Recent Increase in Winter Hazy Days over Central India and the Arabian Sea

Abstract

Indian subcontinent is greatly vulnerable to air pollution, especially during the winter season. Here, we use 15 years (2003–2017) of satellite and model reanalysis datasets over India and adjoining Seas to estimate the trend in hazy days (i.e. days with high aerosol loading) during the dry winter season (November to February). The number of hazy days is increasing at the rate of ~2.6 days per year over Central India. Interestingly, this is higher than over the Indo-Gangetic Plain (~1.7 days/year), a well known global hotspot of particulate pollution. Consistent increasing trends in absorbing aerosols are also visible in the recent years. As a result, the estimated atmospheric warming trends over Central India are two-fold higher than that over Indo-Gangetic Plain. This anomalous increment in hazy days over Central India is associated with the relatively higher increase in biomass burning over the region. Moreover, the trend in aerosol loading over the Arabian Sea, which is located downwind to Central India, is also higher than that over the Bay of Bengal during the dry winter season. Our findings not only draw attention to the rapid deteriorating air quality over Central India, but also underline the significance of increasing biomass burning under the recent climate change.

Full text

1
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
Recent Increase in Winter Hazy
Days over Central India and the
Arabian Sea
Abin Thomas
1, Chandan Sarangi2* & Vijay P. Kanawade
1*
Indian subcontinent is greatly vulnerable to air pollution, especially during the winter season. Here,
we use 15 years (2003–2017) of satellite and model reanalysis datasets over India and adjoining Seas
to estimate the trend in hazy days (i.e. days with high aerosol loading) during the dry winter season
(November to February). The number of hazy days is increasing at the rate of ~2.6 days per year over
Central India. Interestingly, this is higher than over the Indo-Gangetic Plain (~1.7 days/year), a well
known global hotspot of particulate pollution. Consistent increasing trends in absorbing aerosols are
also visible in the recent years. As a result, the estimated atmospheric warming trends over Central
India are two-fold higher than that over Indo-Gangetic Plain. This anomalous increment in hazy days
over Central India is associated with the relatively higher increase in biomass burning over the region.
Moreover, the trend in aerosol loading over the Arabian Sea, which is located downwind to Central
India, is also higher than that over the Bay of Bengal during the dry winter season. Our ndings not
only draw attention to the rapid deteriorating air quality over Central India, but also underline the
signicance of increasing biomass burning under the recent climate change.
Aerosols are ubiquitous in the atmosphere. While natural aerosols constitute the largest fraction of global aero-
sol burden, regional hotspots of high aerosol loading coincide with regions of high population density, urban-
ization and industrialization, or biomass burning. e long-term measurements of aerosols1–3, observational
campaigns4–6, and remote sensing from ground and space7,8 have provided requisite datasets to improve our
understanding of the physical, optical and chemical properties of aerosols. e extensive observational and mod-
elling eorts over the last two decades have aided remarkably to advance our knowledge of aerosols inuence of
the climate9. For instance, the cooling eect from increasing aerosols has masked about one-third of the increas-
ing greenhouse warming over the past half-century10. However, the lack of understanding of the variability of
aerosols and their eects at regional scale contribute to the existing large uncertainty in aerosol feedbacks in
future climate predictions.
e Indian subcontinent and adjoining Seas experience a tropical and sub-tropical climate and have been a
focus of the study for aerosols over the last two decades. Observational campaigns (e.g. Indian Space Research
Organization-Geosphere Biosphere Programme, ISRO-GBP; Indian Ocean Experiment, INDOEX; Arabian Sea
Monsoon Experiment, ARMEX) have shown large aerosol negative radiative forcing at the surface and relatively
large atmospheric warming than top of the atmosphere (TOA)11–13. e wintertime hazardous air pollution sce-
narios over the densely populated regions of India have recently received the utmost scientic attention. e win-
ter season is characterized by a shallower boundary layer, lower wind speed and low precipitation, leading to the
accumulation of aerosols near the surface2,14,15. Along with the obvious health impacts, studies have shown that
the high aerosol loading in winter can reduce radiation reaching to the surface by about 25%, thereby decreasing
crop yield16–18. Nonetheless, aerosols act as cloud condensation nuclei and aect cloud formation and rainfall19.
erefore, accurate knowledge of trend in aerosols over India during the winter season is extremely essential for
reducing uncertainty in future climate, health, and economic predictions.
In India, several recent studies have used satellite and ground based measurements spanning over decade to
nd the trend in aerosol loading over Indian region2,20–28. Dey and Girolamo20 and Dey et al.26, found a signicant
rise in anthropogenic aerosols over the Indian subcontinent during 2000–2010 using MISR aerosol optical depth
(AOD) data. ese studies highlighted that the rural areas of IGP are more polluted than that of urban cities in the
1Centre for Earth, Ocean and Atmospheric Sciences, University of Hyderabad, Hyderabad, Telangana, 500046, India.
2Pacic Northwest National Laboratory, Richland, Washington, 99352, USA. *email: chandansarangi591@gmail.
com; vijaykanawade03@yahoo.co.in
OPEN
Content courtesy of Springer Nature, terms of use apply. Rights reserved

2
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
www.nature.com/scientificreports/
peninsular India. Kaskaoutis et al.22, found an increasing trend in AOD over Kanpur (located in IGP), especially
during the post-monsoon and winter seasons based on Aerosol Robotic Network (AERONET) AOD data from
2001 to 2010. Hsu et al.25, also found an increasing trend in ne mode anthropogenic aerosols over North India
and Bay of Bengal (BoB), particularly during the dry winter and post-monsoon seasons using SeaWiFS measure-
ment from 1997 to 2010. Ramachandran et al.23, attributed the incremental trend in aerosols over New Delhi to
anthropogenic aerosols and over Northeast India to increase in forest re and biomass burning emissions. Babu
et al.2, showed an increasing trend in anthropogenic contribution to total aerosol loading during the dry winter
season. Moorthy et al.21, also found an increasing trend in aerosol loading in the current decade than its value in
1985. Srivasthava24 highlighted that the more than 70% of the Indian subcontinent shows a positive trend in AOD
from 2 to 6% during the winter and pre-monsoon seasons, with a trend of >6% over BoB. Kumar et al.27, recently
found a relatively high aerosol loading over IGP as compared to other parts of India, but a statistically insigni-
cant increasing trend of 0.002 AOD/year using MODIS-TERRA and nine ground-based stations data. ey also
observed a strong seasonality in aerosol loading with the dominance of ne mode aerosols over IGP, especially
during the dry winter season. Most of these previous studies highlight the increasing trend of aerosol loading over
highly polluted IGP and northern BoB in last two decades due to anthropogenic emissions.
Using 15 years of satellite (MODIS and OMI) observations and reanalysis (MERRA-2) data products, we illus-
trate that aerosol loading over Central India and the Arabian Sea during the dry winter season is increasing at a
greater rate than that over IGP and the BoB in the recent years. We focused our analyses for the dry winter season
(November to February) since the number of hazy days is highest during the season. A hazy day is referred to as
the day with high aerosol loading (i.e. AOD greater than 66th percentile value over a location).
Results and Discussion
Trend in the number of hazy days. e entire time period (2003–2017) is split into two sub-periods; past
years (2003–2007) and recent years (2013–2017) to highlight changes in aerosol loading in the current decade
compared to that of the previous decade. Figure1 shows averaged spatial distribution of MODIS columnar AOD
and OMI UV aerosol index (UV-AI) over India and adjoining Seas for the past and the recent years. e increased
Figure 1. Averaged spatial distribution of MODIS AOD for November through February over the time periods
(a) 2003–2007 (past years), (b) 2013–2017 (recent years) and (c) the percentage dierence between the recent
and the past years. e study regions are bounded by solid black line shown in (b); IGP: Indo-Gangetic Plain,
AS: Arabian Sea, CI: Central India and BoB: Bay of Bengal. Averaged spatial distribution of OMI UV-AI for
November through February over the time periods (d) 2003–2007, (e) 2013–2017, and (f) the percentage
dierence between the recent and the past years.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

3
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
www.nature.com/scientificreports/
aerosol loading in the recent years (2013–2017) is clearly evident, analogous to several previous studies2,20–25,27.
e rate of change in the aerosol loading between the past and the recent years is distinctive regionally due to
the variability in aerosol emission rates and latitudinal diverse climatic conditions. OMI retrieved UV-AI, which
is a measure of UV-absorbing aerosol particles such as soot/smoke and mineral dust, show similar distinctive
enhancement as that of AOD over Central India. is suggests the incremental dominance of absorbing aerosols
over Central India in the recent years.
We rst compare the number of days from low to high aerosol loading between the past and the recent years.
In order to do this, the daily AOD observations are segregated into three percentile bins: AOD less than 33rdper-
centile, between the value 33rdand 66th percentile,and greater than 66th percentile values over each 1° × 1° grid
based on the AOD values for November through February of the year 2003. e bins are identied as three dis-
tinct aerosol loading regimes; low (<33rd percentile), medium (33–66th percentiles), and high (>66th percentile).
en, the number of days for each of these regimes is counted during the past (Fig.2(a)) and the recent (Fig.2(b))
time periods. e number of days with low and medium aerosol loading has reduced in the recent years as com-
pared to the past years (Fig.2(a,b)). But, the days with high aerosol loading (>66th percentile) have increased
in the recent years over India and adjoining Seas (Fig.2(a,b)). Interestingly, the rate of increase in the number
of high aerosol loading (i.e. hazy) days over CI is higher (~2.6 days/season) than over IGP (~1.8 days/season)
(Fig.S1). is rate is also higher over the AS (~1.9 days/season) than over the BoB (~1.1 days/season). Noticeably,
all the four regions have about two-fold high aerosol loading days in the year 2017 as compared to the year 2003
(Fig.S1).is indicates that the aerosol burden over India is typically shied to higher values in the recent decade
relative to the previous decade. It should be noted that MODIS retrieved AOD over IGP in the dry winter season
is oen plagued under foggy conditions, but we have removed AOD > 1.0 to avoid fog/cloud contaminated AOD
retrievals in our analysis29,30. Wintertime fog thickens overnight and on till about an hour aer sunrise. en,
Figure 2. Trend in the number of days with aerosol loading from low to high regimes. (a) e number of days
with low (<33rd percentile value of AOD), medium (33–66th) and high (>66th) aerosol loading for November-
February of 2003–2007 (b) same as (a) except that for 2013–2017. (c) Trend in the number of days with <33rd,
33–66th and >66th percentile values of AOD for November–February of 2003–2017. Black dots indicate
statistical signicance using Student’s t-test at a condence interval of 95%.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

4
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
www.nature.com/scientificreports/
it gradually disperse or thin out by noontime via eective ventilation31 and since AQUA instrument onboard
MODIS has overpass over India at about 1:30 pm local time, chances of fog to inuence the results are negligible.
Further, the trend in the number of days for each of the percentile regime is calculated from 2003 to 2017
(Fig.2(c)). e trend in the number of days with low to medium aerosol loading days is found to be negative
over India and adjoining Seas. But, the trend in the number of days with high aerosol loading days is found to
be positive, with signicantly positive values over CI and the AS as compared to IGP, analogous to our nd-
ing from Fig.2(b). Previous studies have reported increasing aerosol loading trend over India using both the
ground-based2,21,22,28 and satellite observations1,20,23,24,32. But, we nd that the increasing aerosol loading trend
is more pronounced over CI and the AS, with >2 days per season over the time period from 2003 to 2017.
Moorthy et al.33, also revealed that the rate of increase in columnar AOD was more rapid (~4%) during the time
period from 2000 to 2011 compared to the previous decade and that Central Peninsular India showed the highest
increasing trend (3.63%/year) than the Southern Peninsular and Northern India, with higher rates during the
winter season as compared to other seasons. A recent study also showed less signicant increasing trends in AOD
over IGP over last decade27.
e dierence in the percentage frequency distribution for MODIS AOD between the past and the recent years
shows a decreasing frequency of low aerosol loading (AOD bin of 0.1), whereas it shows an increasing frequency
of medium to high aerosol loading (AOD bin >0.4) (Fig.3(a)) over all study regions. We also nd that MERRA-2
reanalysis data is able to reproduce similar intensity-frequency variation in the AOD (Fig.3(b)). In order to exam-
ine the changes in various aerosol species associated with the observed increasing trend in total aerosol loading,
we calculate the percentage frequency distribution for species-wise AOD separately for the past and the recent
years for dierent regions of India (Fig.S2). e dierences in the percentage frequency distribution for these
variables between the past and the recent years are then plotted in Fig.3(c,d). e BC and OC AOD is clubbed
together as they are co-emitted from various local sources, both anthropogenic aerosols and biomass burning.
e natural emissions like sea salt (SS) and dust AOD, which are mostly transported into CI and IGP domain, are
also shown in Fig.S3. Intensity-frequency variability similar to that seen for composite AOD (Fig.3(a,b) is pres-
ent for BC + OC and sulfate (Fig.3(c,d)) species over both IGP and CI regions. Interestingly, similar changes are
not seen for the case of transported species i.e. SS and dust AOD which indicates that the change in total aerosol
loading may be attributed to increase in the local emissions. Moreover, the intensity-frequency variation is more
prominent over CI and the AS compared to that over IGP. Further, the dierences in the vertical aerosol mixing
ratio proles illustrate signicant enhancement in BC and OC from the surface to 700hPa (~3.1 km) which imply
near-surface sources (Fig.S5). But, the anomalous enhancements in BC and OC mixing ratios are higher over CI
in the recent years than over the IGP.
Interestingly, the intensity-frequency changes in AOD observed over the AS region is more prominent com-
pared to the BoB region. In addition, the anthropogenic species over the AS, particularly BC, OC, and sulfate,
are found to be three times higher than SS aerosols during the winter season (Figs.3 and S3). e mean wind
Figure 3. e percentage dierence in the frequency distribution of (a) MODIS AOD, (b) MERRA-2 AOD,
(c) MERRA-2 (BC + OC) AOD and (d) MERRA-2 sulfate AOD between the past and the recent years over
dierent study regions.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

5
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
www.nature.com/scientificreports/
circulation clearly illustrates predominant northeasterly ow over the continental region mostly owing into the
AS i.e. the AS is located in the downwind regions of CI (Fig.S4). us, the BC, OC, and sulfate aerosols over the
AS are probably transported from local anthropogenic emissions over the CI. is also explains the high percent-
age change in aerosol loading over the AS than the BoB in the recent years (Fig.3). is suggests that the increase
in continental aerosol emissions has profound impact on the aerosol loading over the adjacent marine regions.
is nding is dierent from previous investigators as those studies indicated that the aerosol loading over the
BoB is higher than over the AS, and that the relative contribution of anthropogenic aerosol mass tend to be higher
over the BoB34–37. It could be noted that most of these studies used data prior to the year 2012, except the recent
study by Srivastava24, further highlighting the signicance of our nding on aerosol perturbations in the recent
years (2013–2017).
Rapid urbanization in developing nations like India is generally the primary source of the overall aerosol bur-
den38,39. For example, out of the ten most populous metropolitan areas in India, ve of them lie within CI region
(Bangalore, Hyderabad, Mumbai, Nagpur and Pune) and these cities are known for rapid change in the land use
and land cover over the last decade40. It should be noted that forests, shrubs, and cropland contributes to a large
fraction of the land cover over Central India. As a result, the biomass burning activities peak within the two central
states (Madhya Pradesh and Maharashtra), accounting for about 36% of the total re counts in India41. Besides,
the Eastern Ghats in Central Eastern India is a dense active re hotspot, owing to shiing cultivation practices and
clearing of mixed deciduous forest in the late winter season42,43. erefore, biomass burning (e.g. forest res, crop
residue burning, trash/wood burning), which is a major sources of aerosol loading over India44, can contribute
signicantly to this anomalous enhancement in AOD in the recent decade. Figure4(a–c) presents spatial map of
1° × 1° gridded total re counts over India during the past and the recent years and dierence between them. e
spatio-temporal variability in re counts over CI (Fig.4(a,b)) is consistent with the spatial variability in aerosol
loading and AI (Fig.1). Further, the monthly re count dierence (Fig.S6) correlates spatially with the monthly
trend in the number of days with high aerosol loading (Fig.S7), particularly over CI. We have then counted the
Figure 4. (a) Spatial map of 1° × 1° gridded total re counts (FC) for November-February of 2003–2007 (past
years). (b) Same as (a) except that for 2013–2017 (recent years). (c) e dierence between the past and the
recent years. (d) e change in percentage contribution of re counts over CI and IGP for November-February
of 2003 to 2017.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

6
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
www.nature.com/scientificreports/
number of re hotspots for each year over IGP and CI, and calculated the percentage contribution of re counts
from both the regions. Note that the number of 1° × 1° gridded pixels in IGP and CI are equal (n = 75). e rela-
tive percentage of re counts over CI and IGP are plotted in Fig.4(d). e percentage contribution of re counts
over CI has increased at a rate of 20% from 2003 to 2017 indicating that the re counts over CI is higher than that
over IGP in the recent years. us, the greater rate of enhancement in high aerosol loading days over CI in the
recent years (Fig.2(b)) can be associated with relatively large enhancement in re emissions over CI (Fig.4(c))
compared to IGP. Moreover, a recent study show that a large fraction of aerosols from biomass burning activities
over northern western states can also extend over parts of Central India45 as well as northwestern parts of Bay of
Bengal34. erefore, the observed enhancement in the recent aerosol loading over CI can be a combined result of
rapid urbanization, enhanced localized re emissions, and long-range transport of aerosols from the IGP region.
Change in aerosol direct radiative forcing. Aerosol direct radiative forcing counteracts, in part, the
warming due to greenhouse gases10,46, but the eects may vary temporally and spatially. High aerosol direct radi-
ative forcing is usually found over India due to the combined eect of dust, smoke and other aerosols32. In this
section, we present the change in the ADRF at the top of the atmosphere, on the atmosphere and at the surface
using MERRA-2 data. FiguresS8–S10 presents the averaged spatial distribution of the ADRF at the TOA, on
the atmosphere and at the surface, respectively, during the past and the recent years and the dierence between
these two time periods. Nair et al.47, have shown that aerosol forcing at the TOA is modulated mostly by anthro-
pogenic aerosols. It is clear that the increasing trends in aerosol loading over India and adjoining Seas, especially
o the eastern and western coasts, led to increased cooling at the TOA in recent years (Fig.S8). Kaskaoutis et
al.34, have observed signicant fraction of soot aerosols over northwestern BoB during the winter. In fact, the
aerosol-induced atmospheric forcing eciency was found higher for the BoB (31 W/m2) as compared to the AS
(18 W/m2)48, but we nd that ADRF on the atmosphere is higher over the AS in the recent years. It is apparent that
there is increased atmospheric warming (Fig.S8), increasedsurfacecooling (Fig.S9), and increased TOA cooling
(Fig.S10) over India and adjoining Seas owing to increased aerosol loading (Figs.1 and 2) in the recent years.
To illustrate how the forcing has changed over the study period, we calculate ADRF for each region separately
(Fig.S11). e overall atmospheric (positive) and surface (negative) forcing is highest over IGP compared to CI
and adjoining Seas (Fig.S11). But, what is interesting is that the dierence between the recent and the past years
in atmospheric warming over CI (4.50 W/m2) overtakes IGP (2.01 W/m2) (TableS2). Concurrently, the mean dif-
ference in ADRF on the atmosphere between the recent and the past years is higher over the AS (3.67 W/m2) than
that of the BoB (0.48 W/m2), as opposed to previous studies48,49. is suggests that aerosols exerted as much as
seven times more atmospheric warming over the AS than the BoB in recent years. e regionally averaged ADRF
on the surface and at the TOA are also higher over the AS than the BoB (TableS2), contradicting to the previous
study which showed larger values over the BoB for February 200350. ese ndings imply that the increasing re
activity over CI has altered the forcing at much expected level over CI and downwind AS than the polluted IGP.
e enhanced atmospheric heating and surface cooling over CI and the AS can lead to increase in lower tropo-
spheric stability. e stable atmospheric condition would then favor more accumulation of aerosols close to the
surface and can further accelerate occurrence of hazy days (Fig.S12), and thus, creating a positive feedback mech-
anism. Nonetheless, the increase in LTS can itself be inuenced by the increase in aerosol-induced atmospheric
warming51. erefore, this aerosol-LTS coupling can induce a positive feedback cycle on aerosol accumulation in
the boundary layer and enhance aerosol loading over CI and the AS. Moreover, enhanced atmospheric stability
may also enhance low-level cloud amount over the AS and the BoB, leading to a negative feedback on the climate
system in a warming anthropogenic future.
Conclusions
Indian subcontinent, one of the world’s fastest growing regions in terms of urbanization and population, is greatly
vulnerable to particulate pollution. e winter time hazy scenarios and their radiation feedbacks exert a profound
impact on the weather and climate. Using 15 years (2003–2017) of satellite and reanalysis datasets, this study
investigates the trend in the number of hazy days (i.e. days with high aerosol loading) and the aerosol-induced
direct radiation feedbacks on the surface-atmosphere system over India and adjoining Seas for the dry winter
season (November–February).
e major ndings of this study are as follows;
1. Overall, aerosol loading over India and adjoining Seas is rapidly increasing in the recent years.
2. e number of hazy days are increasing at the rate of about 2 days per year over India, with a higher rate
over CI (~2.6 days/year) than that of over IGP (~1.7 days/year).
3. Since the AS is located downwind to the CI, the number of hazy days over the AS is also higher than that
over the BoB during the dry winter season.
4. Collocated similar enhancements in UV aerosol index as that of AOD suggests the dominance of absorb-
ing aerosols over Central India in the recent years.
5. Consequently, aerosol-induced atmospheric warming (4.50 W/m2) and surface cooling (−9.44 W/m2) due
to aerosol direct radiative forcing is highest over CI as compared to other study regions in the recent years.
e enhanced atmospheric warming over CI is about two-fold to that of over IGP (2.01 W/m2).
6. e higher aerosol loading over CI is attributed to the recent increase in biomass burning activities over
the region.
7. Surprisingly, aerosols exerted as much as seven times more atmospheric warming over the AS in the recent
years than over the BoB. is contradicts to majority of previous studies which showed higher atmospheric
warming over the BoB than the AS during the winter season. is nding is substantiated by the mean
changes in wind speed and lower tropospheric stability over CI and the AS.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

7
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
www.nature.com/scientificreports/
Although the high aerosol loading is observed over IGP than other regions, our study reveals that aerosol
loading over CI has escalated greatly in the recent years. us, our ndings provide new insights to better con-
strain aerosols role in the climate over Indian subcontinent.
Methods
In this study, we have used the geographical region bounded by latitude, 7–38°N, and longitude, 66–94°E. is
region further divided into four sub-regions viz., Indo-Gangetic Plain (IGP), Central India (CI), Arabian Sea
(AS) and Bay of Bengal (BoB) (Fig.1b). e columnar composite AOD from Moderate-resolution Imaging
Spectroradiometer (MODIS) onboard AQUA is used. The modeled aerosol species concentration (sulfate,
BC, OC, sea salt, and dust) in terms of their optical depth and radiation uxes are used from the Modern-Era
Retrospective Analysis for Research and Application - version 2 (MERRA-2). ese observations and modelled
values at daily resolution over the time period from 2003 to 2017 for the time period from October through
February are used. Also the composition of aerosols and the meteorological conditions have insignificant
monthly variability during the dry winter season18. TableS1 summarizes data products and their temporal and
spatial resolution.
MODIS. e Moderate-resolution Imaging Spectroradiometer sensor onboard polar orbiting Earth Observing
System (EOS) satellite AQUA ies at an altitude of 705 km, with a swath width of 2330 km and equator crossing
at 13:30 local time. It measures the reected solar radiance and terrestrial emission in a wavelength band ranging
from 0.41–14.4 μm divided into 36 channels, categorized with horizontal resolutions varying between 0.25 and 1
square kilometer52. Here, we have used the level3 Dark Target and Deep Blue combined AOD of collection 6.1 at
0.55 µm at a grid resolution of 1° × 1° 53–55. e estimated maximum uncertainty is approximately ±0.05 × AOD
over Oceans and ±0.15 × AOD over continents. e technical details, algorithm and validation details can be
found in Remer et al.52. e collection 6 MODIS active re location product at 1 km resolution was also used
as proxy for biomass burning hot spots. e product utilizes thermal anomalies in infrared wavelengths (4 and
11 μm)56.
OMI. In addition to MODIS, AURA Ozone Monitoring Instrument (OMI) daily UV aerosol index (UV-AI)
is used (OMAERUVd.003) in this analysis. AURA OMI is a nadir-viewing spectrometer onboard NASA’s Aura
satellite. It measures direct and backscattered solar radiation in the UV-visible range from 264 to 504 nm. e
retrieval technique and validation is given in Bucsela et al.57. UV-AI is based on a spectral contrast method in a
UV region where the ozone absorption is negligible. Positive values of UV-AI indicate absorbing aerosols (smoke
and dust) whereas near-zero or negative values indicate non-absorbing aerosols (sulfate and sea salt) and clouds58.
While, UV-AI is good indicator of absorbing aerosols, its value is dependent on the smoke plume altitude59.
e OMI retrieval algorithm for aerosol detection has been validated with ground-based measurements60. OMI
UV-AI is available from the NASA Goddard Earth Sciences, Data and Information Services Center (GES DISC;
http://disc.sci.gsfc.nasa.gov).
MERRA-2. e Modern-Era Retrospective Analysis for Research and Application - version 261 uses the
Goddard Earth Observing System-5 (GEOS-5) atmospheric general circulation model62 that is coupled with the
Goddard Global Ozone Chemistry Aerosol Radiation and Transport model (GOCART)63. e GOCART model
simulates AOD for ve major aerosol species like OC, BC, sea salt, dust and sulfate using multi-satellite based
(MODIS,AVHRR and MISR) and ground-based (AERONET) AOD data. GEOS-5 model provides the data from
1980 to present in hourly and monthly gridded data with the resolution of 0.5° × 0.625° in latitude and longitude
from the surface to the top layer of 0.01 hPa with 72 vertical levels64. e MERRA-2 simulated AOD compares
well with ground-based and satellite measurements globally64,65. It is freely available from NASA Goddard Earth
Sciences (GES) Data and Information Services Center (DISC) https://disc.gsfc.nasa.gov/. In order to calculate
clear-sky aerosol direct radiative forcing (ADRF) from the radiative uxes from MERRA-2 data65, the dierence
between the radiation ux under clear sky condition in the presence of aerosol and without aerosol is calculated.
e hourly variables; SWGNTCLR (surface net downward shortwave ux assuming clear-sky), SWGNTCLRCLN
(surface net downward shortwave ux assuming clear-sky and no aerosol), LWGNTCLR (surface net down-
ward longwave ux assuming clear sky) and LWGNTCLRCLN (surface net downward longwave ux assuming
clear-sky and no aerosol) are used to calculate ADRF at surface (ADRFSURF). Concurrently, ADRF at the top of
the atmosphere (ADRFTOA) is calculated from hourly variables, SWTNTCLR (TOA net downward shortwave ux
assuming clear sky), SWTNTCLRCLN (TOA net downward shortwave ux assuming clear-sky and no aerosol),
LWTUPCLR (upwelling longwave ux at TOA assuming clear-sky) and LWTUPCLRCLN (upwelling longwave
ux at TOA assuming clear-sky and no aerosol). ese radiation variables can be found in the MERRA-2 prod-
uct, tavg1_2d_rad_Nx. Overall, MERRA-2 simulated radiative uxes agrees well with CERES EBAF Edition 2.8
satellite product over 2001–201566. MERRA-2 simulated both shortwave and longwave radiative uxes are used to
derive the total aerosol direct radiative forcing (ARDF). ADRF at the surface (ADRFSUR) and at top of the atmos-
phere (ADRFTOA) are calculated by the following formula.
=+−+
=+−+
ADRF (SWGNTCLRLWGNTCLR) (SWGNTCLRCLN LWGNTCLRCLN)
ADRF (SWTNTCLRLWTUPCLR) (SWTNTCLRCLN LWTUPCLRCLN)
SUR
TOA
e ADRF on the atmosphere, which indicates the energy trapped by all aerosols in the atmosphere, is calculated
by taking the dierence between ADRF at the TOA and ADRF at the surface.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

8
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
www.nature.com/scientificreports/
GDAS. We used potential temperature at the free troposphere pressure level (i.e. 700 hPa) and the surface
to calculate lower tropospheric stability (LTS) using the NOAA-NCEP (National Oceanic and Atmospheric
Administration National Centers for Environmental Prediction) Global Data Assimilation System (GDAS)
assimilated meteorological datasets67. e variables are available at 1° × 1° spatial resolution with 21 vertical lev-
els (1000 hPa −100 hPa) at synoptic hours. LTS is calculated as; LTS = θ700 − θsurface.
Data availability
Satellite (MODIS-AQUA and AURA-OMI) and model reanalysis (MERRA-2) datasets are freely accessible to the
public from irrespective websites (Refer to TableS1 in the Supplementary Material).
Received: 21 June 2019; Accepted: 29 October 2019;
Published: xx xx xxxx
References
1. Andrews, E. et al. Climatology of aerosol radiative properties in the free troposphere. Atmos. es. 102, 365–393, https://doi.
org/10.1016/j.atmosres.2011.08.017 (2011).
2. Babu, S. S. et al. Trends in aerosol optical depth over Indian region: Potential causes and impact indicators. J. Geophys. es. Atmos.
118(11), 794–711,806, https://doi.org/10.1002/2013JD020507 (2013).
3. Putaud, J. P. et al. A European aerosol phenomenology – 3: Physical and chemical characteristics of particulate matter from 60 rural,
urban, and erbside sites across Europe. Atmos. Environ. 44, 1308–1320, https://doi.org/10.1016/j.atmosenv.2009.12.011 (2010).
4. Moorthy, . ., Satheesh, S. ., Babu, S. S. & Dutt, C. B. S. Integrated Campaign for Aerosols, gases and adiation Budget (ICAB):
An overview. J. Earth Syst. Sci. 117, 243–262, https://doi.org/10.1007/s12040-008-0029-7 (2008).
5. Tripathi, S. N. et al. Measurements of atmospheric parameters during Indian Space esearch Organization Geosphere Biosphere
Programme Land Campaign II at a typical location in the Ganga basin: 1. Physical and optical properties. J. Geophys. es. Atmos.
111, https://doi.org/10.1029/2006JD007278 (2006).
6. Quinn, P. . & Bates, T. S. egional aerosol properties: Comparisons of boundary layer measurements from ACE 1, ACE 2,
Aerosols99, INDOEX, ACE Asia, TAFOX, and NEAQS. J. Geophys. es. Atmos. 110, https://doi.org/10.1029/2004jd004755 (2005).
7. emer, L. A. et al. Global aerosol climatology from the MODIS satellite sensors. J. Geophys. es. Atmos. 113, https://doi.
org/10.1029/2007JD009661 (2008).
8. Holben, B. N. et al. AEONET—A Federated Instrument Networ and Data Archive for Aerosol Characterization. emote Sens.
Environ. 66, 1–16, https://doi.org/10.1016/S0034-4257(98)00031-5 (1998).
9. IPCC. Climate Change 2013: e Physical Science Basis. Contribution of Woring Group I to the Fih Assessment eport of the
Intergovernmental Panel on Climate Change, [Stocer, T. F. et al. (ed.)]. Cambridge University Press, Cambridge, United ingdom
and New Yor, NY, USA, 1535 pp (2013).
10. Storelvmo, T., Leirvi, T., Lohmann, U., Phillips, P. C. B. & Wild, M. Disentangling greenhouse warming and aerosol cooling to
reveal Earth’s climate sensitivity. Nat. Geosci. 9, 286, https://doi.org/10.1038/ngeo2670
11. David, L. M. et al. Aerosol Optical Depth Over India. J. Geophys. es. Atmos. 123, 3688–3703, https://doi.org/10.1002/2017JD027719
(2018).
12. asaoutis, D. G., Badarinath, . V. S.,  harol, . S., Sharma, . A. & ambezidis, H. D. Variations in the aerosol optical properties
and types over the tropical urban site of Hyderabad, India. J. Geophys. es. Atmos. 114, https://doi.org/10.1029/2009JD012423
(2009).
13. Satheesh, S. ., Moorthy, . ., aufman, Y. J. & Taemura, T. Aerosol optical depth, physical properties and radiative forcing over
the Arabian Sea. Meteorol. Atmos. Phys. 91, 45–62, https://doi.org/10.1007/s00703-004-0097-4 (2006).
14. Pan, X. et al. A multi-model evaluation of aerosols over South Asia: common problems and possible causes. Atmos. Chem. Phys. 15,
5903–5928, https://doi.org/10.5194/acp-15-5903-2015 (2015).
15. Prijith, S. S., ajeev, ., ampi, B. V., Nair, S. . & Mohan, M. Multi-year observations of the spatial and vertical distribution of
aerosols and the genesis of abnormal variations in aerosol loading over the Arabian Sea during Asian summer monsoon season. J.
Atmospheric Sol.-Terr. Phys. 105–106, 142–151, https://doi.org/10.1016/j.jastp.2013.09.009 (2013).
16. Burney, J. & amanathan, V. ecent climate and air pollution impacts on Indian agriculture. Proc. Natl. Acad. Sci. USA 111,
16319–16324, https://doi.org/10.1073/pnas.1317275111 (2014).
17. Chameides, W. L. et al. Case study of the eects of atmospheric aerosols and regional haze on agriculture: An opportunity to
enhance crop yields in China through emission controls? Proc. Natl. Acad. Sci. USA 96, 13626–13633, https://doi.org/10.1073/
pnas.96.24.13626 (1999).
18. Latha, ., Murthy, B. S., Lipi, ., Srivastava, M. . & umar, M. Absorbing Aerosols, Possible Implication to Crop Yield - A
Comparison between IGB Stations. Aerosol Air Qual. es. 17, 693–705, https://doi.org/10.4209/aaqr.2016.02.0054 (2017).
19. Sarangi, C., anawade, V. P., Tripathi, S. N., omas, A. & Ganguly, D. Aerosol-induced intensication of cooling eect of clouds
during Indian summer monsoon. Nat. Commun. 9, 3754, https://doi.org/10.1038/s41467-018-06015-5 (2018).
20. Dey, S. & Di Girolamo, L. A decade of change in aerosol properties over the Indian subcontinent. Geophys. es. Lett. 38, https://doi.
org/10.1029/2011GL048153 (2011).
21. Moorthy, . ., Babu, S. S., Manoj, M. . & Satheesh, S. . Buildup of aerosols over the Indian egion. Geophys. es. Lett. 40,
1011–1014, https://doi.org/10.1002/grl.50165 (2013).
22. asaoutis, D. G. et al. Variability and trends of aerosol properties over anpur, northern India using AEONET data (2001–10).
Environ. es. Lett. 7, 024003 (2012).
23. amachandran, S., edia, S. & Srivastava, . Aerosol optical depth trends over dierent regions of India. Atmos. Environ. 49,
338–347, https://doi.org/10.1016/j.atmosenv.2011.11.017 (2012).
24. Srivastava, . Trends in aerosol optical properties over South Asia. Int. J. Climatol. 37, 371–380, https://doi.org/10.1002/joc.4710
(2017).
25. Hsu, N. C. et al. Global and regional trends of aerosol optical depth over land and ocean using SeaWiFS measurements from 1997
to 2010. Atmos. Chem. Phys. 12, 8037–8053, https://doi.org/10.5194/acp-12-8037-2012 (2012).
26. De y, S. et al. Variability of outdoor ne particulate (PM2.5) concentration in the Indian Subcontinent: A remote sensing approach.
emote Sens. Environ. 127, 153–161, https://doi.org/10.1016/j.rse.2012.08.021 (2012).
27. umar, M. et al. Long-term aerosol climatology over Indo-Gangetic Plain: Trend, prediction and potential source elds. Atmos.
Environ. 180, 37–50, https://doi.org/10.1016/j.atmosenv.2018.02.027 (2018).
28. Manoj, M. ., Satheesh, S. ., Moorthy, . ., Gogoi, M. M. & Babu, S. S. Decreasing Trend in Blac Carbon Aerosols Over the
Indian egion. Geophys. es. Lett. 46, 2903–2910, https://doi.org/10.1029/2018GL081666 (2019).
29. Sarangi, C., Tripathi, S. N., anawade, V. P., oren, I. & Pai, D. S. Investigation of the aerosol–cloud–rainfall association over the
Indian summer monsoon region. Atmos. Chem. Phys. 17, 5185–5204, https://doi.org/10.5194/acp-17-5185-2017 (2017).
30. oren, I., aufman, Y. J., osenfeld, D., emer, L. A. & udich, Y. Aerosol invigoration and restructuring of Atlantic convective
clouds. Geophys. es. Lett. 32, https://doi.org/10.1029/2005gl023187 (2005).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

9
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
www.nature.com/scientificreports/
31. Gaur, A., Tripathi, S. N., anawade, V. P., Tare, V. & Shula, S. P. Four-year measurements of trace gases (SO2, NOx, CO, and O3) at
an urban location, anpur, in Northern India. J. Atmos. Chem. 71, 283–301, https://doi.org/10.1007/s10874-014-9295-8 (2014).
32. aufman, Y. J., Tanré, D. & Boucher, O. A satellite view of aerosols in the climate system. Nature 419, 215, https://doi.org/10.1038/
nature01091 (2002).
33. rishna Moorthy, ., Suresh Babu, S., Manoj, M. . & Satheesh, S. . Buildup of aerosols over the Indian egion. Geophys. es. Lett.
40, 1011–1014, https://doi.org/10.1002/grl.50165 (2013).
34. asaoutis, D. G. et al. Extremely large anthropogenic-aerosol contribution to total aerosol load over the Bay of Bengal during
winter season. Atmos. Chem. Phys. 11, 7097–7117, https://doi.org/10.5194/acp-11-7097-2011 (2011).
35. Lawrence, M. G. & Lelieveld, J. Atmospheric pollutant outow from southern Asia: a review. Atmos. Chem. Phys. 10, 11017–11096,
https://doi.org/10.5194/acp-10-11017-2010 (2010).
36. amachandran, S. & Jayaraman, A. Spectral aerosol optical depths over Bay of Bengal and Chennai: II-sources, anthropogenic
inuence and model estimates. Atmos. Environ. 37, 1951–1962, https://doi.org/10.1016/S1352-2310(03)00084-0 (2003).
37. Sanwlani, N., Chauhan, P. & Navalgund, . . Characterization and transport of aerosols over the Bay of Bengal during the winter
monsoon: a comparative study using in-situ and satellite measurements. Int. J. emote Sens. 32, 1253–1267, https://doi.
org/10.1080/01431160903527454 (2011).
38. Han, L., Zhou, W. & Li, W. Growing Urbanization and the Impact on Fine Particulate Matter (PM2.5) Dynamics. Sustainability 10,
1696, https://doi.org/10.3390/su10051696 (2018).
39. Haque, I. & Patel, P. P. Growth of metro cities in India: trends, patterns and determinants. Urban es. Pract. 11, 338–377, https://doi.
org/10.1080/17535069.2017.1344727 (2018).
40. athee, G. In Urbanization in Asia: Governance, Infrastructure and the Environment (eds ala Seetharam Sridhar & Guanghua Wan)
215-238 (Springer India, 2014).
41. Sahu, L. . et al. egional biomass burning trends in India: Analysis of satellite re data. J. Earth Syst. Sci. 124, 1377–1387, https://
doi.org/10.1007/s12040-015-0616-3 (2015).
42. Prasad, V. ., Badarinath, . V. S. & Eaturu, A. Biophysical and anthropogenic controls of forest res in the Deccan Plateau, India. J.
Environ. Manage. 86, 1–13, https://doi.org/10.1016/j.jenvman.2006.11.017 (2008).
43. Vadrevu, . P., Laso, ., Giglio, L. & Justice, C. Vegetation res, absorbing aerosols and smoe plume characteristics in diverse
biomass burning regions of Asia. Environ. es. Lett. 10, 105003, https://doi.org/10.1088/1748-9326/10/10/105003 (2015).
44. Venataraman, C. et al. Emissions from open biomass burning in India: Integrating the inventory approach with high-resolution
Moderate esolution Imaging Spectroradiometer (MODIS) active-re and land cover data. Glob. Biogeochem. Cycles 20, https://doi.
org/10.1029/2005gb002547 (2006).
45. Sarar, S., Singh, . P. & Chauhan, A. Crop esidue Burning in Northern India: Increasing reat to Greater India. J. Geophys. es.
Atmos. 123, 6920–6934, https://doi.org/10.1029/2018jd028428 (2018).
46. Wild, M. et al. From Dimming to Brightening: Decadal Changes in Solar adiation at Earth’s Surface. Science 308, 847–850, https://
doi.org/10.1126/science.1103215 (2005).
47. Nair, V. S., Babu, S. S., Manoj, M. ., Moorthy, . . & Chin, M. Direct radiative eects of aerosols over South Asia from observations
and modeling. Clim. Dyn. 49, 1411–1428, https://doi.org/10.1007/s00382-016-3384-0 (2017).
48. Moorthy, . ., Nair, V. S., Babu, S. S. & Satheesh, S. . Spatial and vertical heterogeneities in aerosol properties over oceanic regions
around India: Implications for radiative forcing. Q. J. oyal Meteorol. Soc. 135, 2131–2145, https://doi.org/10.1002/qj.525 (2009).
49. De y, S., Sarar, S. & Singh, . P. Comparison of aerosol radiative forcing over the Arabian Sea and the Bay of Bengal. Adv. Space es.
33, 1104–1108, https://doi.org/10.1016/S0273-1177(03)00737-3 (2004).
50. Vinoj, V., Babu, S. S., Satheesh, S. ., Moorthy, . . & aufman, Y. J. adiative forcing by aerosols over the Bay of Bengal region
derived from shipborne, island-based, and satellite (Moderate-esolution Imaging Spectroradiometer) observations. J. Geophys.
es. Atmos. 109, https://doi.org/10.1029/2003JD004329 (2004).
51. Sarangi, C., Tripathi, S. N., Mishra, A. ., Goel, A. & Welton, E. J. Elevated aerosol layers and their radiative impact over anpur
during monsoon onset period. J. Geophys. es. Atmos. 121, 7936–7957, https://doi.org/10.1002/2015jd024711 (2016).
52. emer, L. A. et al. e MODIS Aerosol Algorithm, Products, and Validation. J. Atmos. Sci. 62, 947–973, https://doi.org/10.1175/
jas3385.1 (2005).
53. Levy, . C. et al. e Collection 6 MODIS aerosol products over land and ocean. Atmos. Meas. Tech. 6, 2989–3034, https://doi.
org/10.5194/amt-6-2989-2013 (2013).
54. Hsu, N. C. et al. Enhanced Deep Blue aerosol retrieval algorithm: e second generation. J. Geophys. es. Atmos. 118, 9296–9315,
https://doi.org/10.1002/jgrd.50712 (2013).
55. Sayer, A. M. et al. MODIS Collection 6 aerosol products: Comparison between Aqua’s e-Deep Blue, Dar Target, and “merged” data
sets, and usage recommendations. J. Geophys. es. Atmos. 119(13), 965–913,989, https://doi.org/10.1002/2014jd022453 (2014).
56. Giglio, L., Descloitres, J., Justice, C. O. & aufman, Y. J. An Enhanced Contextual Fire Detection Algorithm for MODIS. emote
Sens. Environ. 87, 273–282, https://doi.org/10.1016/S0034-4257(03)00184-6 (2003).
57. Bucsela, E. J. et al. A new stratospheric and tropospheric NO2 retrieval algorithm for nadir-viewing satellite instruments:
applications to OMI. Atmos. Meas. Tech. 6, 2607–2626, https://doi.org/10.5194/amt-6-2607-2013 (2013).
58. Torres, O. et al. Aerosols and surface UV products from Ozone Monitoring Instrument observations: An overview. Journal of
Geophysical esearch: Atmospheres 112, https://doi.org/10.1029/2007jd008809 (2007).
59. Ec, T. F. et al. Characterization of the optical properties of biomass burning aerosols in Zambia during the 1997 ZIBBEE eld
campaign. J. Geophys. es. Atmos. 106, 3425–3448, https://doi.org/10.1029/2000jd900555 (2001).
60. Curier, . L. et al. etrieval of aerosol optical properties from OMI radiances using a multiwavelength algorithm: Application to
western Europe J. Geophys. es. Atmos. 113, https://doi.org/10.1029/2007jd008738 (2008).
61. G e l aro, . et al. e Modern-Era etrospective Analysis for esearch and Applications, Version 2 (MEA-2). J. Climatol. 30,
5419–5454, https://doi.org/10.1175/jcli-d-16-0758.1 (2017).
62. Molod, A., Taacs, L., Suarez, M. & Bacmeister, J. Development of the GEOS-5 atmospheric general circulation model: evolution
from MEA to MEA2. Geosci. Model Dev. 8, 1339–1356, https://doi.org/10.5194/gmd-8-1339-2015 (2015).
63. C hin, M. et al. Tropospheric aerosol optical thicness from the GOCAT model and comparisons with satellite and Sun photometer
measurements. J. Atmos. Sci. 59, 461–483, 10.1175/1520-0469(2002)059<0461:Taot>2.0.Co;2 (2002).
64. Buchard, V. et al. e MEA-2 Aerosol eanalysis, 1980 Onward. Part II: Evaluation and Case Studies. J. Clim. 30, 6851–6872,
https://doi.org/10.1175/jcli-d-16-0613.1 (2017).
65.  andles, C. A. et al. e MEA-2 Aerosol eanalysis, 1980 Onward. Part I: System Description and Data Assimilation Evaluation.
J. Clim. 30, 6823–6850, https://doi.org/10.1175/jcli-d-16-0609.1 (2017).
66. Hinelman, L. M. e Global adiative Energy Budget in MEA and MEA-2: Evaluation with espect to CEES EBAF Data.
J. Clim. 32, 1973–1994, https://doi.org/10.1175/jcli-d-18-0445.1 (2019).
67. David, F. P. & John, C. D. e National Meteorological Center’s Spectral Statistical-Interpolation Analysis System. Mon. Weather ev.
120, 1747–1763, https://doi.org/10.1175/1520-0493(1992)120 (1992).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

10
SCIENTIFIC REPORTS | (2019) 9:17406 | https://doi.org/10.1038/s41598-019-53630-3
www.nature.com/scientificreports
www.nature.com/scientificreports/
Acknowledgements
V.P.K. would like to thank the University Grants Commission, Government of India for Start-Up Grant (Ref. No.
F.4-5/230-FRP/2015/BSR) and Department of Science & Technology (DST)-SERB Grant (ECR/2016/001333).
Satellite (MODIS-AQUA and AURA-OMI) and reanalysis (MERRA-2) datasets used in this study are
acknowledged. Authors are thankful to anonymous reviewers for their critical comments and suggestions that
helped to improve the impact of the paper.
Author contributions
C.S. and V.P.K. conceived the study. A.T. and C.S. did the data analysis. A.T. and V.P.K. wrote the initial
manuscript. All authors reviewed the manuscript.
Competing interests
e authors declare no competing interests.
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/s41598-019-53630-3.
Correspondence and requests for materials should be addressed to C.S. or V.P.K.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons license, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
© e Author(s) 2019
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com