murray et al. (2020). TMREL (theoretical mini
Climate Impacts (COALESCE) (Venkataraman et al., 2020). It includes mum risk exposure level) assumed in this study is 5.47 μg/m3 based on
emissions from seven major species: black carbon (BC), organic carbon the past eleven years of PM2.5 concentrations (work still not published).
(OC), sulphur dioxide (SO2), oxides of nitrogen (NOx), ammonia (NH3),
Mineral matter (MM) and non-methane volatile organic compounds 2.3. Observational data for model evaluation
(NMVOC). The anthropogenic sectors were mapped and merged with
different emission inventories as listed in Table S2. The daily varying For model evaluation, Climate Research Unit (CRU) Temperature
anthropogenic emissions from all the sectors were input in the surface version 4 (CRUTEM4) observational monthly dataset for the surface air
layer except for the forest fires and energy production sectors, which temperature for 2015 with 0.5◦ × 0.5◦ (~50 km) resolution is used. It
were distributed vertically in the model. Emissions for anthropogenic includes various station records and homogenized station data (Harris
NMVOC species were created from the 400+ molecular species for the et al., 2014). The planetary boundary layer (PBL) height and relative
18 source-sector activities and were speciated based on the RADM2 humidity (RH) variables were obtained from the global atmospheric
chemical mechanism (Carter, 2015). Biogenic Volatile Organic Com reanalysis ERA-Interim produced by the ECMWF with a horizontal res
pounds (BVOC) are important emissions as they account for ~90% of olution of about 80 km and at a daily temporal scale. A global level-3
global VOC emissions (Guenther et al., 1994). Therefore, we simulated daily gridded aerosol optical depth (AOD) dataset was obtained from
these emissions using v2.04 of the Model of Emissions of Gases and Moderate Resolution Imaging Spectroradiometer (MODIS, MYD08_D3
Aerosols from Nature (MEGAN) (Guenther et al., 2006) which is a product) at 500 nm with 1o X 1o spatial resolution which includes both
standard component of v3.3 WRF-Chem. Sea salt and dust emissions Terra (approximate local overpass time at about 10.30 local solar time
(Chin et al., 2002; Ginoux et al., 2001) were online parameterised as a (LST)) and Aqua (overpass at 13.30 LST) satellites. The ground-based
function of prevalent wind speed at 10 m for water and non-urban land observations of AOD were obtained from Aerosol Robotic Network
surfaces with sparse vegetation, respectively. (AERONET) at Kanpur (26.5◦ N, 80.2◦ E), Jaipur (26.9◦ N, 75.8◦ E) and
Gandhi College (25.9◦ N, 84.1◦ E) for model evaluation. A new and
2.1.2. Dust boundary condition mapping improved level 3 global aerosol optical depth dataset is used from a
In this study, we opted for dust opt = 3, which is the Air Force Multi-angle imaging SpectroRadiometer (MISR, MIL3MAE_4). It con
Weather Agency (AFWA) Georgia Institute of Technology–Goddard tains a statistical summary of column aerosol 555 nm optical depth, and
3
a monthly aerosol compositional type frequency histogram. This data Fig. 1 shows the evaluation of 2m mean temperature (T-2m), surface
product is a global summary of relevant Level 2 aerosol parameters, reaching shortwave radiation (SSR), PBL and relative humidity (RH).
averaged over a month and reported on a geographic grid, with a res Overall, the meteorological variables are well simulated by the model.
olution of 0.5◦ by 0.5◦ . From the Clouds and Earth’s Radiant Energy The modelled T-2m showed better agreement with observations among
Systems (CERES) energy balanced and filled (EBAF Ed4.1) level 3 all the variables. The seasonal variation of T-2m is seen to have been
product gridded at 1o X 1o, global monthly mean Top-of-Atmosphere captured by the model in comparison with CRU, where the mean ab
(TOA) and surface (SFC) longwave (LW), shortwave (SW), and net solute magnitude in months of July, August, September, and October is
(NET) fluxes under clear and all-sky conditions were obtained at underestimated by approximately − 0.4 ◦ C and overestimates in other
monthly resolution for 2015. Surface PM2.5 concentrations were months ranging from 0.2 to 0.6 ◦ C. Also, the modelled annual spatial
downloaded for various cities from 2013 to 2017 from the Central distribution of T-2m is well matched, with a correlation coefficient (R) of
Pollution Control Board (CPCB) regulated by the Ministry of Environ 0.9 and a normalised mean bias (NMB) value of 0.3, and a root mean
ment and Forests (https://cpcb.nic.in/air-quality-data/and https://cpc squared error (RMSE) value of 4 οC. Many modelling studies have shown
b.nic.in/namp-data/). PM2.5 mass concentration is measured by using similar agreement with observations related to temperature (Li et al.,
the tapered element oscillating microbalance (TEOM), which uses a 2020; Nguyen et al., 2019a, 2019b; Wang et al., 2019b). For radiation,
small vibrating glass tube whose frequency changes with the mass of the model overestimated the observational values of CERES. Though the
PM2.5 deposited on it. For the 2015 year, the CPCB measurement in model well captured the seasonal cycle of the SSR, the overestimation of
cludes very few locations and may not provide robust validation with the values is found every month except for the monsoon period. The
model output. Therefore, to increase the robustness of the model eval highest amount of solar incoming radiation was observed in the summer
uation for surface PM2.5 concentration by including more stations, a season and the lowest in the monsoon season in pertinence to the
few adjacent years were considered, which was also done in a previous presence of cloud cover obstructing sunlight during this time. The
study (Conibear et al., 2018). We gathered data from 75 unique loca modelled annual spatial distribution of SSR captured the north-south
tions, where a location is representative of the average of all the mea spatial gradient, with NMB value of − 0.3 and RMSE value of 36
surement sites. The chemical species composition of 10 different W/m2. The lower prediction of cloud fraction by the model consequently
locations were collected from different literature (Gawhane et al., 2019; leads to an overestimation of SSR (Fig. S2). The solar radiation flux
Pipalatkar et al., 2014; Ram et al., 2012; Rastogi et al., 2014; Samiksha dataset from AERONET showed satisfactory agreement (NMB = +7%)
et al., 2017; Singh et al., 2017). We also evaluated model PM1 (partic compared to the BASE run over the Kanpur site for January to July 2015
ulate matter of diameter less than 1 μm) and speciated concentrations (Fig. S3). In the case of PBL height, the seasonal variation was emulated
with the South West Asian Aerosol Monsoon Interaction (SWAAMI) well by the model compared to ERA-INTERIM values. The higher values
project aircraft dataset over the Indian subcontinent for June–July 2016 of PBL in summer allow additional vertical mixing of pollutants,
(Brooks et al., 2019). The campaign consists of 17 flight tracks and whereas shallower PBL in winter entraps more pollutants near the sur
onboard instruments like an aerosol mass spectrometer (AMS) and face. The model underestimated values of PBL for all the months except
Single Particle Soot Photometer (SP2). for the winter months. Also, the annual spatial distribution was captured
well, with NMB values of − 0.4 and RMSE values of 350 m. The model
3. Results and discussion underestimated the spatial and monthly values of RH compared to the
ERA-INTERIM values. The monthly evolution pattern of modelled RH
The results presented include a comprehensive model evaluation values resembles that of ERA-INTERIM, where the highest RH values are
against observation datasets of re-analysis meteorology and PM2.5 and observed in the monsoon and the lowest in the summer month of May.
chemical species from the ground station and aircraft measurements. The model captured the annual spatial gradient of RH, with NMB values
Further, we investigate the mechanism of aerosol radiation interaction of − 0.3 and RMSE values of 17%. Such underestimation in RH values
by comparing the differences between BASE and NoARI for annual and can impact the model’s aerosol formation and evolution process. Thus,
different seasons. The seasons included summer (March-April-May), as for an overall observation, the model can capture the meteorological
monsoon (June-July-August), post-monsoon (September-October- parameters such as T-2m, SSR, PBL and RH.
November) and winter (December-January-February). We also discuss
its implication on premature mortality due to exposure to PM2.5 in each 3.1.2. Evaluation of aerosol optical depth (AOD)
case. We evaluated Aerosol optical depth (AOD), a useful satellite-derived
variable available in both temporal and spatial resolutions from satellite
3.1. Model evaluation observation, to understand the simulation of aerosol columnar abun
dance, as seen in Fig. 2. The model underestimated the seasonal cycle of
3.1.1. Meteorology evaluation AOD when compared with the MODIS and MISR satellite observations.
Atmospheric aerosols are influenced by meteorological conditions, The simulated monthly mean AOD values were closer to MISR than
including surface reaching radiation, temperature and PBL, which MODIS satellite observations. Among different months, a large under
govern atmospheric physical and chemical processes. Both the temper estimation (~40–50%) in mean AOD values was found in the monsoon
ature and solar irradiance determine the photolysis rate and influence months (JJA). The annual spatial distribution was consistent with the
many atmospheric chemical reactions like O3 or NO2 production. The Indo-Gangetic plain (IGP) and the northern plain, capturing the highest
atmospheric vertical temperature profile determines the vertical diffu values of AOD synonymous with the satellite observations. Studies have
sion intensity of an air parcel. Wind speed and direction determine the attributed this highest AOD value to long-ranged dust transport from the
horizontal transportation and vertical mixing of chemical species. Arabian and the Thar deserts, densely populated regions and industri
Planetary boundary layer (PBL) height dictates dilution and dispersion alization (Das et al., 2015; Pan et al., 2015). However, the MISR satellite
of pollutants, and even cloud formation. Thus, the wind and temperature values were most coherent with model simulated values (R = 0.7, NMB
profiles, along with the boundary layer height are the main factors = − 0.4) compared with MODIS satellite values (R = 0.6, NMB = 0.5). A
affecting the turbulent diffusion of pollutants. Thus, for accurate esti similar agreement of AOD values has been found in various modelling
mation of aerosol distribution in the atmosphere, the model should be studies over India (David et al., 2018; Ganguly et al., 2012; Patil et al.,
able to simulate these meteorological parameters correctly. Herein, the 2019; Sharma et al., 2022). Studies have pointed out probable reasons
differences in the temporal and spatial variation of aerosol and meteo underlying such underestimations. The online dust parameterization in
rological variables between simulated and observational/reanalysis the model can produce an inadequate amount of regional flux; for
datasets at the same grid level have been carried out and evaluated. example, a recent study by Bhattacharya et al. (2022) applied a different
4
Fig. 1. Monthly evaluation over Indian landmass of observations and model (BASE): (a1) 2-m mean temperature (T-2m, oC), (b1) surface shortwave radiation (SSR,
W/m2), (c1) planetary boundary layer height (PBL, m) and (d1) relative humidity (RH, %). Standard deviation represents the variability across space are shown for
both observations and the model. Annual mean spatial distribution: (a2, a3) T-2m from CRU and BASE; (b2, b3) SSR from CERES and BASE; (c2, c3) PBL from ERA-
INTERIM and BASE and (d2,d3) relative humidity from ERA-INTERIM, and BASE.
correction factor to the Indian region to improve dust emission fluxes. mainly underestimates the AERONET values and is quite comparable
David et al. (2018) applied dust correction in size bins 5, 6 and 7 to with the satellite observations for the exact days of AERONET mea
improve the spatial pattern of AOD over India. Sanap et al. (2014) found surement. Studies have also shown that aerosol retrievals from satellites
the considerable biases in simulating the 850 hPa wind field (which mostly have large negative biases relative to AERNOET data over highly
plays an important role in the transport of dust from adjacent deserts) polluted sites in India and China (David et al., 2018; Mangla et al., 2020;
would be the possible reason for the poor representation of dust AOD Shi et al., 2019) and other world stations (Green et al., 2009, Almazroui,
and, in turn, total AOD over the Indian region in CMIP5 models. The 2019). The polluted sites in the urban region like Kanpur and Jaipur
different chemical mechanisms can lead to different AOD estimations, as showed consistency between AERONET and modelled AOD. Whereas a
they are significantly influenced by the abundances of secondary aerosol rural site at Gandhi College showed a larger bias than Kanpur, even
formations like sulphate and nitrate (Balzarini et al., 2015). We also when they are located in the same state. Since the Gandhi College site is
compared the ground observations of AERONET with the modelled as located downwind of populated urban cities like Delhi, Kanpur, and
well as satellite-derived AOD values (Fig. S4). In general, the model Lucknow, it experiences a mixture of urban and rural aerosol emissions
5
Fig. 2. For aerosol optical depth (AOD): (a1) monthly evaluation over Indian landmass of observations and model (BASE), (a2-a4) annual mean spatial distribution
of AOD from MODIS, MISR and BASE simulation.
(Ansari and Ramachandran, 2022), which is captured by AERONET. optimised model set-up using the SMoG-India inventory and meteoro
However, the model fails to capture the local rural aerosol loading at logical data assimilation shows surface PM2.5 concentration values
Gandhi College due to a lack of details in the emission inventory. within an NMB of − 0.10, which is a significant improvement over pre
Anthropogenic emissions from the regionally representative SMoG-India vious studies (Bran and Srivastava, 2017; Jena et al., 2020; Karambelas
emission inventory were earlier found to improve AOD simulation et al., 2018; Ojha et al., 2020). The seasonal NMB values are relatively
substantially (David et al., 2018) in a different model GEOS-Chem. Yet more extensive, ranging from − 0.02 to − 0.30. In observations, the
the continuing underestimation of AOD in this simulation points to the spatial distribution of modelled annual averaged surface PM2.5 con
likely underestimation of online calculated dust emission, relative hu centration was satisfactorily captured with a north-south and west-east
midity, and the parameterization of aerosol water, which governs par gradient. The model simulated the highest values of surface PM2.5
ticle size and AOD. concentrations in the northern IGP region with an annual mean as high
as 180 μg/m3. However, this region had a considerable underestimation
3.1.3. Evalution of PM and chemical constituents (NMB = − 0.5 to 0.6) compared to observations across the country,
The evaluation of PM2.5 and its chemical composition in India is indicating a likely persisting underestimation of emissions of primary
impeded by the paucity of available data, resulting in a focused evalu PM2.5 constituents or its precursors in this region. The annual surface
ation of a single pollutant or a single/few locations (Bran and Srivastava, PM2.5 concentration in the rest of India showed relatively better
2017; Conibear et al., 2018; Jena et al., 2020; Karambelas et al., 2018; agreement with biases of NMB < − 0.3.
Ojha et al., 2020). This study extensively evaluated modelled particulate As mentioned earlier, aerosol chemical species were evaluated
matter (PM) and its chemical species composition at the surface and against both ground-station measurements and aircraft observations
altitude profiles (Fig. 3). Comparison between the ground-station ob from the SWAAMI campaign. Among different chemical species, SO4,
servations from 75 locations across India averaged annually with an nitrate (NO3), ammonium (NH4) were predicted by the model within
Fig. 3. (a) Seasonal and annual comparison of model skill for the base simulation relative to surface PM2.5 CPCB observations between 2015 and 2017 using the
normalised mean bias (NMB). Seasons are defined as Winter (DJF), Post-monsoon (SON), Monsoon (JJA) and Summer (MAM). (b) Annual surface PM2.5 concen
tration from BASE simulation compared with CPCB annual PM2.5 concentration (represented in circles) along with comparison of species composition at 10 different
sites over India (represented as bar graphs). (c) The difference in the modelled BASE simulation and observed PM1 aerosol concentrations from the SWAAMI flight
tracks in June–July 2016. The chemical species obtained from SWAAMI flight are compared with model simulations for 5 different regions (outlined in different
colors). The observations are represented as circles and model outputs are represented as bar in the figure.
6
NMB of − 0.7 across 10 ground-stations, indicating reasonable model measurements (Conibear et al., 2018; GBD, 2018). We believe mini
skill in capturing spatial differences in atmospheric transport and mising the uncertainties by choosing better representative regional
chemistry. Many atmospheric models overpredict NO3 concentrations emissions and corrections to dust boundary mapping (refer to section
(Fairlie et al., 2010; GBD, 2018) and a few possible reasons for such 2.1.2) helped our model simulation predict particulate pollutants better.
overestimation have been identified in the literature as uncertainties in
NH3 and NOx emissions in terms of both sectoral disaggregation and
3.2. Changes induced by aerosol radiative feedback
temporal scales, larger availability of NH4 concentrations to react and
form nitrates, misrepresentation in the model mechanism such as
In the case of aerosol-radiation interaction, it reduces shortwave
overestimation of N2O5 hydrolysis in aerosols and underestimation of
radiation reaching the surface, resulting in a decrease in surface tem
the dry deposition of HNO3 (GBD, 2018; Heald et al., 2012; Zhang et al.,
perature. However, the increase in mid-atmospheric temperature cre
2012). However, a larger underestimation was found in OC (NMB – − 0.2
ates temperature inversion to enhance in ARI scenario leading to
nationally and − 0.6 regionally). The current model chemistry
atmospheric stability and lower planetary boundary height (PBL).
(RADM2-SORGAM) did not satisfactorily simulate secondary organic
Hence, the vertical dispersion of pollutants is reduced. Also, horizontal
aerosol (SOA) with extremely low concentrations across the region.
transportation is reduced because of the reduction in wind speed,
Sectors with large uncertainty in emissions, like residential biomass fuel
leading to further accumulation of PM2.5 and increased concentration.
use, agricultural residue burning and traditional informal industries like
brick production, are all the largest emitters of OC. There is likely to be a
3.2.1. Meteorological changes
more considerable persisting uncertainty in OC emission, which is being
As discussed before, aerosols can directly impact meteorological
evaluated in a concomitant inventory development study (Venkatara
variables like surface reaching shortwave radiation (SSR), 2-m temper
man et al., 2020).
ature (T-2m) and PBL height. The model showed a reduction in the
We also evaluated modelled PM1 at different flight track altitudes
annual and seasonal averages of daily values for all these variables due
obtained from SWAAMI aircraft measurements during the June–July
to the ARI effect (Fig. 4). The mean reduction over India’s landmass in
period of 2016 (Fig. 3c). The difference between modelled values and
SSR, T-2m and PBL were − 15 W/m2 (− 8%), − 0.21 ◦ C (− 0.83%), and
ones from SWAAMI ranged from − 5 to +9 μg/m3. The model clearly
− 48 m (− 10%) respectively. The northern and IGP regions show the
showed underestimation in the northwest region and central west India
maximum reduction in SSR, T-2m and PBL of 15%, 2% and 12%,
and a mixed response in the rest of the regions. The other chemical
respectively. Among all seasons, the summer season experienced the
species modelled were approximately similar to that from SWAAMI.
highest mean reduction in SSR, T-2m and PBL of − 25 W/m2 (− 10%),
However, the magnitude of simulated OC was largely underpredicted
− 0.25 ◦ C (− 1%), and − 100 m (− 13%), respectively. The post-monsoon
when compared with other chemical species. Such differences in the
and winter seasons show a similar reduction, whereas the monsoon
output simulated by the model can influence the understanding of the
season shows the lowest reduction in these met variables. The spatial
magnitude of the regional pollution levels.
pattern of increase or decrease found in T-2m looks similar to PBL
Previous regional modelling studies have shown a relatively higher
because PBL is influenced by its contact with the planetary surface. The
underestimation of PM2.5 and its constituents over India, with NMB
depth of the PBL depends on the surface temperature and terrain;
values reaching up to − 0.7 (Bran and Srivastava, 2017; Jat et al., 2021;
therefore, it varies with the seasons and is found to be higher over land,
Jena et al., 2020; Karambelas et al., 2018; Ojha et al., 2020). Some
especially in the desert where thermal mixing is prominent (Fig. S5). For
studies have shown similar underestimation with NMB values < − 0.2;
precipitation, ARI causes a mixed behaviour regionally during all sea
however, they are compared against a significantly smaller database of
sons, except for monsoon. The changes in precipitation showed an
Fig. 4. Spatial distribution due to aerosol radiation interaction (BASE - NoARI) to meteorological variables: (a1-a5) surface shortwave radiation (W/m2), (b1-b5)
temperature at 2m (oC), and (c1-c5) PBLH (m) for annual and seasonal mean of 2015.
7
overall increase in northern India and a decrease in southern India concentrations resulting from ARI were about 3 μg/m3 (6%), which
during the monsoon (Fig. S6). These changes concurred with the ARI- appears to be a modest increase over the mean PM2.5 levels. Larger
induced changes in surface temperature over northern India i.e., enhancements in surface PM2.5 concentrations occurred seasonally in
increased precipitation would lead to a reduction in surface temperature summer (6 μg/m3 - 15%) and post-monsoon (10 μg/m3 - 10%), but were
(Fig. 4b). minimum in monsoon season (1 μg/m3 - 3%) as expected. Regionally,
We examined associations between the changes in meteorological larger changes occurred in the north-western region, where the highest
variables, which were annually averaged across all grid cells over India annual average change reached up to 40% with PM2.5 enhancement of
(Fig. S7). A higher value of a positive or a negative correlation coeffi 30 μg/m3. This region co-occurs with the location where the PM2.5
cient emphasises their strong relationships with each other. For surface concentration is largest and is also dominated by primary con
example, the decrease in SSR strongly impacts the decreases in T-2m, or centrations of BC, OC and dust (Fig. S8). Fig. S7 suggests the potential
the decrease in T-2m strongly influences PBL. Similar reductions were reasons for such enhancements, where Pearson’s correlation coefficient
found in earlier studies (Hong et al., 2017b; Nguyen et al., 2019a, between the changes in PM2.5 and meteorological variables has strong
2019b; Sekiguchi et al., 2018; Wang et al., 2016a; Xing et al., 2017). negative correlations. It is suggestive of the reduction in downwelling
solar radiation causing cooler surface temperatures which in turn
3.2.2. Changes in surface PM2.5 concentrations weakened the thermal turbulence resulting in thinning of PBL due to
The reductions observed in meteorological variables over India ARI. This further caused stabilisation in the atmosphere and further
coincide with the increases observed in PM2.5 concentrations (Fig. 5). enhanced surface PM2.5 concentrations (Hong et al., 2017b; Nguyen
For 2015, the annual average PM2.5 concentrations were found to be 60 et al., 2019a, 2019b; Sekiguchi et al., 2018; Wang et al., 2016a; Xing
μg/m3 over the Indian landmass. Among all seasons, wintertime surface et al., 2017). Sub-regional level changes in annual average PM2.5 surface
PM2.5 concentrations were the highest, reaching 160 μg/m3 over the concentrations from ARI feedbacks found in India were greater than
northern and IGP regions (Fig. S8). In winter, the vertical mixing of the those found in other regions of the world (Forkel et al., 2015b; Wang
pollutant is inhibited by meteorological conditions such as shallower et al., 2014, 2016b), but were found to be similar in the case of the haze
PBL heights (Fig. S5), steeper temperature lapse rate, calmer wind episodes (Li et al., 2020). This finding indicates the importance of ARI in
conditions and greater abundances of anthropogenic aerosols from enhancing PM2.5 in regions of high aerosol loading, especially in Indian
biomass burning activities (Ghude et al., 2017; Mehta, 2015). Averaged regions. It is worth noting that even when the aerosol loading is low
spatially over India, enhancements in the annual surface PM2.5 during monsoon season (Fig. S9), the ARI effect still occurs with
Fig. 5. Spatial distribution due to aerosol radiation interaction (BASE - NoARI) to surface concentrations of: (a1-a5) PM2.5, (b1-b5) black carbon (BC), (c1-c5) organic
carbon (OC), and (d1-d5) dust, for annual and seasonal mean of 2015.
8
minuscule impacts. The model has shown a significant underestimation annual average, SO4 concentration decreased by 10%, with a maximum
in AOD values which interacts with radiative transfer models for esti reduction in the monsoon season of 30%. The reduction is mainly due to
mating radiative fluxes. Therefore, it could be implied that better the inhibited aqueous phase and gas-phase reactions of SO2 to form
simulation of AOD would result in even larger increases in PM2.5 surface sulphate. Aqueous phase reactions between SO2 and H2O2 or SO2 and O3
concentrations from ARI feedbacks in the Indian region. are the major pathways for sulphate formation, although the gas-phase
reaction of SO2 with OH is also important (Erikson, 1972; Harris et al.,
3.2.3. Changes in surface aerosol chemical composition 2012; Tanaka et al., 1994; Zhang et al., 2018). The findings suggest that
Different regions of India contribute differently to primary and sec in summer, the gas-phase reaction can be dominant as the spatial dis
ondary aerosol concentrations annually and across all seasons. The tribution of SO4 resembles that of OH and whereas in winter, the
modelled annual average contribution of primary aerosol constituents to aqueous phase reaction may be prevalent as the spatial distribution of
total PM2.5 concentrations was found to be approximately 60%, 40% SO4 and O3 looks alike (Fig. S13). Gas-phase oxidation of SO2 by OH
and 50% over north-northwest India, eastern India and southern India, occurs in summer, whereas aqueous-phase oxidation of SO2 by O3 and
respectively. The spatial distribution of BC, OC, NH4 and NO3 concen H2O2 occurs in winter. Atmospheric low temperatures and high relative
trations dominated over the north and IGP region, whereas dust and SO4 humidity or weak solar radiation in the atmosphere make a conducive
concentrations were majorly found over the north and north-western environment for aqueous phase oxidation of SO2 in aerosols (Liang and
region and the eastern region, respectively (Fig. S10). We also found Jacobson, 1999). Our study showed higher relative humidity in winter
that due to ARI, the primary aerosol concentrations increased all over when compared to summer (Fig. S14), which supports the earlier find
India (Fig. 5). The major percentage enhancements were found due to ings from the literature of a more probable oxidation pathway through
positive changes in dust concentrations reaching up to 8 μg/m3 (40%) an aqueous phase reaction to occur in winter. The increase in SO4
over the north-western region. The increased surface PM2.5 concentra concentrations observed in post-monsoon may be due to the dominance
tions due to ARI resulted from increased primary aerosol concentrations of both gas-phase and aqueous-phase reactions, resulting in the further
(Fig. 5). The strong positive correlation found between surface PM2.5 conversion of sulphur dioxide to sulphate (Fig. S13). The increases in
concentration and primary components (Fig. S7) suggests that the in oxidants concentrations during post-monsoon can be due to reduced
crease in PM2.5 is directly dependent on the increases found in BC, OC planetary boundary layer height.
and dust. In the model simulation, BC and OC emissions are held con Due to the ARI, the change in nitrate and ammonium has shown
stant, whereas the dust emissions are parameterised online. The similar regional and seasonal differences (Fig. 6). The annual average
increased dust concentrations were due to increased local dust emissions has shown an increase of 1.3 μg/m3 (+8.5%) in NO−3 and 0.3 μg/m3
in all seasons due to the ARI effect (Fig. S11). Also, we found an (+6%) in NH+ 4 . The highest increases in NH4 and NO3 were found in the
+ −
increased dust advection from the nearby countries, especially in the post-monsoon and winter seasons compared to the summer and
summer and monsoon seasons (Fig. S12), which caused enhanced dust monsoon seasons. The model includes two pathways for nitrate forma
concentrations. tion (NOx to nitric acid HNO3 oxidation): the NO2 oxidation by OH
The secondary aerosols formation behaves differently compared to during the daytime and the hydrolysis of dinitrogen pentoxide (N2O5) at
the primary aerosols; therefore, the effect of ARI on sulphate, ammo night (Archer-Nicholls et al., 2016). During winter, low temperatures
nium and nitrate was investigated through their formation pathways. and high RH favour the formation of NO−3 and NH+ 4 (Fig. 4, Fig. S14).
Given the same meteorological conditions, sulphate decreases annually High temperatures and low relative humidity conditions cause a shift of
and in all seasons except post-monsoon due to the ARI (Fig. 6). On an equilibrium from volatile particulate ammonium nitrate to ammonia
Fig. 6. Spatial distribution due to aerosol radiation interaction (BASE - NoARI) to secondary aerosol concentrations: (a1-a5) sulphate (SO4), (b1-b5) Ammonium
(NH4), (c1-c5) Nitrate (NO3), for annual and seasonal mean of 2015.
9
and nitric acid (HNO3), resulting in their high abundances during shortwave radiation and temperature, modulate the photochemistry and
summer than in winter (Fig. S16). The similar increases/decreases of oxidation rate and second, the lower PBL increases stability and mod
NH+ 4 and NO3 due to ARI is established by their high positive correlation
−
ulates abundances of precursor concentration and consequently reaction
between them (Fig. S7), and their formation largely depends on rates (Xing et al., 2015). The decreased O3 concentrations could be
ammonia. attributed to the increased NOx titration due to enhanced atmospheric
Overall, the change in annual surface PM2.5 concentration due to ARI stability and reduced PBL height (Fig. 4). The reduction in the photolysis
comes majorly from the contribution of primary aerosol changes (up to rate due to the decreased surface shortwave radiation caused by the ARI
80% over the north-western region) rather than secondary aerosol resulted in a decreased O3 concentration. The increased O3 concentra
changes. It may be inferred that the ARI causes changes to primary tion in the summer and post-monsoon seasons could be associated with
aerosol concentrations, which are influenced by the changes in meteo the lower PBL height that increased the impact of entrainment of O3
rological parameters. In India, the anthropogenic primary aerosol transport from upwind polluted areas. The increases in NO2 and SO2
emissions are dominant due to fossil and biomass-burning activities were found annually and seasonally due to ARI (Fig. 7). The annual
(Venkataraman et al., 2005), and the dust comes from long-range spatial average increases found were +0.2 ppb (+3.4%) in NO2 and
transport from South Africa or local sources (Dey et al., 2004; Pan +0.1 ppb (+1.7%) in SO2 concentrations. The largest seasonal en
dithurai et al., 2008; Sijikumar et al., 2016). Therefore, if the primary hancements for NO2 and SO2 occur in summer and post-monsoon, with
aerosol emissions continue to increase, then ARI’s influence will further changes up to +1.2 ppb (4.8%) and +0.9 ppb (13%), respectively, over
increase surface PM2.5 concentrations and degrade air quality (Xing northern India. In the post-monsoon season, the regions with high SO2
et al., 2016). concentrations undergo oxidation by high O3 concentrations to form
sulphate. Similar changes in ozone and its precursors’ concentration due
3.2.4. Changes in surface ozone concentrations to ARI are also reported in other studies where percentage annual
The O3 concentration ranged from 27 to 62 ppb during the year, with average changes are between − 0.92% and − 1.56% (Nguyen et al.,
an average value of 37 ppb in the BASE simulation (Fig. S15). In contrast 2019a, 2019b), which is similar to this study.
with the uniform enhancements in surface PM2.5 concentrations, the
surface O3 concentrations showed a mix of responses due to ARI (Fig. 7).
Averaged spatially over India, reductions in the annual averages of 3.3. Changes in premature mortality attributable to PM2.5 exposure
surface O3 concentrations resulting from ARI were about − 0.5 ppb
(− 1.4%). Larger enhancements in surface O3 concentrations occurred Apart from being a highly populated country, India also experiences
seasonally, in summer by 10 ppb (20%) and post-monsoon by 7 ppb severe air pollution concentrations when compared to other world na
(15%), but reductions were at their largest in the monsoon season by 15 tions. The co-occurrence of these two factors can cause significant im
ppb (− 35%) and in winter by 6 ppb (− 10%). Regionally, increases in pacts on mortality due to ARI’s enhancement effect on
surface O3 concentrations occurred in the north-western Indian region PM2.5concentrations. In this section, we explore the differences in esti
annually and seasonally, where the highest annual average change mating premature mortality due to differences in exposure to higher
reached up to 2.5% with an O3 enhancement of 1.2 ppb. These changes PM2.5 concentrations induced by accounting for ARI.
in surface O3 concentrations as a response to ARI competes between two Fig. 8a shows the population-weighted PM2.5 concentrations in
effects: first, changes in meteorological conditions, such as a decrease in different states of India from the BASE simulation. The Northern Indian
states are exposed to the annual population-weighted PM2.5
Fig. 7. Spatial distribution due to aerosol radiation interaction (BASE - NoARI) to ozone and its precursors: (a1-a5) ozone (O3); (b1-b5) nitrogen dioxide (NO2), (c1-
c5) sulphur dioxide (SO2), for annual and seasonal mean of 2015.
10
Fig. 8. Annual spatial average plot for 2015: (a) population weighted PM2.5 concentration in BASE simulation; (b) total premature deaths in BASE simulation, (c)
difference of population weighted PM2.5 concentration between BASE and NoARI simulation and (d) difference of premature deaths between BASE and NoARI
simulation. The percentage of deaths due to ARI are written for first five top states. The pie chart in the bottom figure represent the contribution from the six causes.
concentrations reaching 120 μg/m3, IGP states reaching 70 μg/m3, and chemical transport models (Conibear et al., 2018). Across India, the
the rest of the Indian states experience exposure lower than 70 μg/m3. population-weighted PM2.5 ranges from 1 to 12 μg/m3 due to ARI
The annual total premature deaths in Indian states due to exposure to (Fig. 8c). The highest population-weighted PM2.5 concentrations occur
PM2.5 concentrations in the BASE scenario are shown in Fig. 8b. The in North-western states like Punjab, Haryana, Rajasthan, and Gujarat,
total number of deaths in India due to accounted causes is approximately where changes in primary aerosol concentrations like BC, OC and dust
6 lakhs (95% uncertainty level 4.5–7.6 lakh). The top Indian states with have been highest due to ARI. A large fraction of primary PM results
the highest number of deaths are Uttar Pradesh, Bihar, Maharashtra and from anthropogenic activity, which causes enhancements in ambient PM
West Bengal, which are also highly populous states. The estimated (and consequently associated higher aerosol radiative effects) and also
premature deaths in the BASE case are underestimated by approxi co-occurs in high and moderately high populous regions (Fig. S17).
mately 1 lakh compared to another study, which used the same method These results are consistent with other world regions like East Asia,
(Pandey et al., 2021). This difference in the number of deaths can be due North America and Europe, where greater enhancements of surface
to the underestimation of modelled PM2.5 concentrations. Other PM2.5 due to the incorporation of aerosol radiative interactions also
modelling studies reported a varying estimation of premature mortality collocate with higher population density regions (Xing et al., 2016). The
attributable to PM2.5 in India i.e., 0.6 million in 2010–2011 (Lelieveld differences in the exposure of higher surface PM2.5 concentrations
et al., 2015; Ghude et al., 2016) and 0.8 million in 2013 (Gao et al., induced by accounting for ARI can cause differences in the estimated
2018) and ~1.2 million in 2017 (Dicker et al., 2018). The differences in premature mortality of approximately 26 thousand (95% uncertainty
these reported values are partly due to variations in ambient PM2.5 level 15–39 thousand) (Fig. 8d). The diseases such as COPD and IHD
across years, in addition to differences in adopted health functions and contribute to ~80% of the total difference in the estimated pre-mature
11
deaths (bottom pie graph in Fig. 8d). The states with the highest number stabilisation induced by meteorological variables and primary aerosol
of pre-mature deaths are Rajasthan, Uttar Pradesh, Gujarat, Madhya concentration increases.
Pradesh and Maharashtra, comprising ~70% of total deaths due to the There are several limitations in this work, which could be addressed
differences in exposure of surface PM2.5 concentrations on accounting in future studies. The study uses a pair of single-year simulations, while
only the ARI effects. further insights could be obtained from multi-year simulations, both in
Here we discuss the implications of regional vulnerability to ARI terms of understanding inter-annual variability and the complex inter
impacts and the importance of including ARI when simulating with air action between air pollution and radiative feedback. The chemical
quality models. The Indian states have been classified into four levels of mechanism implemented (RADM2-SORGAM) resulted in negligibly low
“high”, “moderately high”, “moderate”, and “low” under the three cat SOA formation over India, which requires further investigation, with
egories of (i) population, differences in (ii) exposure to PM2.5 concen evaluation among chemical mechanisms. The possible ways of
tration and (iii) differences in the estimation of premature mortality improving SOA formation in SORGAM are to consider the inclusion of
(Fig. S17). We analysed the relationship among these levels across (a) oxidation reactions of biogenic monoterpenes (McKeen et al., 2007;
different categories for each state (described in Table S5). The results Tuccella et al., 2012) and isoprene (Azmi et al., 2022); (b) the interac
show that even a “moderately highly” populated state, when exposed to tion between NMVOC and the liquid phase aerosols (Volkamer et al.,
changes in PM2.5, even up to 12 μg/m3, can cause a “high” number of 2007); (c) additional SOA precursors such as volatile primary organic
deaths. Prior studies examining historical responses due to aerosol aerosol (POA), semi-VOC and intermediate VOCs, and glyoxal (Zhang
radiative effects have shown exacerbated health impacts from the ARI- et al., 2016) and (d) explicit representation of monoterpenes and ses
causing enhancements in PM2.5 concentrations, especially in South quiterpenes. Also, the volatility basis set (VBS) SOA module has shown
Asian countries, and are considered important for future health assess significant improvement in simulating SOA closer to measurements
ments due to aerosols (Xing et al., 2016). If ARI is not represented well in (Farina et al., 2010; Ahmadov et al., 2012) as it treats the organic
air quality models, then particulate concentration values can vary gas/particle partitioning within a spectrum of volatilities using a satu
largely and restrict our understanding of air quality implications on ration vapour concentration as the surrogate of volatility (Donahue
health and the ecosystem, especially for highly polluted regions like et al., 2006; Ahmadov et al., 2012). The boundary conditions of mete
India. We believe that such differences can substantially affect the orological variables used are for the same base year of simulation, thus
implementation of environmental and health-related policies. capturing the impacts of a given “climate” state. Appropriate changes in
simulation set-up can attempt to disentangle the effects of emissions
4. Conclusions changes and climate feedback on future levels of air pollution.
We find that in the polluted north-western and IGP regions, aerosol
The responses of various meteorological and particulate pollution direct radiative effects significantly enhance surface-level particulate
variables to aerosol direct radiative effects were evaluated over the pollution and related premature mortality. This is of particular concern,
South Asia region, using the online coupled WRF Chem model, with a with expected future increases in regional emissions. These findings
pair of one-year simulations with and without ARI coupling to meteo suggest the importance of using fully coupled meteorology-chemistry
rology and chemistry. The model performance was statistically evalu models, rather than one-way meteorology-driven chemical transport
ated against reanalysis data and in-situ and satellite observations, models, to analyse air pollution in the Indian region to account for ARI
including spatial and seasonal variation. For meteorological variables, enhancements of surface air pollution and its effects.
the model was able to reproduce the observations competently. How
ever, the model underestimated PBL, RH and AOD and overestimated Funding
SSR. For PM2.5, the model performance was bound within NMB of − 0.1
compared with an extensive database of observations, indicating This work was supported by the NCAP-COALESCE project funded by
considerably better performance than what was reported in earlier the Ministry of Environment Forests and Climate Change, India [grant
studies for this region. This results from the combined benefits of a NO.14/10/2014-CC (Vol. II)]
regionally representative emission inventory optimised dust boundary
conditions and meteorological data assimilation. A larger NMB of − 0.02 CRediT authorship contribution statement
to − 0.27 seasonally of PM2.5 concentration is indicative of differences in
meteorological and chemical processes across seasons. Evaluation of Arushi Sharma: Methodology, Validation, Formal analysis, Inves
PM2.5 precursors and chemical constituents had greater variability, tigation, Data curation, Writing – original draft, Conceptualization,
indicating the larger influence of model processes. Overall, an under Writing – review & editing, Resources, Writing – review & editing.
estimation in AOD points to improvements needed in the model simu Chandra Venkataraman: Supervision, Project administration, Funding
lation of relative humidity and aerosol water, while an underestimation acquisition, Conceptualization, Writing – review & editing. Kaushik
of OC points to improvements needed in emission inventory and sec Muduchuru: Methodology, Investigation. Vikas Singh: Resources,
ondary organic aerosol estimation, especially over northern India. Writing – review & editing. Amit Kesarkar: Resources, Writing – review
The impact of aerosol direct radiative effects on meteorology and air & editing. Sudipta Ghosh: Resources, Writing – review & editing.
quality is clearly noticeable when comparing simulations with and Sagnik Dey: Resources, Writing – review & editing.
without ARI coupling to meteorology. Regarding meteorology, SSR, T-
2m, and PBLH showed a reduction due to ARI. Spatially and annually- Declaration of competing interest
averaged changes were − 8% for SSR, − 0.83% for T-2m, and − 10%
for PBLH, with the largest regional changes in northern and IGP; and the The authors declare that they have no known competing financial
largest seasonal changes in SSR and PBL variables during summer. interests or personal relationships that could have appeared to influence
Analysing the relationships among the responses of meteorological the work reported in this paper.
variables and pollutant concentrations indicated that a decrease in T-2m
and PBLH were highly associated with the decrease in SSR. For partic Data availability
ulate pollution, the concentration of PM2.5, SO2, and NO2 showed en
hancements, while O3 concentration mostly showed a reduction. Data will be made available on request.
Perhaps the most interesting finding was a spatial asymmetry with the
largest increase in particulate pollution induced in northwest India by
the ARI effect. The increase in PM2.5 concentration was highly related to
12
Acknowledgements Daly, A., Zannetti, P., 2007. Air Pollution Modeling – an Overview. The Arab School for
Science and Technology (ASST) and The EnviroComp Institute.
Das, S., Dey, S., Dash, S.K., Giuliani, G., Solmon, F., 2015. Dust aerosol feedback on the
We acknowledge the NCAP-COALESCE project funded by the Min Indian summer monsoon: sensitivity to absorption property. J Geophys Res
istry of Environment Forests and Climate Change through grant NO.14/ Atmospheres 120, 9642–9652. https://doi.org/10.1002/2015jd023589.
10/2014-CC (Vol. II). The views expressed in this document are solely David, L.M., Ravishankara, A.R., Kodros, J.K., Venkataraman, C., Sadavarte, P., Pierce, J.
R., Chaliyakunnel, S., Millet, D.B., 2018. Aerosol optical depth over India. J Geophys
those of the authors and do not necessarily reflect those of the Ministry. Res Atmospheres 123, 3688–3703. https://doi.org/10.1002/2017jd027719.
The Ministry does not endorse any products or commercial services Dee, D.P., Uppala, S.M., Simmons, A.J., Berrisford, P., Poli, P., Kobayashi, S., et al., 2011.
mentioned in this publication. We thank Prof. Hugh Coe for assisting The ERA-Interim reanalysis: configuration and performance of the data assimilation
system. Q. J. Roy. Meteorol. Soc. 137, 553–597. https://doi.org/10.1002/qj.828.
with the SWAAMI dataset collected during the joint MoES-NERC pro Dey, S., Tripathi, S.N., Singh, R.P., Holben, B.N., 2004. Influence of dust storms on the
gramme and funded through the NERC grant number NE/L013886/1. aerosol optical properties over the Indo-Gangetic basin. J Geophys Res Atmospheres.
We also thank Dr. Sourangsu Chowdhury, Dr. Kushal Tibrewal and Ms. https://doi.org/10.1029/2004jd004924, 1984 2012 109.
Dicker, D., Nguyen, G., Abate, D., Abate, et al., 2018. Global, regional, and national age-
Shashi Tiwari for helping with premature mortality calculation. sex-specific mortality and life expectancy, 1950–2017: a systematic analysis for the
Global Burden of Disease Study 2017. Lancet Lond Engl 392, 1684–1735. https://
doi.org/10.1016/s0140-6736(18)31891-9.
Appendix A. Supplementary data
Donahue, N.M., Robinson, A.L., Stanier, C.O., Pandis, S.N., 2006. Coupled partitioning,
dilution, and chemical aging of semivolatile organics. Environ. Sci. Technol. 40 (8),
Supplementary data to this article can be found online at https://doi. 2635–2643. https://doi.org/10.1021/es052297c.
org/10.1016/j.atmosenv.2023.119609. Emmons, L.K., Walters, S., Hess, P.G., Lamarque, J.-F., Pfister, G.G., Fillmore, D.,
Granier, C., Guenther, A., Kinnison, D., Laepple, T., Orlando, J., Tie, X., Tyndall, G.,
Wiedinmyer, C., Baughcum, S.L., Kloster, S., 2010. Description and evaluation of the
References model for ozone and related chemical Tracers, version 4 (MOZART-4). Geosci. Model
Dev. (GMD) 3, 43–67. https://doi.org/10.5194/gmd-3-43-2010.
Erikson, T.E., 1972. Sulfur isotope effects. Acta Chem. Scand. 26, 581–584.
Ackermann, I.J., Hass, H., Memmesheimer, M., Ebel, A., Binkowski, F.S., Shankar, U.,
Fairlie, T.D., Jacob, D.J., Dibb, J.E., Alexander, B., Avery, M.A., Donkelaar, A. van,
1998. Modal aerosol dynamics model for Europe development and first applications.
Zhang, L., 2010. Impact of mineral dust on nitrate, sulfate, and ozone in transpacific
Atmos. Environ. 32 https://doi.org/10.1016/S1352-2310(98)00006-5.
Asian pollution plumes. Atmos. Chem. Phys. 10, 3999–4012. https://doi.org/
Ahmadov, R., McKeen, S.A., Robinson, A.L., Bahreini, R., Middlebrook, A.M., De
10.5194/acp-10-3999-2010.
Gouw, J.A., Meagher, J., Hsie, E.Y., Edgerton, E., Shaw, S., Trainer, M., 2012.
Farina, S.C., Adams, P.J., Pandis, S.N., 2010. Modeling global secondary organic aerosol
A volatility basis set model for summertime secondary organic aerosols over the
formation and processing with the volatility basis set: implications for anthropogenic
eastern United States in 2006. J. Geophys. Res. Atmos. 117 (D6) https://doi.org/
secondary organic aerosol. J. Geophys. Res. Atmos. 115 (D9) https://doi.org/
10.1029/2011JD016831.
10.1029/2009JD013046.
Almazroui, M., 2019. A comparison study between AOD data from MODIS deep blue
Forkel, R., Balzarini, A., Baró, R., Bianconi, R., Curci, G., Jiménez-Guerrero, P., Hirtl, M.,
collections 51 and 06 and from AERONET over Saudi Arabia. Atmos. Res. 225,
Honzak, L., Lorenz, C., Im, U., Pérez, J.L., Pirovano, G., José, R.S., Tuccella, P.,
88–95. https://doi.org/10.1016/j.atmosres.2019.03.040.
Werhahn, J., Žabkar, R., 2015a. Analysis of the WRF-Chem contributions to AQMEII
Ansari, K., Ramachandran, S., 2022. Aerosol Characteristics over Indo-Gangetic Plain
phase2 with respect to aerosol radiative feedbacks on meteorology and pollutant
from Ground-Based AERONET and MERRA-2/CAMS Model Simulations.
distributions. Atmos. Environ. 115, 630–645. https://doi.org/10.1016/j.
Atmospheric Environment, 119434.
atmosenv.2014.10.056.
Archer-Nicholls, S., Lowe, D., Schultz, D.M., McFiggans, G., 2016.
Forkel, R., Balzarini, A., Baró, R., Bianconi, R., Curci, G., Jiménez-Guerrero, P., Hirtl, M.,
Aerosol–radiation–cloud interactions in a regional coupled model: the effects of
Honzak, L., Lorenz, C., Im, U., Pérez, J.L., Pirovano, G., José, R.S., Tuccella, P.,
convective parameterisation and resolution. Atmos. Chem. Phys. 16, 5573–5594.
Werhahn, J., Žabkar, R., 2015b. Analysis of the WRF-Chem contributions to AQMEII
https://doi.org/10.5194/acp-16-5573-2016.
phase2 with respect to aerosol radiative feedbacks on meteorology and pollutant
Atwater, M.A., 1970. Planetary albedo changes due to aerosols. Science 170, 64–66.
distributions. Atmos. Environ. 115 https://doi.org/10.1016/j.
https://doi.org/10.1126/science.170.3953.64.
atmosenv.2014.10.056.
Azmi, S., Sharma, M., Nagar, P.K., 2022. NMVOC emissions and their formation into
Forkel, R., Werhahn, J., Hansen, A.B., McKeen, S., Peckham, S., Grell, G., Suppan, P.,
secondary organic aerosols over India using WRF-Chem model. Atmos. Environ. 287,
2012. Effect of aerosol-radiation feedback on regional air quality – a case study with
119254 https://doi.org/10.1016/j.atmosenv.2022.119254.
WRF/Chem. Atmos. Environ. 53, 202–211. https://doi.org/10.1016/j.
Balzarini, A., Pirovano, G., Honzak, L., Žabkar, R., Curci, G., Forkel, R., Hirtl, M., José, R.
atmosenv.2011.10.009.
S., Tuccella, P., Grell, G.A., 2015. WRF-Chem model sensitivity to chemical
Ganguly, D., Rasch, P.J., Wang, H., Yoon, J., 2012. Climate response of the South Asian
mechanisms choice in reconstructing aerosol optical properties. Atmos. Environ. 115
monsoon system to anthropogenic aerosols. J. Geophys. Res. Atmospheres 1984.
https://doi.org/10.1016/j.atmosenv.2014.12.033.
https://doi.org/10.1029/2012jd017508, 2012 117, n/a-n/a.
Barker, D.M., Huang, W., Guo, Y.-R., Bourgeois, A.J., Xiao, Q.N., 2004. A three-
Gao, M., Beig, G., Song, S., Zhang, H., Hu, J., Ying, Q., Liang, F., Liu, Y., Wang, H., Lu, X.,
dimensional variational data assimilation system for MM5: implementation and
Zhu, T., Carmichael, G.R., Nielsen, C.P., McElroy, M.B., 2018. The impact of power
initial results. Mon. Weather Rev. 132, 897–914. https://doi.org/10.1175/1520-
generation emissions on ambient PM2.5 pollution and human health in China and
0493(2004)132<0897:atvdas>2.0.co. ;2.
India. Environ. Int. 121, 250–259. https://doi.org/10.1016/j.envint.2018.09.015.
Bhattacharya, A., Venkataraman, C., Sarkar, T., Sharma, A.K., Sharma, A., Anand, S.,
Gao, M., Carmichael, G.R., Wang, Y., Saide, P.E., Yu, M., Xin, J., Liu, Z., Wang, Z., 2016.
Ganguly, D., Bhawar, R., Dey, S., Ghosh, S., 2022. An analysis of the aerosol lifecycle
Modeling study of the 2010 regional haze event in the North China Plain. Atmos.
over India: COALESCE intercomparison of three general circulation models.
Chem. Phys. 16, 1673–1691. https://doi.org/10.5194/acp-16-1673-2016.
J Geophys Res Atmospheres 127. https://doi.org/10.1029/2022jd036457.
Gao, Y., Zhang, M., Liu, Z., Wang, L., Wang, P., Xia, X., Tao, M., Zhu, L., 2015. Modeling
Bran, S.H., Srivastava, R., 2017. Investigation of PM2.5 mass concentration over India
the feedback between aerosol and meteorological variables in the atmospheric
using a regional climate model. Environ. Pollut. 224, 484–493. https://doi.org/
boundary layer during a severe fog–haze event over the North China Plain. Atmos.
10.1016/j.envpol.2017.02.030.
Chem. Phys. 15, 4279–4295. https://doi.org/10.5194/acp-15-4279-2015.
Brooks, J., Allan, J.D., Williams, P.I., Liu, D., Fox, C., Haywood, J., Langridge, J.M.,
Gautam, R., Hsu, N.C., Lau, K.-M., Tsay, S.-C., Kafatos, M., 2009. Enhanced pre-monsoon
Highwood, E.J., Kompalli, S.K., O’Sullivan, D., Babu, S.S., Satheesh, S.K., Turner, A.
warming over the Himalayan-Gangetic region from 1979 to 2007. Geophys. Res.
G., Coe, H., 2019. Vertical and horizontal distribution of submicron aerosol chemical
Lett. 36 https://doi.org/10.1029/2009gl037641.
composition and physical characteristics across northern India during pre-monsoon
Gawhane, R.D., Rao, P.S.P., Budhavant, K., Meshram, D.C., Safai, P.D., 2019.
and monsoon seasons. Atmos. Chem. Phys. 19, 5615–5634. https://doi.org/
Anthropogenic fine aerosols dominate over the Pune region, Southwest India.
10.5194/acp-19-5615-2019.
Meteorol. Atmos. Phys. 131, 1497–1508. https://doi.org/10.1007/s00703-018-
Carter, W.P.L., 2015. Development of a database for chemical mechanism assignments
0653-y.
for volatile organic emissions. J. Air Waste Manage. 65, 1171–1184. https://doi.org/
GBD, 2018. GBD MAPS working group. 2018. In: Burden of Disease Attributable to Major
10.1080/10962247.2015.1013646.
Air Pollution Sources in India. Special Report 21. Health Effects Institute, Boston,
Chen, H., Zhuang, B., Liu, J., Li, S., Wang, T., Xie, X., Xie, M., Li, M., Zhao, M., 2020.
MA.
Regional climate responses in East Asia to the black carbon aerosol direct effects
Ghude, S.D., Bhat, G.S., Prabhakaran, T., et al., 2017. Winter fog experiment over the
from India and China in summer. J. Clim. 33, 9783–9800. https://doi.org/10.1175/
indo-gangetic plains of India. Curr. Sci. India 112, 767–784. https://doi.org/
jcli-d-19-0706.1.
10.18520/cs/v112/i04/767-784.
Chin, M., Ginoux, P., Kinne, S., Torres, O., Holben, B.N., Duncan, B.N., Martin, R.V.,
Ghude, S.D., Chate, D.M., Jena, C., Beig, G., Kumar, R., Barth, M.C., Pfister, G.G.,
Logan, J.A., Higurashi, A., Nakajima, T., 2002. Tropospheric aerosol optical
Fadnavis, S., Pithani, P., 2016. Premature mortality in India due to PM2.5 and ozone
thickness from the GOCART model and comparisons with satellite and sun
exposure. Geophys. Res. Lett. 43, 4650–4658. https://doi.org/10.1002/
photometer measurements. J. Atmos. Sci. 59, 461–483. https://doi.org/10.1175/
2016gl068949.
1520-0469(2002)059<0461:taotft>2.0.co. ;2.
Giglio, L., Randerson, J.T., Werf, G.R., 2013. Analysis of daily, monthly, and annual
Conibear, L., Butt, E.W., Knote, C., Arnold, S.R., Spracklen, D.V., 2018. Residential
burned area using the fourth-generation global fire emissions database (GFED4).
energy use emissions dominate health impacts from exposure to ambient particulate
J. Geophys.Res.Biogeosciences. 118, 317–328. https://doi.org/10.1002/jgrg.20042.
matter in India. Nat. Commun. 9, 617. https://doi.org/10.1038/s41467-018-02986-
7.
13
Ginoux, P., Chin, M., Tegen, I., Prospero, J., Holben, B., Dubovik, O., Lin, S.-J., 2001. Liu, J., Scheuer, E., Dibb, J., Diskin, G.S., Ziemba, L.D., Thornhill, K.L., Anderson, B.E.,
Sources and global distributions of dust aerosols simulated with the GOCART model. Wisthaler, A., Mikoviny, T., Devi, J.J., Bergin, M., Perring, A.E., Markovic, M.Z.,
J. Geophys. Res. 106 (20), 255-20,27. Schwarz, J.P., Campuzano-Jost, P., Day, D.A., Jimenez, J.L., Weber, R.J., 2015.
Green, M., Kondragunta, S., Ciren, P., Xu, C., 2009. Comparison of GOES and MODIS Brown carbon aerosol in the North American continental troposphere: sources,
aerosol optical depth (AOD) to aerosol robotic network (AERONET) AOD and abundance, and radiative forcing. Atmos. Chem. Phys. 15, 7841–7858. https://doi.
IMPROVE PM2.5 mass at bondville, Illinois. J. Air Waste Manage. 59, 1082–1091. org/10.5194/acp-15-7841-2015.
https://doi.org/10.3155/1047-3289.59.9.1082. Mangla, R., J, I., C, S.S., 2020. Inter-comparison of multi-satellites and aeronet AOD over
Guenther, A., Karl, T., Harley, P., Wiedinmyer, C., Palmer, P.I., Geron, C., 2006. Indian region. Atmos. Res. 240, 104950 https://doi.org/10.1016/j.
Estimates of global terrestrial isoprene emissions using MEGAN (model of emissions atmosres.2020.104950.
of gases and aerosols from nature). Atmos. Chem. Phys. 6, 3181–3210. https://doi. McKeen, S., et al., 2007. Evaluation of several PM2.5 forecast models using data collected
org/10.5194/acp-6-3181-2006. during the ICARTT/NEAQS 2004 field study. J. Geophys. Res. 112, D10S20. https://
Guenther, A., Zimmerman, P., Wildermuth, M., 1994. Natural volatile organic compound doi.org/10.1029/2006JD007608.
emission rate estimates for U.S. woodland landscapes. Atmos. Environ. 28, Mehta, M., 2015. A study of aerosol optical depth variations over the Indian region using
1197–1210. https://doi.org/10.1016/1352-2310(94)90297-6. thirteen years (2001–2013) of MODIS and MISR Level 3 data. Atmos. Environ. 109,
Guttikunda, S.K., Goel, R., Pant, P., 2014. Nature of air pollution, emission sources, and 161–170. https://doi.org/10.1016/j.atmosenv.2015.03.021.
management in the Indian cities. Atmos. Environ. 95, 501–510. https://doi.org/ Mondal, A., Sah, N., Sharma, A., Venkataraman, C., Patil, N., 2020. Absorbing aerosols
10.1016/j.atmosenv.2014.07.006. and high temperature extremes in India: a general circulation modelling study. Int. J.
Harris, E., Sinha, B., Hoppe, P., Crowley, J.N., Ono, S., Foley, S., 2012. Sulfur isotope Climatol. https://doi.org/10.1002/joc.6783.
fractionation during oxidation of sulfur dioxide: gas-phase oxidation by OH radicals Murray, C.J.L., Aravkin, A.Y., Zheng, P., et al., 2020. Global burden of 87 risk factors in
and aqueous oxidation by H2O2, O3 and iron catalysis. Atmos. Chem. Phys. 12, 204 countries and territories, 1990–2019: a systematic analysis for the Global
407–423. https://doi.org/10.5194/acp-12-407-2012. Burden of Disease Study 2019. Lancet 396, 1223–1249. https://doi.org/10.1016/
Harris, I., Jones, P.D., Osborn, T.J., Lister, D.H., 2014. Updated high-resolution grids of s0140-6736(20)30752-2.
monthly climatic observations – the CRU TS3.10 Dataset. Int. J. Climatol. 34, Nguyen, D.L., 2014. A brief review of air quality models and their applications. J. Atmos.
623–642. https://doi.org/10.1002/joc.3711. Clim. Change 1.
Heald, C.L., C, J.L., Lee, T., Benedict, K.B., Schwandner, F.M., Li, Y., Clarisse, L., Nguyen, G.T.H., Shimadera, H., Sekiguchi, A., Matsuo, T., Kondo, A., 2019a.
Hurtmans, D.R., Damme, M.V., Clerbaux, C., Coheur, P.-F., Philip, S., Martin, R.V., Investigation of aerosol direct effects on meteorology and air quality in East Asia by
Pye, H.O.T., 2012. Atmospheric ammonia and particulate inorganic nitrogen over using an online coupled modeling system. Atmos. Environ. 207, 182–196. https://
the United States. Atmos. Chem. Phys. 12, 10295–10312. https://doi.org/10.5194/ doi.org/10.1016/j.atmosenv.2019.03.017.
acp-12-10295-2012. Nguyen, G.T.H., Shimadera, H., Uranishi, K., Matsuo, T., Kondo, A., Thepanondh, S.,
Heinold, B., Tegen, I., Schepanski, K., Hellmuth, O., 2008. Dust radiative feedback on 2019b. Numerical assessment of PM2.5 and O3 air quality in continental Southeast
Saharan boundary layer dynamics and dust mobilization. Geophys. Res. Lett. 35 Asia: Baseline simulation and aerosol direct effects investigation. Atmos. Environ.
https://doi.org/10.1029/2008gl035319. 219, 117054 https://doi.org/10.1016/j.atmosenv.2019.117054.
Henriksson, S.V., Laaksonen, A., Kerminen, V.-M., Räisänen, P., Järvinen, H., Ojha, N., Sharma, A., Kumar, M., Girach, I., Ansari, T.U., Sharma, S.K., Singh, N.,
Sundström, A.-M., Leeuw, G. de, 2011. Spatial distributions and seasonal cycles of Pozzer, A., Gunthe, S.S., 2020. On the widespread enhancement in fine particulate
aerosols in India and China seen in global climate-aerosol model. Atmos. Chem. matter across the Indo-Gangetic Plain towards winter. Sci. Rep-uk 10, 5862. https://
Phys. 11, 7975–7990. https://doi.org/10.5194/acp-11-7975-2011. doi.org/10.1038/s41598-020-62710-8.
Hoesly, R.M., Smith, S.J., Feng, L., Klimont, Z., Janssens-Maenhout, G., Pitkanen, T., Pan, X., Chin, M., Gautam, R., Bian, H., Kim, D., Colarco, P.R., Diehl, T.L., Takemura, T.,
Seibert, J.J., Vu, L., Andres, R.J., Bolt, R.M., Bond, T.C., Dawidowski, L., Kholod, N., Pozzoli, L., Tsigaridis, K., Bauer, S., Bellouin, N., 2015. A multi-model evaluation of
Kurokawa, J., Li, M., Liu, L., Lu, Z., Moura, M.C.P., O’Rourke, P.R., Zhang, Q., 2017. aerosols over South Asia: common problems and possible causes. Atmos. Chem.
Historical (1750–2014) anthropogenic emissions of reactive gases and aerosols from Phys. 15, 5903–5928. https://doi.org/10.5194/acp-15-5903-2015.
the Community Emissions Data System (CEDS). Geosci. Model Dev. (GMD) 11, Pandey, A., Brauer, M., Cropper, M.L., et al., 2021. Health and economic impact of air
369–408. https://doi.org/10.5194/gmd-11-369-2018. pollution in the states of India: the Global Burden of Disease Study 2019. Lancet
Hong, C., Zhang, Q., Zhang, Y., Tang, Y., Tong, D., He, K., 2017a. Multi-year downscaling Planet. Health 5, e25–e38. https://doi.org/10.1016/s2542-5196(20)30298-9.
application of two-way coupled WRF v3.4 and CMAQ v5.0.2 over east Asia for Pandey, A., Venkataraman, C., 2014. Estimating emissions from the Indian transport
regional climate and air quality modeling: model evaluation and aerosol direct sector with on-road fleet composition and traffic volume. Atmos. Environ. 98,
effects. Geosci. Model Dev. (GMD) 10, 2447–2470. https://doi.org/10.5194/gmd- 123–133. https://doi.org/10.1016/j.atmosenv.2014.08.039.
10-2447-2017. Pandithurai, G., Dipu, S., Dani, K.K., Tiwari, S., Bisht, D.S., Devara, P.C.S., Pinker, R.T.,
Hong, C., Zhang, Q., Zhang, Y., Tang, Y., Tong, D., He, K., 2017b. Multi-year downscaling 2008. Aerosol radiative forcing during dust events over New Delhi, India. J Geophys
application of two-way coupled WRF v3.4 and CMAQ v5.0.2 over east Asia for Res Atmospheres 113. https://doi.org/10.1029/2008jd009804, 1984 2012.
regional climate and air quality modeling: model evaluation and aerosol direct Patil, N., Venkataraman, C., Muduchuru, K., Ghosh, S., Mondal, A., 2019. Disentangling
effects. Geosci. Model Dev. (GMD) 10, 2447–2470. https://doi.org/10.5194/gmd- sea-surface temperature and anthropogenic aerosol influences on recent trends in
10-2447-2017. South Asian monsoon rainfall. Clim. Dynam. 52, 2287–2302. https://doi.org/
Hu, R.-M., Blanchet, J.-P., Girard, E., 2005. Evaluation of the direct and indirect radiative 10.1007/s00382-018-4251-y.
and climate effects of aerosols over the western Arctic. J Geophys Res Atmospheres. Pipalatkar, P., Khaparde, V.V., Gajghate, D.G., Bawase, M.A., 2014. Source
https://doi.org/10.1029/2004jd005043, 1984 2012 110. apportionment of PM2.5 using a CMB model for a centrally located Indian city.
Jacobson, M.Z., 2001. Global direct radiative forcing due to anthropogenic and natural Aerosol Air Qual. Res. 14, 1089–1099. https://doi.org/10.4209/aaqr.2013.04.0130.
aerosols multicomponent. J. Geophys. Res. 106, 1551–1568. https://doi.org/ Ram, K., Sarin, M.M., Tripathi, S.N., 2012. Temporal trends in atmospheric PM2.5,
10.1029/2000JD900514. PM10, elemental carbon, organic carbon, water-soluble organic carbon, and optical
Jat, R., Gurjar, B.R., Lowe, D., 2021. Regional pollution loading in winter months over properties: impact of biomass burning emissions in the Indo-Gangetic plain. Environ.
India using high resolution WRF-Chem simulation. Atmos. Res. 249, 105326 https:// Sci. Technol. 46, 686–695. https://doi.org/10.1021/es202857w.
doi.org/10.1016/j.atmosres.2020.105326. Rastogi, N., Singh, A., Singh, D., Sarin, M.M., 2014. Chemical characteristics of PM2.5 at
Jena, C., Ghude, S.D., Kulkarni, R., Debnath, S., Kumar, R., Soni, V.K., Acharja, P., a source region of biomass burning emissions: evidence for secondary aerosol
Kulkarni, S.H., Khare, M., Kaginalkar, A.J., Chate, D.M., Ali, K., Nanjundiah, R.S., formation. Environ. Pollut. 184, 563–569. https://doi.org/10.1016/j.
Rajeevan, M.N., 2020. Evaluating the sensitivity of fine particulate matter (PM2.5) envpol.2013.09.037.
simulations to chemical mechanism in Delhi. Atmospheric Chem.Phys.Discuss. 2020 Sabari, E.E.M., Reynoso, J.A.G., 2020. Meteorological Data Assimilation for Air Quality
1–28. https://doi.org/10.5194/acp-2020-673. Modeling with WRF-Chem: Central Mexico Case Study. Atmósfera. https://doi.org/
Johnson, A.F., 2020. Valuing Science Informing Decisions - Health Institute Annual 10.20937/atm.52804.
Report. DS Graphics, Lowell, MA. Sadavarte, P., Venkataraman, C., 2014. Trends in multi-pollutant emissions from a
Karambelas, A., Holloway, T., Kiesewetter, G., Heyes, C., 2018. Constraining the technology-linked inventory for India: I. Industry and transport sectors. Atmos.
uncertainty in emissions over India with a regional air quality model evaluation. Environ. 99, 353–364. https://doi.org/10.1016/j.atmosenv.2014.09.081.
Atmos. Environ. 174, 194–203. https://doi.org/10.1016/j.atmosenv.2017.11.052. Sadavarte, P., Venkataraman, C., Cherian, R., Patil, N., Madhavan, B.L., Gupta, T.,
Lelieveld, J., Evans, J.S., Fnais, M., Giannadaki, D., Pozzer, A., 2015. The contribution of Kulkarni, S., Carmichael, G.R., Adhikary, B., 2016. Seasonal differences in aerosol
outdoor air pollution sources to premature mortality on a global scale. Nature 525 abundance and radiative forcing in months of contrasting emissions and rainfall over
(7569), 367–371. northern South Asia. Atmos. Environ. 125, 512–523. https://doi.org/10.1016/j.
Li, Jiawei, Han, Z., Wu, Y., Xiong, Z., Xia, X., Li, Jie, Liang, L., Zhang, R., 2020. Aerosol atmosenv.2015.10.092.
radiative effects and feedbacks on boundary layer meteorology and PM2.5 chemical Samiksha, S., Raman, R.S., Nirmalkar, J., Kumar, S., Sirvaiya, R., 2017. PM10 and PM2.5
components during winter haze events over the Beijing-Tianjin-Hebei region. Atmos. chemical source profiles with optical attenuation and health risk indicators of paved
Chem. Phys. 20, 8659–8690. https://doi.org/10.5194/acp-20-8659-2020. and unpaved road dust in Bhopal, India. Environ. Pollut. 222, 477–485. https://doi.
Li, M., Wang, T., Xie, M., Zhuang, B., Li, S., Han, Y., Chen, P., 2017. Impacts of aerosol- org/10.1016/j.envpol.2016.11.067.
radiation feedback on local air quality during a severe haze episode in Nanjing Sanap, S.D., Ayantika, D.C., Pandithurai, G., Niranjan, K., 2014. Assessment of the
megacity, eastern China. Tellus B 69, 1339548. https://doi.org/10.1080/ aerosol distribution over Indian subcontinent in CMIP5 models. Atmos. Environ. 87,
16000889.2017.1339548. 123–137. https://doi.org/10.1016/j.atmosenv.2014.01.017.
Liang, J., Jacobson, M.Z., 1999. A study of sulfur dioxide oxidation pathways over a Schell, B., Ackermann, I.J., Hass, H., Binkowski, F.S., Ebel, A., 2001. Modeling the
range of liquid water contents, pH values, and temperatures. J Geophys Res formation of secondary organic aerosol within a comprehensive air quality model
Atmospheres 104, 13749–13769. https://doi.org/10.1029/1999jd900097. system. J Geophys Res Atmospheres 106, 28275–28293. https://doi.org/10.1029/
2001jd000384.
14
Schnaiter, M., Horvath, H., Möhler, O., Naumann, K.-H., Saathoff, H., Schöck, O.W., Venkataraman, C., Habib, G., Eiguren-Fernandez, A., Miguel, A.H., Friedlander, S.K.,
2003. UV-VIS-NIR spectral optical properties of soot and soot-containing aerosols. 2005. Residential biofuels in South Asia: carbonaceous aerosol emissions and climate
J. Aerosol Sci. 34, 1421–1444. https://doi.org/10.1016/s0021-8502(03)00361-6. impacts. Science 307, 1454–1456. https://doi.org/10.1126/science.1104359.
Sekiguchi, A., Shimadera, H., Kondo, A., 2018. Impact of aerosol direct effect on Volkamer, R., San Martini, F., Molina, L.T., Salcedo, D., Jimenez, J.L., Molina, M.J.,
wintertime PM2.5 simulated by an online coupled meteorology-air quality model 2007. A missing sink for gas-phase glyoxal in Mexico City: formation of secondary
over East Asia. Aerosol Air Qual. Res. 18, 1068–1079. https://doi.org/10.4209/ organic aerosol. Geophys. Res. Lett. 34 (19), 19807 https://doi.org/10.1029/
aaqr.2016.06.0282. 2007GL030752.
Sharma, A., Bhattacharya, A., Venkataraman, C., 2022. Influence of aerosol radiative Wang, D., Jiang, B., Lin, W., Gu, F., 2019a. Effects of aerosol-radiation feedback and
effects on surface temperature and snow melt in the Himalayan region. Sci. Total topography during an air pollution event over the North China Plain during
Environ. 810, 151299 https://doi.org/10.1016/j.scitotenv.2021.151299. December 2017. Atmos. Pollut. Res. 10, 587–596. https://doi.org/10.1016/j.
Shi, H., Xiao, Z., Zhan, X., Ma, H., Tian, X., 2019. Evaluation of MODIS and two apr.2018.10.006.
reanalysis aerosol optical depth products over AERONET sites. Atmos. Res. 220, Wang, D., Jiang, B., Lin, W., Gu, F., 2019b. Effects of aerosol-radiation feedback and
75–80. https://doi.org/10.1016/j.atmosres.2019.01.009. topography during an air pollution event over the North China Plain during
Sijikumar, S., Aneesh, S., Rajeev, K., 2016. Multi-year model simulations of mineral dust December 2017. Atmos. Pollut. Res. 10, 587–596. https://doi.org/10.1016/j.
distribution and transport over the Indian subcontinent during summer monsoon apr.2018.10.006.
seasons. Meteorol. Atmos. Phys. 128, 453–464. https://doi.org/10.1007/s00703- Wang, J., Allen, D.J., Pickering, K.E., Li, Z., He, H., 2016a. Impact of aerosol direct effect
015-0422-0. on East Asian air quality during the EAST-AIRE campaign. J Geophys Res
Simmons, A.J., Poli, P., Dee, D.P., Berrisford, P., Hersbach, H., Kobayashi, S., Peubey, C., Atmospheres 121, 6534–6554. https://doi.org/10.1002/2016jd025108.
2014. Estimating low-frequency variability and trends in atmospheric temperature Wang, J., Allen, D.J., Pickering, K.E., Li, Z., He, H., 2016b. Impact of aerosol direct effect
using ERA-Interim. Q. J. Roy. Meteorol. Soc. 140, 329–353. https://doi.org/ on East Asian air quality during the EAST-AIRE campaign. J Geophys Res
10.1002/qj.2317. Atmospheres 121, 6534–6554. https://doi.org/10.1002/2016jd025108.
Singh, N., Murari, V., Kumar, M., Barman, S.C., Banerjee, T., 2017. Fine particulates over Wang, J., Wang, S., Jiang, J., Ding, A., Zheng, M., Zhao, B., Wong, D.C., Zhou, W.,
South Asia: review and meta-analysis of PM2.5 source apportionment through Zheng, G., Wang, L., Pleim, J.E., Hao, J., 2014. Impact of aerosol–meteorology
receptor model. Environ. Pollut. 223, 121–136. https://doi.org/10.1016/j. interactions on fine particle pollution during China’s severe haze episode in January
envpol.2016.12.071. 2013. Environ. Res. Lett. 9, 094002 https://doi.org/10.1088/1748-9326/9/9/
Sokolik, I.N., Toon, O.B., 1996. Direct radiative forcing by anthropogenic airborne 094002.
mineral aerosols. Nature 381, 681–683. https://doi.org/10.1038/381681a0. Wiedinmyer, C., Yokelson, R.J., Gullett, B.K., 2014. Global emissions of trace gases,
Srivastava, P., Naja, M., 2021. Characteristics of carbonaceous aerosols derived from particulate matter, and hazardous air pollutants from open burning of domestic
long-term high-resolution measurements at a high-altitude site in the central waste. Environ. Sci. Technol. 48, 9523–9530. https://doi.org/10.1021/es502250z.
Himalayas: radiative forcing estimates and role of meteorology and biomass burning. Xing, J., Mathur, R., Pleim, J., Hogrefe, C., Gan, C., Wong, D.C., Wei, C., Wang, J., 2015.
Environ. Sci. Pollut. Res. 28, 14654–14670. https://doi.org/10.1007/s11356-020- Air pollution and climate response to aerosol direct radiative effects: a modeling
11579-1. study of decadal trends across the northern hemisphere. J Geophys Res Atmospheres
Stockwell, W.R., Middleton, P., Chang, J.S., Tang, X., 1990. The second generation 120 (12). https://doi.org/10.1002/2015jd023933, 221-12,236.
regional acid deposition model chemical mechanism for regional air quality Xing, J., Wang, J., Mathur, R., Pleim, J., Wang, S., Hogrefe, C., Gan, C.-M., Wong, D.C.,
modeling. J Geophys Res Atmospheres 95, 16343–16367. https://doi.org/10.1029/ Hao, J., 2016. Unexpected benefits of reducing aerosol cooling effects. Environ. Sci.
jd095id10p16343. Technol. 50, 7527–7534. https://doi.org/10.1021/acs.est.6b00767.
Tanaka, N., Rye, D.M., Xiao, Y., Lasag, A.C., 1994. Use of stable sulfur isotope Xing, J., Wang, J., Mathur, R., Wang, S., Sarwar, G., Pleim, J., Hogrefe, C., Zhang, Y.,
systematics for evaluating oxidation reaction pathways and in-cloud-scavenging of Jiang, J., Wong, D.C., Hao, J., 2017. Impacts of aerosol direct effects on tropospheric
sulfur dioxide in the atmosphere H2 S•"’• 4 •,• H2SO 4. Geophys. Res. Lett. 21, ozone through changes in atmospheric dynamics and photolysis rates. Atmos. Chem.
1519–1522. Phys. 17, 9869–9883. https://doi.org/10.5194/acp-17-9869-2017.
Tasoglou, A., Saliba, G., Subramanian, R., Pandis, S.N., 2017. Absorption of chemically Yu, H., Kaufman, Y.J., Chin, M., Feingold, G., Remer, L.A., Anderson, T.L., Balkanski, Y.,
aged biomass burning carbonaceous aerosol. J. Aerosol Sci. 113, 141–152. https:// Bellouin, N., Boucher, O., Christopher, S., DeCola, P., Kahn, R., Koch, D., Loeb, N.,
doi.org/10.1016/j.jaerosci.2017.07.011. Reddy, M.S., Schulz, M., Takemura, T., Zhou, M., 2006. A review of measurement-
Tuccella, P., Curci, G., Visconti, G., Bessagnet, B., Menut, L., Park, R.J., 2012. Modeling based assessments of the aerosol direct radiative effect and forcing. Atmos. Chem.
of gas and aerosol with WRF/Chem over Europe: evaluation and sensitivity study. Phys. 6, 613–666. https://doi.org/10.5194/acp-6-613-2006.
J. Geophys. Res. Atmos. 117 (D3) https://doi.org/10.1029/2011JD016302. Zhang, L., Jacob, D.J., Knipping, E.M., Kumar, N., Munger, J.W., Carouge, C.C.,
Twomey, S., 1991. Aerosols, clouds and radiation. Atmos. Environ. Part A Gen. Top. 25, Donkelaar, A. van, Wang, Y.X., Chen, D., 2012. Nitrogen deposition to the United
2435–2442. https://doi.org/10.1016/0960-1686(91)90159-5. States: distribution, sources, and processes. Atmos. Chem. Phys. 12, 4539–4554.
Venkataraman, C., Bhushan, M., Dey, S., Ganguly, D., Gupta, T., Habib, G., Kesarkar, A., https://doi.org/10.5194/acp-12-4539-2012.
Phuleria, H., Raman, R.S., 2020. Indian network project on carbonaceous aerosol Zhang, He, J., Zhu, S., Gantt, B., 2016. Sensitivity of simulated chemical concentrations
emissions, source apportionment and climate impacts (COALESCE) Indian network and aerosol-meteorology interactions to aerosol treatments and biogenic organic
project on carbonaceous aerosol emissions, source apportionment and climate emissions in WRF. Chem. J. Geophys. Res. 121 (10), 6014–6048. https://doi.org/
impacts (COALESCE). Bull. Am. Meteorol. Soc. 101, E1052–E1068. https://doi.org/ 10.1002/2016JD024882.
10.1175/bams-d-19-0030.1. Zhang, R., Sun, X., Shi, A., Huang, Y., Yan, J., Nie, T., Yan, X., Li, X., 2018. Secondary
Venkataraman, C., Brauer, M., Tibrewal, K., Sadavarte, P., Ma, Q., Cohen, A., inorganic aerosols formation during haze episodes at an urban site in Beijing, China.
Chaliyakunnel, S., Frostad, J., Klimont, Z., Martin, R.V., Millet, D.B., Philip, S., Atmos. Environ. 177, 275–282. https://doi.org/10.1016/j.atmosenv.2017.12.031.
Walker, K., Wang, S., 2018. Source influence on emission pathways and ambient Zhao, C., Liu, X., Leung, L.R., Johnson, B., McFarlane, S.A., G, W.I., Fast, J.D., Easter, R.,
PM2.5 pollution over India (2015–2050). Atmos. Chem. Phys. 18, 8017–8039. 2010. The spatial distribution of mineral dust and its shortwave radiative forcing
https://doi.org/10.5194/acp-18-8017-2018. over North Africa: modeling sensitivities to dust emissions and aerosol size
treatments. Atmos. Chem. Phys. 10, 8821–8838. https://doi.org/10.5194/acp-10-
8821-2010.
15

Aerosol Radiactive

Uploaded by

Copyright:

Available Formats

Aerosol Radiactive

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Aerosol Radiactive

Uploaded by

Copyright:

Available Formats

Atmospheric Environment 298 (2023) 119609

Contents lists available at ScienceDirect

Aerosol radiative feedback enhances particulate pollution over India: A

• Online coupled WRF-Chem simulations

1. Introduction et al., 2011). This diversity results in seasonally varying

You might also like