Page 1
Synoptic weather conditions and aerosol episodesover Indo-Gangetic Plains, India
D. G. Kaskaoutis • E. E. Houssos • D. Goto • A. Bartzokas • P. T. Nastos •
P. R. Sinha • S. K. Kharol • P. G. Kosmopoulos • R. P. Singh • T. Takemura
Received: 29 May 2013 / Accepted: 13 January 2014
� Springer-Verlag Berlin Heidelberg 2014
Abstract The present study focuses on identifying the
main atmospheric circulation characteristics associated
with aerosol episodes (AEs) over Kanpur, India during the
period 2001–2010. In this respect, mean sea level pressure
(MSLP) and geopotential height of 700 hPa (Z700) data
obtained from the NCEP/NCAR Reanalysis Project were
used along with daily Terra-MODIS AOD550 data. The
analysis identifies 277 AEs [AOD500 [ AOD500 ? 1ST-
DEV (standard deviation)] over Kanpur corresponding to
13.2 % of the available AERONET dataset, which are
seasonally distributed as 12.5, 9.1, 14.7 and 18.6 % for
winter (Dec–Feb), pre-monsoon (Mar–May), monsoon
(Jun–Sep) and post-monsoon (Oct–Nov), respectively. The
post-monsoon and winter AEs are mostly related to
anthropogenic emissions, in contrast to pre-monsoon and
monsoon episodes when a significant component of dust is
found. The multivariate statistical methods Factor and
Cluster Analysis are applied on the dataset of the AEs
days’ Z700 patterns over south Asia, to group them into
discrete clusters. Six clusters are identified and for each of
them the composite means for MSLP and Z700 as well as
their anomalies from the mean 1981–2010 climatology are
studied. Furthermore, the spatial distribution of Terra-
MODIS AOD550 over Indian sub-continent is examined to
identify aerosol hot-spot areas for each cluster, while the
SPRINTARS model simulations reveal incapability in
reproducing the large anthropogenic AOD, suggesting need
of further improvement in model emission inventories.
This work is the first performed over India aiming to
analyze and group the atmospheric circulation patterns
D. G. Kaskaoutis (&)
Department of Physics, School of Natural Sciences,
Shiv Nadar University, Dadri Tehsil 203207, India
e-mail: [email protected] ;
[email protected]
E. E. Houssos � A. Bartzokas
Laboratory of Meteorology, Department of Physics,
University of Ioannina, 45110 Ioannina, Greece
D. Goto
National Institute for Environmental Studies (NIES),
Tsukuba, Ibaraki 305-8506, Japan
P. T. Nastos
Laboratory of Climatology and Atmospheric Environment,
Faculty of Geology and Geoenvironment, University of Athens,
15784 Zografou, Greece
P. R. Sinha
National Balloon Facility, Tata Institute of Fundamental
Research, ECIL Post 5, Hyderabad 500 062, India
S. K. Kharol
Department of Physics and Atmospheric Science,
Dalhousie University, Halifax, Canada
P. G. Kosmopoulos
Institute for Environmental Research and Sustainable
Development, National Observatory of Athens, Athens, Greece
R. P. Singh
School of Earth and Environmental Sciences, Schmid College
of Science and Technology, Chapman University, Orange,
CA 92866, USA
T. Takemura
Research Institute for Applied Mechanics, Kyusyu University,
Fukuoka, Japan
123
Clim Dyn
DOI 10.1007/s00382-014-2055-2
Page 2
associated with AEs over Indo-Gangetic Plains and to
explore the influence of meteorology on the accumulation
of aerosols.
Keywords Aerosol episodes � Factor–cluster analysis �Weather clusters � Kanpur � India
1 Introduction
Several studies during the last decade have highlighted the
role of aerosols over northern India on radiative forcing,
regional climate, hydrological cycle and monsoon (Lau
et al. 2006; Dey and Tripathi 2007; Gautam et al. 2009a, b,
2010; Srivastava et al. 2011, 2012a). The aerosol climatic
effects over south Asia are still highly uncertain due to
variety of aerosol types, mixing processes in the atmo-
sphere and large spatio-temporal variation (Lawrence and
Lelieveld 2010; Srivastava et al. 2012b). A better under-
standing of the aerosol characteristics and their linkage to
meteorological conditions and atmospheric dynamics needs
further analysis from ground-based instruments, satellite
observations and model simulations as well as examination
of the role of meteorology and weather patterns (Lau et al.
2010; Krishnan et al. 2012).
The northern part of India, Indo-Gangetic Plains
(IGP), is one of the most aerosol laden, densely-popu-
lated and agriculturally productive areas (Kishcha et al.
2011), with a strong intra-annual and intra-seasonal aer-
osol variability (Singh et al. 2004; Dey and di Girolamo,
2010; Henriksson et al. 2011; Kaskaoutis et al. 2011) due
to changes in anthropogenic and natural emissions,
atmospheric and meteorological dynamics (Verma et al.
2012). Formation of fog and haze conditions over the
whole IGP region is common during winter (Gautam
et al. 2007), while in pre-monsoon and early monsoon
seasons IGP is under the influence of frequent dust
storms originated from Thar Desert or transported from
the Arabian Peninsula and Middle East (El-Askary et al.
2006; Prasad and Singh 2007a). During post-monsoon
season, crop residue burning takes place over north-
western IGP and smoke plumes cover the whole region
(Sharma et al. 2010). Natural coarse aerosols are effi-
ciently mixed with fine-mode anthropogenic ones (Dey
and Tripathi 2008; Srivastava and Ramachandran 2012)
forming a thick aerosol layer often called as ‘‘brownish
clouds’’ (Ramanathan et al. 2005) with serious implica-
tions on atmosphere, biosphere, climate and hydrological
cycle.
The significant interrelation of aerosols with the
monsoon system (Bollasina and Nigam 2009; Bhawar
and Devara 2010), drought conditions (Vijayakumar
et al. 2012), Quassi-Biennial Oscillation (QBO) in the
tropopause (Abish and Mohanakumar 2011), El-Nino
Southern Oscillation (Abish and Mohankumar 2013),
local anthropogenic emissions (Reddy and Venkatar-
aman 2002) and long-range transport over Indian sub-
continent (Guleria et al. 2011) motivated us to examine
the synoptic weather conditions that are associated with
aerosol episodes (AEs) over central IGP. Gkikas et al.
(2009) examined the frequency and intensity of AEs
along with their seasonal variability over Mediterranean
Basin by using daily Terra-MODIS observations during
the period 2000–2007. Recently, Gkikas et al. (2012)
classified the synoptic weather conditions that favor the
AEs over the Mediterranean identifying 8 characteristic
weather clusters. Local, regional and synoptic meteo-
rology are important factors for the atmospheric aerosol
lifetime, abundance and properties (Carmona and Alpert
2009; Kaskaoutis et al. 2012a), since they control the
aerosol emissions, convection, transport, dry and wet
deposition (Nastos 2012). Reversely, anthropogenic
aerosols over India have been recognized to have a
significant climatic response to the monsoon system
(Rotstayn and Lohmann 2002; Wang et al. 2009; Bo-
llasina et al. 2011; Ganguly et al. 2012). It is, therefore,
of significant importance to study the synoptic weather
systems that are associated with aerosol extremes over
northern India.
The present study focuses on grouping and analyzing the
synoptic circulation patterns that dominate over south Asia
during AEs over Kanpur (26.5�N, 80.2�E), central IGP, for
the period 2001–2010. The AEs correspond to daily values
of AOD500 [ AOD500 ? 1STDEV (standard deviation), as
obtained from Kanpur-AERONET site. The synoptic pat-
terns on the AE days are grouped objectively in discrete,
characteristic clusters, revealing, in this way, the main
modes of atmospheric circulation favoring accumulation of
aerosols over the region, either locally produced or long-
range transported. The clusters are defined by using the
multivariate statistical methods of Factor (S-mode) and
Cluster Analysis (K-means) on the meteorological maps at
700-hPa atmospheric pressure level over south Asia. This
is the first study conducted over India aiming at associating
severe aerosol episodes with synoptic meteorology clus-
ters. For each cluster the AOD distribution is analyzed via
MODIS observations and aerosol hot-spot areas are
revealed. Furthermore, Spectral Radiation-Transport
Model for Aerosol Species (SPRINTARS) model is applied
to examine the degree of satisfactory simulation of the AEs
over central IGP. The results will help in better under-
standing the strong variations in aerosol properties over
IGP associated with meteorological anomalies and
dynamics of weather.
D. G. Kaskaoutis et al.
123
Page 3
2 Data set
2.1 AERONET data
The initial dataset used is the AOD500 values obtained from
Kanpur AERONET station (Level 2, cloud screened and
quality assured) during the period January 2001–December
2010 for the identification of the AEs, as the days with
daily-mean AOD500 above a critical threshold (AOD ?
1STD), i.e. AOD500 [ 0.928. Except of the AOD500, the
Angstrom exponent (a) and single scattering albedo (SSA)
values from direct and almucantar retrievals via Cimel
(CE-318) sun/sky radiometer are also used following the
uncertainties described elsewhere (Dubovik et al. 2000;
Smirnov et al. 2000). Details of the Kanpur AERONET
station, the atmospheric conditions over the region and
long-term aerosol variability and properties are discussed
by many (Singh et al. 2004; Eck et al. 2010, 2012; Giles
et al. 2011; Kaskaoutis et al. 2012b). The study period
includes 2,095 daily AOD500 values, with a mean of
0.627 ± 0.302, from which 277 have been identified as AE
days.
2.2 NCEP/NCAR re-analysis data
For the identification of the main atmospheric circulation
modes associated with the AEs over IGP, daily values (at
12:00 UTC, ?5:30 IST) of mean sea level pressure
(MSLP) and geopotential heights at 700 hPa (Z700)
obtained from the National Center for Environmental
Prediction/National Center for Atmospheric Research
(NCEP/NCAR) Reanalysis project (Kalnay et al. 1996) are
used. The data set corresponds to all AE days during the
period 2001–2010, at 2.5� 9 2.5� spatial resolution, cov-
ering a broad region of south-central Asia (45�–100�E and
5�–50�N).
The MSLP patterns are used to examine the influence of
near surface pressure systems to the accumulation of
aerosols over Kanpur and/or their transport from sources
near the city. The 700 hPa geopotential height has been
selected to investigate long-range dust transport in the
atmosphere, which although it may take place up to 5 km,
is more pronounced at about 3,000 m (Gautam et al. 2010;
Misra et al. 2012), the average height of this pressure level.
2.3 MODIS data
Terra-MODIS AOD data are available from March 2000
(under cloud free conditions) with good accuracy over dark
surfaces (Levy et al. 2010) by means of two separate
algorithms over land and ocean (Kaufman and Tanre
1998). The aerosol properties are derived by the inversion
of the MODIS-observed reflectance using pre-computed
radiative transfer look-up tables based on aerosol models,
on which cloud-screening procedure was initially applied.
Several validation studies over AERONET land locations
(including Kanpur) reveal that about 72 % of the retrievals
fall within expected uncertainty levels of ±0.05 ±
0.15*AOD (Remer et al. 2008; Levy et al. 2010; Shi et al.
2011), which is an improvement of the previous version
(collection 4). In the present study, the collection 5 (C005)
Level 3 (1� 9 1�) Terra-MODIS AOD550 is used on the AE
days over Indian sub-continent and adjoining oceanic
regions (3–32�N, 57–94�E) following the dark-target
approach with lack of data over the deserts (Levy et al.
2007) in order to reveal the aerosol hot-spot areas.
2.4 SPRINTARS model
The global three-dimensional aerosol model Radiation-
Transport Model for Aerosol Species (SPRINTARS)
Spectral has been used to simulate aerosols over Kanpur on
the AE days. SPRINTARS (Takemura et al. 2000, 2005,
2009) is implemented in both an atmospheric general cir-
culation model (AGCM) developed by the University of
Tokyo, National Institute for Environmental Studies and
the Japan Agency for Marine-Earth and Technology (K-1
Developers 2004; Watanabe et al. 2010) and, a radiation
transfer with a k-distribution scheme, MSTRN-8, (Nakaj-
ima et al. 2000; Sekiguchi and Nakajima 2008). The model
simulates AOD for various aerosol components (Black
Carbon (BC), Organic Carbon (OC), sulfate, soil dust and
sea salt) with prescribed optical properties, while the
simulations were conducted with nudged meteorological
fields from NCEP/NCAR reanalysis (wind field, water
vapor and temperature) with the T106 horizontal resolution
(1.1� 9 1.1�). Emission processes for dust with 6 bins
ranging from 0.1 to 10 lm are online-calculated in the
model according to an empirical relation by Gillette
(1978), depending on near-surface wind speed above a
threshold of 6.5 ms-1, vegetation, leaf area index, soil
moisture and snow amount (Takemura et al. 2009; Goto
et al. 2011a). For sea salt, emission process is a function of
wind speed using Monahan’s parameterization (Monahan
et al. 1986) and is online-calculated with 4 bins ranging
from 0.1 to 10 lm (Takemura et al. 2009). For the
anthropogenic components (BC, OC and sulfates) emission
processes and inventories are described by Diehl et al.
(2012) and are widely used in the AeroCom project (http://
aerocom.met.no/). Recently, SPRINTARS was used to
simulate aerosol emission and properties over India (Goto
et al. 2011b, c; Kaskaoutis et al. 2012c) and, in the present
work, model simulations of various aerosol parameters
(AOD, a, SSA, mass column loading, mass concentration)
have been performed over Kanpur aiming to investigate the
capability of the model in reproducing the AEs. The
Synoptic weather conditions
123
Page 4
SPRINTARS AODs have been compared with various
satellites including MODIS under AeroCom project (Kinne
et al. 2006), while the dust loading, which dominates in the
current simulations, was also compared with various
measurements under the same project (Huneeus et al.
2011). Both studies revealed that SPRINTARS is well
within the uncertainty of the other global aerosol models
and have generally agreement with observations.
3 Factor and cluster analysis
At first, a data matrix of daily atmospheric circulation at
Z700 level is constructed. It consists of 437 columns cor-
responding to the 437 grid points, covering south-central
Asia and, 277 rows corresponding to the 277 AE days over
IGP during the period 2001–2010. In order to reduce the
dimensionality of the data matrix, Factor Analysis (Jolliffe
1986; Manly 1986) is carried out to group objectively time-
series highly correlated to each other. The original time-
series are substituted by new, fewer, characteristic and
uncorrelated ones, the Factors, consisting of the so-called
factor scores. It is found that the 437 initial grid point time-
series can be reduced to eight (8) Factors only, accounting
for 88 % of the total variance. Thus, a new matrix
(277 9 8) is constructed, in which the columns correspond
to the 8 Factors’ time-series, and the rows correspond to the
277 daily atmospheric circulations described by the 8
factor scores values only, instead of the 437 Z700 values.
In order to classify objectively the 277 atmospheric
circulations into homogeneous clusters K-Means Cluster
Analysis (Sharma 1996) is applied on the 277 9 8 factor
scores matrix. To decide the optimum number of clusters,
so that they are as much as possible homogeneous and
discrete, the ‘‘Jump’’ criterion (Sugar and James 2003) is
used. Thus, the 277 atmospheric circulations are classified
into 6 clusters. Detailed information about the Factor and
Cluster analysis techniques can be found elsewhere
(Houssos and Bartzokas 2006; Houssos et al. 2011). For
each cluster, the mean patterns of MSLP and Z700 are
drawn. Furthermore, MSLP and Z700 anomalies from the
30-year average (1981–2010, NCEP/NCAR) are calculated
and presented, in order to reveal the most significant fea-
tures of the synoptic patterns associated with high aerosol
loads over Kanpur.
It has to be mentioned that both multivariate statistical
methods are applied only to the Z700 data matrix and not
to the MSLP one in order to avoid problems resulting from
the high altitude of the Tibetan Plateau ([4,000 m). Spe-
cifically, the MSLP field over Tibet could be characterized
as inaccurate and this would lead to erroneous groupings in
the application of Factor Analysis. On the other hand, the
700 hPa level is relatively close to most places of the
Tibetan Plateau and the atmospheric circulation presented
is far more realistic. Thus, on the MSLP pattern of each
cluster we will focus on atmospheric circulation outside
Tibet.
4 Results and discussions
4.1 Temporal distribution of aerosol episodes
The frequency of occurrence of AE days over Kanpur
during the period 2001–2010 is shown in Fig. 1. A large
gap in AOD data series during pre-monsoon and monsoon
seasons of 2004 as well as during winter and monsoon of
2006 and monsoon to winter of 2007 occurs due to tech-
nical issues and calibration of the instrument; as a result,
low frequency of AE is observed. The 277 AEs (13.21 %
of the total 2,095 daily AOD observations) exhibit higher
frequency during post-monsoon (78 cases, 18.6 %) and
monsoon (76 cases, 14.7 %) seasons, while winter (65
cases, 12.5 %) and pre-monsoon (58 cases, 9.1 %) exhibit
lower absolute and percentage frequencies. The frequency
of occurrence of AEs presents considerable intra-annual
variations as a result of changes in atmospheric and
meteorological conditions, formation of hazy and foggy
conditions during winter, boundary-layer dynamics and
variability in dust outflows. The most important finding is
the frequent AEs during monsoon of 2002 (July) and 2003
(June) and during post-monsoon/winter of 2008 (Novem-
ber/December). The reason for these AEs is the drought
and prolonged dry conditions over the northwestern India
responsible for higher frequency in dust events and longer
aerosol lifetime due to deficit of rainfall (Kaskaoutis et al.
2012c). The second case corresponds to persistent dense
Fig. 1 Frequency of occurrence for AOD peaks for each season and
year at Kanpur AERONET station. There is lack of data during the
period pre-monsoon to monsoon of 2004, monsoon and winter of
2006 and monsoon to winter of 2007. [winter: Dec–Feb; pre-
monsoon: Mar–May; monsoon: Jun–Sep; post-monsoon: Oct–Nov]
D. G. Kaskaoutis et al.
123
Page 5
smoke plumes originated from extensive agriculture crop-
residue burning in the northwestern IGP (Punjab state)
causing significant heating in the middle (700 hPa level)
troposphere over the region (Badarinath et al. 2009). In
general, the occurrence of extreme AODs over IGP during
pre-monsoon and monsoon is higher in the first half of the
study period, while during post-monsoon and winter in the
second half. The former influences the neutral-to-declining
AOD decadal trend observed over Kanpur (Kaskaoutis
et al. 2012b), Delhi (Lodhi et al. 2013) and northern India
(Dey and di Girolamo 2011; Kaskaoutis et al. 2011) in late
pre-monsoon and monsoon seasons, while the latter is
consistent with the significant increase in anthropogenic
AOD over IGP (Dey and di Girolamo 2011; Kaskaoutis
et al. 2011), in close consideration with the strong increase
in BC, SO2, and OC emissions (Lu et al. 2011).
4.2 Classification of the atmospheric circulation
patterns and the corresponding anomalies
Using the methodology discussed in Sect. 3, the 277
atmospheric circulation patterns have been classified into 6
clusters. Figure 2 shows the monthly variation of the fre-
quency of occurrence for each cluster. The mean composite
maps of MSLP and Z700 for each cluster (Figs. 3, 4, 5, 6,
7, 8) reveal the main characteristics of the atmospheric
circulation that are associated or being favorable for aer-
osol accumulation over IGP. For a further elucidation of
the results, the corresponding anomalies maps at MSLP
and Z700 from the 30-years (1981–2010) mean climato-
logical conditions are also presented. This is a common
approach for such climatological studies followed by many
researchers (Kaskaoutis et al. 2012a, c; Nastos 2012).
Furthermore, Kishcha et al. (2012) showed that the oscil-
lations (anomalies) in the wind field (direction, speed and
convergence) may be linked to aerosol trends over coastal
region of Bay of Bengal (BoB) without variations in aer-
osol emissions.
4.2.1 Cluster 1 (62 days—22 %)
Cluster 1 comprises 62 out of the 277 AE days (22 %).
Most of the days are distributed in the post-monsoon sea-
son but there are also a few in April (Fig. 2). The mean
circulation of Cluster 1 at Z700 (Fig. 3b) is characterized
by an intense ridge extended from Africa to the Arabian
Peninsula, while over Kazakhstan, northwestern China and
western Mongolia the atmospheric circulation is almost
zonal. Over India, the MSLP (Fig. 3a) patterns present
almost constant, relatively high values (above 1,010 hPa),
responsible for the lack of significant air flow and favoring
subsidence and atmospheric stability. The consequent dry
conditions, along with the intensification of the agriculture
burning in Punjab in October–November, contribute to
accumulation of fine-mode aerosols in the lower tropo-
sphere over IGP (Table 1). The corresponding anomalies
maps (Fig. 3c, d) show both negligible differences from the
30-year mean over India but higher than usual values over
Iran and the Caspian Sea. These positive anomalies indi-
cate an extension of the Arabian Peninsula ridge to the
north, increasing pressure gradient and winds there and not
to the east, over India, where calm conditions favor aero-
sols accumulation.
4.2.2 Cluster 2 (55 days—20 %)
This Cluster is more frequent in post-monsoon and winter
seasons, as well as during April and May (Fig. 2). Simi-
larly to Cluster 1, the constant geopotential height field
with relatively high values over India (Fig. 4b) favors the
accumulation of fine-mode aerosols over IGP (Table 1).
The main difference from Cluster 1 is the weakening of the
ridge over the Arabian Peninsula with a parallel extension
towards India and the appearance of another ridge over
western China and Kazakhstan. The latter is reflected in the
positive values of the Z700 anomalies map (Fig. 4d) over
the same area. On the surface (Fig. 4a), atmospheric
pressure is constant over India with higher values than in
Cluster 1. The anomalies (Fig. 4c) show positive values
over the northeast part of India and neutral-to-slight neg-
ative ones over the Arabian Peninsula and the Middle East.
Although the mean composite maps of Clusters 1 and 2
present similarities, especially over IGP, the stronger
anticyclonic circulation, dominating in the northern partsFig. 2 Monthly variation of the frequency of occurrence for each
atmospheric circulation cluster
Synoptic weather conditions
123
Page 6
of the study region in Cluster 2, differentiates them sig-
nificantly. However these differences are found away from
IGP and do not seem to affect considerably the aerosol load
and aerosol type over Kanpur (Table 1).
4.2.3 Cluster 3 (38 days—14 %)
This atmospheric circulation occurs during winter, mainly
in January, while in some cases it may occur during late
pre-monsoon and post-monsoon seasons (Fig. 2). The main
characteristic that differentiates this Cluster from the pre-
vious two is the intense dipole of high Z700 values over
southern Arabian Peninsula and low ones over Kazakhstan
(Fig. 5b), and the consequent north–south high pressure
gradient, causing an intense westerly air flow over Iran,
Afghanistan and Turkmenistan, which however, does not
affect much IGP. Nevertheless a weak northeast-southwest
gradient over India could be responsible for aerosol outflow
over northern BoB (Kharol et al. 2011). In the Z700 field,
the high negative anomalies over central Asia and the
positive ones extending from Southern Arabia to Tibetan
Plateau via India (Fig. 5d), are coherent with the dipole of
low and high Z700 values. These positive anomalies are
associated with the extension of the ridge towards the
Arabian Sea (AS) and India causing subsidence of air
masses and favoring the accumulation of aerosols over
central IGP. The atmospheric circulation of Clusters 1, 2
and 3, with the calm conditions over IGP, favors the
accumulation of pollutants over Kanpur region.
4.2.4 Cluster 4 (48 days—17 %)
This Cluster is highly frequent during the beginning of the
monsoon season, mainly in June (Fig. 2), with its main
characteristic being the well-organized Indian Low near the
surface, centered over west India and eastern Pakistan and
Fig. 3 For Cluster 1, identified for the AE days over Kanpur during
the period 2001–2010, the composite mean maps of a Mean Sea Level
Pressure (hPa), b 700 hPa Geopotential Heights (gpm), and compos-
ite anomalies of c Mean Sea Level Pressure (hPa) and d 700 hPa
Geopotential Heights (gpm) from the mean climatology during the
period 1981–2010. [Green-to-red and blue-to-violet correspond to
positive and negative anomalies, respectively]
D. G. Kaskaoutis et al.
123
Page 7
extended up to Iran and the Arabian Peninsula (Figs. 6a, c).
The cyclonic circulation over India, near the surface
(Fig. 6a), induces a strong southwesterly air flow, favoring
the transport of dust aerosols over IGP from the neigh-
boring Thar Desert and even from the southern parts of the
Arabian Peninsula mixed with marine particles. The
cyclonic anomaly over the Caspian Sea (Z700, Fig. 6d)
does not affect the south Asian conditions.
4.2.5 Cluster 5 (51 days—19 %)
The highest frequency in this Cluster is observed during the
late pre-monsoon and early monsoon seasons (Fig. 2). The
atmospheric circulation resembles that of Cluster 4, but the
Indian Low in this Cluster is weaker (Fig. 7a). Further-
more, a secondary low in MSLP centered over eastern
India is evident. Similarly to Cluster 4, the uplift and
transport of dust over IGP either from Thar Desert or, in a
lesser degree, from Arabia is favored. Positive MSLP and
Z700 anomalies influence most of the arid regions in
southwest Asia, the AS and the western Tropical Indian
Ocean, while negative anomalies cover the northern and
eastern parts of India and BoB (Fig. 7c, d). The dipole of
anomalies over Indian Ocean has been associated by
Gadgil et al. (2003) via anomalies in OLR (Outgoing Long-
wave Radiation) to deficit of monsoonal rainfall and
obviously to high AOD values. Furthermore, Ghude et al.
(2011) found westerly wind anomalies of 10 ms-1 at
850 hPa over IGP associated with anticyclonic vorticity,
which prevents convection and leads to the weakening of
the monsoon with a negative deviation in rainfall
(*10 mm/day) over western Ghats, central India and
western IGP in July 2002. Manoj et al. (2011) analyzed the
monsoon intra-seasonal oscillations via anomalies in OLR
and found that long break spells followed by active spells
correspond to atmospheric circulations that favor the
accumulation of aerosols over continental India. Some of
these spells (July 2002, June 2009) are associated with
persistent AE days over Kanpur belonging to Cluster 5.
Furthermore, Kaskaoutis et al. (2012c) examined the
anomalies in rainfall and atmospheric circulation associ-
ated with significant accumulation of aerosols over IGP
Fig. 4 As in Fig. 3, but for Cluster 2
Synoptic weather conditions
123
Page 8
during May–June 2003 and July 2002, highlighting posi-
tive MSLP and Z700 anomalies implying the weakening of
cyclonic circulation, monsoon air flow and rainfall.
4.2.6 Cluster 6 (23 days—8 %)
This is the least frequent Cluster occurring mainly in Jan-
uary (Fig. 2). The almost constant MSLP field over India
(Fig. 8a) favors the accumulation of aerosols over IGP,
which are mainly generated by local sources and anthro-
pogenic activities (Kar et al. 2010). Although the synoptic
conditions over India resemble the ones of Clusters 1, 2 and
3, there is a difference at Z700; namely, a rather strong
zonal flow over IGP (Fig. 8b) favors the transport of
aerosols over eastern IGP and Bangladesh. The MLSP and
Z700 values are much lower in Cluster 6 than those in
Cluster 3 and, therefore, large negative anomalies in both
MSLP and Z700 are shown over the northern India and
Tibetan Plateau accompanied with positive ones over
central-western Asia (Fig. 8c, d). The negative anomalies
at Z700 over Tibetan Plateau enhance the westerly flow
over IGP, which accumulates aerosols over the easternmost
part of the Ganges Basin. The anomalous trough at Z700
over northwestern India associated with a small ridge over
eastern IGP and Bangladesh also helps in accumulation of
pollutants over this region. This is verified from the highest
AODs over central-eastern IGP (Table 1; Fig. 10) com-
pared to those found for the other late post-monsoon and
winter Clusters (1, 2 and 3).
4.3 Aerosol distribution
Aerosols over south Asia exhibit substantial spatio-tem-
poral variability, mainly controlled by population density
and local anthropogenic emissions, convection and
boundary-layer dynamics, transport pathways, local and
regional meteorology and apparently by duration and
intensity of the monsoon rainfall (Dey and di Girolamo
2010; Aloysius et al. 2011). The mean Terra-MODIS
AOD550 distribution over Indian sub-continent for each
cluster is shown in Fig. 9. Kanpur is denoted by a star and
exhibits AOD values ranging from 1.15 (Cluster 3) to 1.27
Fig. 5 As in Fig. 3, but for Cluster 3
D. G. Kaskaoutis et al.
123
Page 9
(Cluster 4), which are in satisfactory agreement with those
of AERONET (Table 1).
Clusters 1, 2 and 3 show large similarities in the mean
AODs and spatial distribution due to similar synoptic
conditions (Figs. 3, 4, 5). The highest AODs are observed
in central IGP while in the rest of continental India low
values (below 0.4) are found. The oceans are low aerosol-
laden areas, except of the southeastern AS, coastal and
northern BoB, which are strongly influenced by the conti-
nental outflow (Kaskaoutis et al. 2011). Despite the general
similarities some differences can be detected, which could
be explained by the local emissions. The high AODs in
Cluster 1 are somewhat shifted towards northwestern IGP,
where significant crop residue burning occurs in post-
monsoon season contributing to the accumulation of fine-
mode aerosols, influencing the high a values for Cluster 1
(Table 1), and resulting to large negative forcing values at
surface and significant atmospheric heating (Sharma et al.
2010, 2012; Mishra and Shibata 2012). Clusters 2, 3 and 6
are mainly associated with AEs occurring during winter,
when aerosols are accumulated in the lower-elevated
Ganges basin (di Girolamo et al. 2004). Note that in Cluster
6 the prevailing zonal flow at Z700 (Fig. 8b) is responsible
for the transport of aerosols through central IGP up to
Bangladesh. Several studies over IGP (Komppula et al.
2012; Srivastava et al. 2012c; Misra et al. 2012) revealed
the presence of a thick aerosol layer near the ground during
late post-monsoon and winter seasons composed of a large
fraction of anthropogenic aerosols and BC from fossil and
bio-fuel burning. The aerosol properties over Kanpur are
similar for Clusters 2, 3 and 6; little higher AOD and lower
a are shown for Cluster 6 (Table 1).
The mean AOD550 spatial distribution for Clusters 4 and
5 reveals a large area of very high values, extending from
Pakistan and northernmost AS to eastern IGP. This is a
result of dust outbreaks in Thar Desert, Arabia and Middle
East (Kuhlmann and Quaas 2010; Gautam et al. 2011) and
the dust transport due to dominant wind flow from western
directions (Figs. 6b, 7b) in late pre-monsoon and early
monsoon season. Some extreme AODs shown over North
Indian Ocean (NIO) are attributed to intense dust outbreaks
from Arabian Peninsula in the beginning of 2000s
Fig. 6 As in Fig. 3, but for Cluster 4
Synoptic weather conditions
123
Page 10
(i.e. 2001–2003) (Kaskaoutis et al. 2012c), while there is
lack of sufficient dataset after 2005. The enhanced dust
concentration leads to the highest AODs and lowest a over
Kanpur for Clusters 4 and 5 (Table 1).
The current analysis shows that the AEs over Kanpur
occur under favorable meteorological and atmospheric
circulation conditions, enhanced subsidence, increased
biomass or fossil fuel combustion and transport of dust
plumes. MODIS and AERONET AODs are satisfactory
correlated on cluster-mean basis (Table 1) and the same
has been found by numerous validation studies over Kan-
pur (Tripathi et al. 2005; Jethva et al. 2007; Prasad and
Singh 2007b; Shi et al. 2011).
A crucial aspect that needs the synergy of satellite
remote sensing is the investigation of how much repre-
sentative the AEs over Kanpur are for a greater area over
IGP and India. In this respect, the correlation coefficient
(r) between MODIS-AODs obtained over Kanpur (pixel
centered on 26.5�N, 80.5�E) and the AODs over the other
pixels in the spatial domain shown in Fig. 9 was calculated.
Such an analysis reveals the degree to which the temporal
variability of extreme AODs over Kanpur is related to that
of aerosol load over other regions in the Indian sub-con-
tinent. For neighboring to Kanpur regions, high positive r
values indicate, there too, AEs of similar type, sources and
transportation on the same day. For remote regions, high
positive r values do not necessarily imply AEs there;
nevertheless they do indicate an aerosol teleconnection
pattern, most probably due to similar atmospheric circu-
lation effects over both regions. High negative r values
(anti-correlation) identify a see-saw aerosol teleconnection
pattern, viz. an increase in extreme AOD values over
Kanpur corresponds to a decrease of AOD values over the
remote region and vice versa. Finally, neutral r values
suggest that the variability of aerosol load over Kanpur is
not related to the one over other regions. Possible reasons
for that may be inconsistency in the sources, emission
rates, long-range transport or deposition (wet and dry)
processes.
Figure 10 shows the spatial distribution of the r values
for each cluster, while the grey pixels correspond to
insufficient dataset for performing the regression analysis,
Fig. 7 As in Fig. 3, but for Cluster 5
D. G. Kaskaoutis et al.
123
Page 11
i.e. data lesser than one-third of the AE days of a cluster.
The analysis revealed that the correlations are statistically
significant at 95 % confidence level (p \ 0.05) for values
of r above 0.45–0.50 depending on the degree of freedom
for each individual dataset. Independently of the cluster,
the results show that the AOD values during AEs over
Kanpur are well correlated (r [ 0.6) with those over the
whole IGP region (Fig. 10), indicating consistency in the
sources. On the other hand, some pockets of very high r
values are found at various regions for each cluster
indicating an aerosol teleconnection pattern. More specif-
ically, the high r values for Cluster 1 detected for pixels
located northwestern of Kanpur indicate the source con-
tribution of the Punjab agriculture fires in post-monsoon.
For Clusters 2 and 3 moderate-to-high correlation also
exists between Kanpur and eastern IGP/northern BoB,
which are in the downwind direction for pollution outflow
in the winter season (Kharol et al. 2011). Notable differ-
ences are observed for Clusters 4 and 5, where very high r
values ([0.6–0.7) cover nearly the whole northern India,
Fig. 8 As in Fig. 3, but for Cluster 6
Table 1 Frequency of the AE days for each cluster, along with AERONET-AOD500, MODIS AOD550 and Angstrom exponent values for
various wavelength ranges
Cluster Frequency AERONET AOD500 MODIS AOD550 a (440–870) a (380–500) a (675–870)
1 62 (22 %) 1.17 ± 0.22 1.02 ± 0.27 1.17 ± 0.32 0.81 ± 0.18 1.34 ± 0.44
2 55 (20 %) 1.17 ± 0.24 1.01 ± 0.30 1.02 ± 0.49 0.69 ± 0.31 1.17 ± 0.59
3 38 (14 %) 1.15 ± 0.19 0.91 ± 0.27 0.94 ± 0.46 0.69 ± 0.27 1.06 ± 0.58
4 48 (17 %) 1.27 ± 0.37 1.05 ± 0.70 0.40 ± 0.43 0.39 ± 0.34 0.42 ± 0.46
5 51 (19 %) 1.25 ± 0.32 1.17 ± 0.58 0.41 ± 0.47 0.38 ± 0.34 0.45 ± 0.52
6 23 (8 %) 1.23 ± 0.27 1.12 ± 0.34 0.90 ± 0.56 0.57 ± 0.36 1.07 ± 0.64
Synoptic weather conditions
123
Page 12
D. G. Kaskaoutis et al.
123
Page 13
northern AS and areas in Pakistan. This highlights the
consistency of the aerosol sources and dust transport
pathways from Arabia and Thar Desert over central IGP
during the May–June period. In contrast, large negative
correlations are found over southern AS, NIO and BoB,
revealing a see-saw teleconnection between these regions
and IGP. A comparison of Figs. 9 and 10 suggests that the
high AODs over southern latitudes are not consistent with
the AEs over Kanpur and are related to intensification/
weakening of dust transport from Arabia towards southern
AS and NIO and a synchronous weakening/intensification
of dust transport from Arabia and Pakistan to IGP. Finally,
the high r values over AS for Cluster 6 indicate a triple
aerosol teleconnection pattern between IGP, AS and east-
ern BoB, during winter.
4.4 Model simulations
This section provides an insight in SPRINTARS simula-
tions concerning aerosol properties over Kanpur
(1.1� 9 1.1� spatial resolution) and examines the degree
that SPRINTARS can reproduce the AERONET parame-
ters during severe AE days. On the other hand, there is a
compelling need for improving the aerosol inventories over
the region in order to make realistic assessment of the
impacts of aerosols on radiative forcing and the south
Asian monsoon. Figure 11 shows the SPRINTARS simu-
lated AOD550 time-series during the period 2001–2010
over Kanpur (total of 3,652 days) superimposed with the
AE days with different colored shape depending on cluster.
According to SPRINTARS, only 34 days (0.93 %) with
AOD550 [ 0.9 were found mostly during May–July period,
indicating a systematic underestimation in AOD compared
to AERONET and MODIS, especially during the winter
season. The results show that the majority of the AOD
peaks (except of years 2004 and 2008) correspond to AEs,
whereas nearly the whole AEs during post-monsoon and
winter are associated with very low SPRINTARS AODs
indicating incapability of the model to reproduce them.
Similarly, most GCMs (General Circulation Models) fail to
reproduce the high AODs over IGP, especially during
winter (Chin et al. 2009; Ganguly et al. 2009).
The analysis revealed that SPRINTARS underestimates
significantly the AERONET AOD values with their dif-
ferences being as high as 0.8–1.0 during the post-monsoon
and winter seasons; only on some few cases in late pre-
monsoon and monsoon SPRINTARS AODs are
comparable to those of AERONET. The low correlation
coefficient (r = 0.49, R2 = 0.24) is characteristic for the
model’s incapability in simulating the high AODs over
IGP. Similar results were obtained from the GOCART
simulations over Kanpur with R value of 0.18 (Chin et al.
2009). Significant differences were also found for a440–870
values, despite the qualitative agreement between AER-
ONET and SPRINTARS concerning the much lower val-
ues in Clusters 4 and 5 (Tables 1, 2). SPRINTARS SSA
simulations (Table 2) were not found to be correlated with
AERONET retrievals, suggesting an incapability of the
model in reproducing the absorbing nature of aerosols,
despite the consistency in the mean values (0.90 ± 0.03 for
AERONET and 0.91 ± 0.04 for SPRINTARS). It should
be noted that the SSA at 550 nm from AERONET was
obtained via interpolation from the values at 440 and
675 nm, which adds more uncertainty in the retrievals.
SPRINTARS was found to systematically overestimate aand SSA in Clusters 4 and 5 dominated by desert dust,
suggesting that the model assumes dust particles of lower
size and absorption.
The different aerosol components that contribute to the
total AOD, mass column loading and surface mass con-
centration over Kanpur are summarized in Table 2 for each
Cluster. For all the three parameters a clear dominance of
the dust component is shown, which is increased for the
AEs in Clusters 4 and 5. Thus, according to SPRINTARS
simulations, dust contributes 59–89 % to the total AOD
(depending on cluster), while its contribution to column
mass loading ranges from 75 to 94 % and to surface con-
centration from 35 to 85 %. Given the extremely low
contribution of sea salt to all of these parameters (see
Table 2), SPRINTARS results suggest that the anthropo-
genic contribution (BC ? sulfate ? OC) to total AOD,
mass loading and concentration is very low, even for the
winter season (Clusters 2, 3, 6), when anthropogenic
aerosols dominate over Kanpur (Singh et al. 2004; Dey and
Tripathi 2008). However, the model’s results show that
anthropogenic AOD is higher in Cluster 6, compared to
that in Clusters 1, 2 and 3 (Table 2), which is in agreement
with AERONET and MODIS (Table 1). It is worth to be
noted the very low contribution of BC to all aerosol
parameters (2.4 % for AOD550, 0.8 % for column loading,
3.7 % for mass concentration) compared to that found from
measurements (*9–11 % for AOD and *6–9 % for sur-
face mass concentration) over IGP (Dey and Tripathi 2007,
2008; Goto et al. 2011b; Kharol et al. 2012). The signifi-
cant model underestimation of the anthropogenic emissions
and their higher contribution to atmospheric aerosols in
post-monsoon and winter is the reason for the low AOD,
aerosol loading and concentration during these seasons.
Systematic SPRINTARS underestimation of particulate
matter (PM10 and PM2.5) and BC, especially during winter,
b Fig. 9 Spatial distribution of the mean Terra MODIS AOD550 over
Indian sub-continent and adjoining oceanic regions for the six
atmospheric circulation clusters. The white areas correspond to lack
of data (bright surfaces) or insufficient number of valid pixels. The
Kanpur site is denoted by a star
Synoptic weather conditions
123
Page 14
D. G. Kaskaoutis et al.
123
Page 15
was also found over Dibrugarh in Brahmaputra valley
(Pathak et al. 2013). Recent studies over India (Goto et al.
2011a, b) revealed better model simulations with measured
BC concentrations, AOD, SSA and a by using the
SPRINTARS assumed (Streets et al. 2003; Takemura et al.
b Fig. 10 Spatial distribution of the correlation coefficient values
obtained from the linear regression between the daily MODIS AODs
over Kanpur and the AODs over the remaining pixels for each
atmospheric circulation cluster. The luck of sufficient days for the
regression analysis, i.e. less than the one-third of the frequency for
each cluster, corresponds to gray
Fig. 11 SPRINTARS model
simulations of AOD550 time-
series over Kanpur during the
period Jan 2001–Dec 2010. The
colored shapes indicate the AE
days for each atmospheric
circulation cluster, as identified
from the AERONET AOD500
time-series
Table 2 SPRINTARS model simulations for various aerosol parameters over Kanpur for the AE days classified in the six atmospheric
circulation clusters
Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Cluster 6
AOD550 0.109 0.182 0.126 0.379 0.478 0.217
a(440–870) 1.210 1.248 1.076 0.632 0.639 1.324
SSA550 0.891 0.884 0.897 0.944 0.939 0.902
AOD BC 0.0054 0.0058 0.0049 0.0061 0.0057 0.0074
AOD Su 0.0202 0.0224 0.0198 0.0386 0.0319 0.0487
AOD OC 0.0167 0.0177 0.0168 0.0166 0.0154 0.0343
AOD SS 0.0002 0.0002 0.0002 0.0005 0.0006 0.0003
AOD Du 0.0665 0.1345 0.0852 0.3274 0.4329 0.1304
Load BC 1.95 2.10 1.78 2.22 2.06 2.67
Load Su 11.86 12.21 9.51 22.34 18.87 14.18
Load OC 10.32 10.70 9.56 10.11 9.47 13.63
Load SS 0.17 0.22 0.19 0.44 0.65 0.24
Load Du 75.01 164.31 99.15 402.60 524.37 173.88
Conc. BC 3.31 3.63 3.45 2.04 2.30 3.46
Conc. Su 5.88 6.43 5.06 9.15 8.36 5.85
Conc. OC 16.19 16.22 16.22 9.55 10.74 16.38
Conc. SS 0.01 0.03 0.01 0.09 0.16 0.02
Conc. Du 13.91 38.12 20.31 101.72 122.53 32.94
Mass column loading in lgm-2, mass concentration in lgm-3. Su sulfate, OC organic carbon, SS Sea salt, Du Dust
Synoptic weather conditions
123
Page 16
2005) aerosol emission inventories rather than those used
in the AeroCom project. It is, therefore, needed to further
modify and improve the emissions, transport, mixing and/
or removal processes, especially for the anthropogenic
components, during aerosol life cycle over IGP to better
simulate the direct and indirect aerosol effects on local
monsoon system.
5 Conclusions
The present study classified the atmospheric circulations
(synoptic meteorological systems) associated with aerosol
episodes (AEs) over Kanpur, India into homogeneous and
discrete clusters and aimed to identify favorable conditions
for the accumulation and/or transport of aerosols over the
region. Days with daily mean AOD500 above the decadal
(2001–2010) mean plus 1STD, recorded at Kanpur AER-
ONET station, have been considered as AEs and for these
days the synoptic weather conditions over south Asia were
analyzed. A statistical methodology scheme comprising
Factor (S-mode) and Cluster (K-means) Analysis was
performed on the Z700 values over southern Asia obtained
from NCEP/NCAR re-analysis and six atmospheric circu-
lation clusters were identified. Clusters 1, 2, 3 and 6 pre-
sented their highest frequency in post-monsoon and winter
seasons, mainly associated with larger Z700 over Arabian
Peninsula extended over Arabian Sea and central India,
while Clusters 4 and 5 (higher frequency in May–July)
showed low MSLP over the arid regions of northwestern
India, Pakistan and Arabia thus, favoring convection and
uplift of dust. For each cluster the anomalies from the mean
climatological situation were used to reveal more regional
details of the atmospheric circulation. The mean composite
spatial distribution of MODIS-AOD550 for 4 of the clusters
showed increased aerosol loading only over IGP, consistent
with the synoptic conditions revealed by these clusters,
which favored the accumulation of aerosols. In the other 2
(Clusters 4 and 5) the atmospheric circulation characteris-
tics favored the long-range dust transport over IGP from
the neighboring Thar Desert and the Arabian Peninsula.
The AEs over Kanpur could be representative for the
neighboring regions over IGP and on certain circumstances
for the downwind areas, like northernmost BoB during
winter. Furthermore, for Clusters 4 and 5 the high AODs
over Kanpur were well correlated with AODs over northern
Arabian Sea, Pakistan and northwestern India suggesting
transport of dust aerosols from these areas during the AE
days. Therefore, the source regions and the strength of
advection depend on atmospheric circulation cluster that,
in turn, influences and modifies the aerosol properties over
central IGP. SPRINTARS simulations of aerosol properties
on the AE days over Kanpur revealed systematic
underestimation of AOD, especially in the winter season,
and very low anthropogenic aerosol (BC, sulfate, OC)
component, while the dust component was much higher for
AEs belonging in Clusters 4 and 5. These results, which are
in agreement with other model (GOCART, GFDL-AM2)
simulations over IGP, suggest necessity of improvement of
the emission inventories and/or aerosol life cycle in the
models in order to simulate better the aerosol impact on
monsoon system and regional climate.
Acknowledgments IIT Kanpur AERONET is operational since
January 2001; one of the authors (RPS) took lead to deploy Kanpur
AERONET after the joint agreement by IIT Kanpur and NASA. Our
sincere thanks to the Kanpur AERONET team and the current PIs
(S.N. Tripathi and B.N. Holben) for making the data available. The
NCEP/NCAR Reanalysis team is also gratefully acknowledged for
providing the meteorological maps. We also acknowledge the MO-
DIS scientists and associated NASA personnel for the production of
the data used in this research effort via Giovanni online data system.
The SPRINTARS calculations were performed by using National
Institute for Environmental Studies (NIES) supercomputer system
(NEC SX-8R/128M16). We would also like to thank many developers
for MIROC AGCM and SPRINTARS and the two anonymous
reviewers for helping us in improving the scientific quality of the
work.
References
Abish B, Mohanakumar K (2011) Biennial Variability in aerosol
optical depth associated with QBO modulated tropical tropo-
pause. Atmos Sci Lett 13:61–66
Abish B, Mohankumar K (2013) Absorbing aerosol variability over
the Indian subcontinent and its increasing dependence on ENSO.
Glob Plan Change 106:13–19
Aloysius M, Prijith SS, Mohan M, Parameswaran K (2011) Role of
dynamics in the advection of aerosols over the Arabian Sea
along the west coast of peninsular India during pre-monsoon
season: a case study based on satellite data and Regional Climate
Model. J Earth Syst Sci 120:269–279
Badarinath KVS, Kharol SK, Sharma AR, Roy PS (2009) Fog Over
Indo-Gangetic Plains—A Study Using Multisatellite Data and
Ground Observations. IEEE J Sel Topics Appl Earth Observ
Rem Sens 2:185–195
Bhawar RL, Devara PCS (2010) Study of successive contrasting
monsoons (2001–2002) in terms of aerosol variability over a
tropical station Pune, India. Atmos Chem Phys 10:29–37
Bollasina M, Nigam S (2009) Indian Ocean SST, evaporation, and
precipitation during the South Asian summer monsoon in IPCC-
AR4 coupled simulations. Clim Dyn 33:1017–1032. doi:10.
1007/s00382-008-0477-4
Bollasina MA, Ming Y, Ramaswamy V (2011) Anthropogenic
aerosols and the weakening of the South Asian summer
monsoon. Science 334:502–505
Carmona I, Alpert P (2009) Synoptic classification of moderate
resolution imaging spectroradiometer aerosols over Israel.
J Geophys Res 114:D072008. doi:10.1029/D010160
Chin M, Diehl T, Dubovik O, Eck TF, Holben BN, Sinyuk A, Streets
DG (2009) Light absorption by pollution, dust, and biomass
burning aerosols: a global model study and evaluation with
AERONET measurements. Ann Geophys 27:3439–3464
Dey S, di Girolamo L (2010) A climatology of aerosol optical and
microphysical properties over the Indian subcontinent from
D. G. Kaskaoutis et al.
123
Page 17
9 years (2000-2008) of Multiangle Imaging Spectroradiometer
(MISR) data. J Geophys Res 115:D15204. doi:10.1029/
2009JD013395
Dey S, di Girolamo L (2011) A decade of change in aerosol properties
over the Indian subcontinent. Geophys Res Lett 38:L14811.
doi:10.1029/2011GL048153
Dey S, Tripathi SN (2007) Estimation of aerosol optical properties
and radiative effects in the Ganga Basin, northern India during
the winter time. J Geophys Res 112:D03203. doi:10.1029/
2006JD007267
Dey S, Tripathi SN (2008) Aerosol direct radiative effects over
Kanpur in the Indo-Gangetic basin, northern India: long-term
(2001-2005) observations and implications to regional climate.
J Geophys Res 113:D04212. doi:10.1029/2007JD009029
di Girolamo L, Bond TC, Bramer D, Diner DJ, Fettinger F, Kahn RA,
Martonchik JA, Ramana MV, Ramanathan V, Rasch PJ (2004)
Analysis of Multi-angle Imaging SpectroRadiometer (MISR) aer-
osol optical depths over greater India during winter 2001–2004.
Geophys Res Lett 31:L23115. doi:10.1029/2004GL021273
Diehl T, Heil A, Chin M, Pan X, Streets D, Schulz M, Kinne S (2012)
Anthropogenic, biomass burning, and volcanic emissions of
black carbon, organic carbon, and SO2 from 1980 to 2010 for
hindcast model experiments. Atmos Chem Phys Discuss
12:24895–24954
Dubovik O, Smirnov A, Holben BN, King MD, Kaufman YJ, Eck TF,
Slutsker I (2000) Accuracy assessments of aerosol properties
retrieved from Aerosol Robotic Network (AERONET) sun and
sky radiance measurements. J Geophys Res 105:9791–9806
Eck TF, Holben BN, Sinyuk A, Pinker RT, Goloub P, Chen H,
Chatenet B, Li Z, Singh RP, Tripathi SN, Reid JS, Giles DM,
Dubovik O, O’Neill NT, Smirnov A, Wang P, Xia X (2010)
Climatological aspects of the optical properties of fine/coarse
mode aerosol mixtures. J Geophys Res 115:D19205. doi:10.
1029/2010JD014002
Eck TF, Holben BN, Reid JS, Giles DM, Rivas MA, Singh RP, Tripathi
SN, Bruegge CJ, Platnick S, Arnold GT, Krotkov NA, Carn SA,
Sinyuk A, Dubovik O, Arola A, Schafer JS, Artaxo P, Smirnov A,
Chen H, Goloub P (2012) Fog- and cloud-induced aerosol
modification observed by the Aerosol Robotic Network (AER-
ONET). J Geophys Res 117:D07206. doi:10.1029/2011JD016839
El-Askary H, Gautam R, Singh RP, Kafatos M (2006) Dust storms
detection over the Indo-Gangetic basin using multi sensor data.
Adv Space Res 37:728–733
Gadgil S, Vinayachandran PN, Francis PA (2003) Droughts of the
Indian summer monsoon: role of clouds over the Indian Ocean.
Curr Sci 85:1713–1719
Ganguly D, Ginoux P, Ramaswamy V, Winker DM, Holben BN,
Tripathi SN (2009) Retrieving the composition and concentra-
tion of aerosols over the Indo-Gangetic basin using CALIOP and
AERONET data. Geophys Res Lett 36:L13806. doi:10.1029/
2009GL038315
Ganguly D, Rasch PJ, Wang H, Yoon J-H (2012) Climate response of
the South Asian monsoon system to anthropogenic aerosols.
J Geophys Res 117:D13209. doi:10.1029/2012JD017508
Gautam R, Hsu NC, Kafatos M, Tsay S-C (2007) Influences of winter
haze on fog/low cloud over the Indo-Gangetic plains. J Geophys
Res 112:D05207. doi:10.1029/2005JD007036
Gautam R, Hsu NC, Lau K-M, Kafatos M (2009a) Aerosol and
rainfall variability over the Indian monsoon region: distributions,
trends and coupling. Ann Geophys 29:3691–3703
Gautam R, Hsu NC, Lau K-M, Kafatos M (2009b) Enhanced pre-
monsoon warming over the Himalayan-Gangetic region from
1979 to 2007. Geophys Res Lett 36:L07704. doi:10.1029/
2009GL037641
Gautam R, Hsu NC, Lau K-M (2010) Premonsoon aerosol charac-
terization and radiative effects over the Indo-Gangetic Plains:
implications for regional climate warming. J Geophys Res
115:D17208. doi:10.1029/2010JD013819
Gautam R, Hsu NC, Tsay SC, Lau K-M, Holben BN, Bell S, Smirnov
A, Li C, Hansell R, Ji Q, Payra S, Aryal D, Kayastha R, Kim KM
(2011) Accumulation of aerosols over the Indo-Gangetic plains
and southern slopes of the Himalayas: distribution, properties
and radiative effects during the 2009 pre-monsoon Season.
Atmos Chem Phys 11:12841–12863
Ghude SD, Kulkarni SH, Kulkarni PS, Kanawade VP, Fadnavis S,
Pokhrel S, Jena C, Beig G, Bortoli D (2011) Anomalous low
tropospheric column ozone over Eastern India during the severe
drought event of monsoon 2002: a case study. Environ Sci Pollut
Res 18:1442–1455
Giles DM, Holben BN, Tripathi SN, Eck T, Newcomb W, Slutsker I,
Dickerson R, Thompson A, Mattoo S, Wang S, Singh R, Sinyuk
A, Schafer J (2011) Aerosol Properties over the Indo-Gangetic
Plain: a 1 Mesoscale Perspective from the TIGERZ Experiment.
J Geophys Res 116:D18203. doi:10.1029/2011JD015809
Gillette D (1978) A wind tunnel simulation of the erosion of soil:
effect of soil texture, sandblasting, wind speed and soil
consolidation on dust production. Atmos Environ 12:1735–1743
Gkikas A, Hatzianastassiou N, Mihalopoulos N (2009) Aerosol events
in the broader 11 Mediterranean basin based on 7-year
(2000–2007) MODIS C005 data Ann Geophys 27:3509–3522
Gkikas A, Houssos EE, Hatzianastassiou N, Bartzokas A (2012)
Synoptic conditions favouring the occurrence of aerosol episodes
over the broader Mediterranean basin. Q J Royal Meteorol Soc
138:932–949
Goto D, Nakajima T, Takemura T, Sudo K (2011a) A study of
uncertainties in the sulfate distribution and its radiative forcing
associated with sulfur chemistry in a global aerosol model.
Atmos Chem Phys 11:10889–10910
Goto D, Takemura T, Nakajima T, Badarinath KVS (2011b) Global
aerosol model-derived black carbon concentration and single
scattering albedo over Indian region and its comparison with
ground observations. Atmos Environ 45:3277–3285
Goto D, Badarinath KVS, Takemura T, Nakajima T (2011c)
Simulation of aerosol optical properties over tropical urban sitein India using a global model and its comparison with ground
measurements. Ann Geophys 29:955–963
Guleria RP, Kuniyal JC, Rawat PS, Sharma NL, Thakur HK, Dhyani PP,
Singh M (2011) The assessment of aerosol optical properties over
Mohal in the northwestern Indian Himalayas using satellite and
ground-based measurements and an influence of aerosol transport
on aerosol radiative forcing. Meteor Atmos Phys 113:153–169
Henriksson SV, Laaksonen A, Kerminen V-M, Raisanen P, Jarvinen
H, Sundstrom A-M, de Leeuw G (2011) Spatial distributions and
seasonal cycles of aerosols in India and China seen in global
climate-aerosol model. Atmos Chem Phys 11:7975–7990
Houssos EE, Bartzokas A (2006) Extreme precipitation events in NW
Greece. Adv Geosci 7:91–96
Houssos EE, Lolis CJ, Gkikas A, Hatzianastassiou N, Bartzokas A
(2011) On the atmospheric circulation characteristics associated
with fog in Ioannina, north-western Greece. Int J Climatol
32:1847–1862
Huneeus N, Schulz M, Balkanski Y, Griesfeller J et al (2011) Global
dust model intercomparison in AeroCom phase I. Atmos Chem
Phys 11:7781–7816
Jethva H, Satheesh SK, Srinivasan J (2007) Evaluation of Moderate-
Resolution Imaging Spectroradiometer (MODIS) Collection 004
(C004) aerosol retrievals at Kanpur, Indo-Gangetic Basin.
J Geophys Res 112:D14216. doi:10.1029/2006JD007929
Jolliffe IT (1986) Principal Component Analysis. Springer, New York
Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L,
Iredell M, Saha S, White G, Woollen J, Zhu Y, Leetmaa A,
Reynolds R, Chelliah M, Ebisuzaki W, Higgins W, Janowiak J,
Synoptic weather conditions
123
Page 18
Mo KC, Ropelewski C, Wang J, Roy J, Dennis J (1996) The
NCEP/NCAR 40-year reanalysis project. Bull Am Meteor Soc
77:437–470
Kar J, Deeter MN, Fishman J, Liu Z, Omar A, Creilson JK, Trepte
CR, Vaughan MA, Winker DM (2010) Wintertime pollution
over the Eastern Indo-Gangetic Plains as observed from
MOPITT, CALIPSO and tropospheric ozone residual data.
Atmos Chem Phys 10:12273–12283
Kaskaoutis DG, Kumar Kharol S, Sinha PR, Singh RP, Badarinath
KVS, Mehdi W, Sharma M (2011) Contrasting aerosol trends
over South Asia during the last decade based on MODIS
observations. Atmos Meas Tech Discuss 4:5275–5323
Kaskaoutis DG, Nastos PT, Kosmopoulos PG, Kambezidis HD
(2012a) Characterizing the long-range transport mechanisms of
different aerosol types over Athens, Greece during 2000–2005.
Int J Climatol 32:1249–1270
Kaskaoutis DG, Singh RP, Gautam R, Sharma M, Kosmopoulos PG,
Tripathi SN (2012b) Variability and trends of aerosol properties
over Kanpur, northern India using AERONET data (2001–10).
Environ Res Lett 7:024003
Kaskaoutis DG, Gautam R, Singh RP, Houssos EE, Goto D, Singh S,
Bartzokas A, Kosmopoulos PG, Sharma M, Hsu NC, Holben
BN, Takemura T (2012c) Influence of anomalous dry conditions
on aerosols over India: transport, distribution and properties.
J Geophys Res 117:D09106. doi:10.1029/2011JD017314
Kaufman YJ, Tanre D (1998) Algorithm for remote sensing of
tropospheric aerosol from MODIS, algorithm theoretical basis
documents (ATBD-MOD-02), 85
Kharol SK, Badarinath KVS, Kaskaoutis DG, Sharma AR, Gharai B
(2011) Influence of continental advection on aerosol character-
istics over Bay of Bengal (BoB) in winter: results from
W-ICARB cruise experiment. Ann Geophys 29:1423–1438
Kharol SK, Badarinath KVS, Sharma AR, Mahalakshmi DV, Singh
D, Krishna Prasad V (2012) Black carbon aerosol variations over
Patiala city, Punjab, India - A study during agriculture crop
residue burning period using ground measurements and satellite
data. J Atmos Solar-Terr Phys 84–85:45–51
Kinne S, Schulz M, Textor C, Guibert S et al (2006) An AeroCom
initial assessment—optical properties in aerosol component
modules of global models. Atmos Chem Phys 6:1815–1834
Kishcha P, Starobinets B, Kalashnikova O, Alpert P (2011) Aerosol
optical thickness trends and population growth in the Indian
subcontinent. Int J Rem Sens 32:9137–9149. doi:10.1080/
01431161.2010.550333
Kishcha P, Starobinets B, Long CN, Alpert P (2012) Unexpected
increasing AOT trends over north-west Bay of Bengal in the
early post-monsoon season. J Geophys Res. doi:10.1029/
2012JD018726
Komppula M, Mielonen T, Arola A, Korhonen K, Lihavainen H,
Hyvarinen A-P, Baars H, Engelmann R, Althausen D, Ansmann
A, Muller D, Panwar TS, Hooda RK, Sharma VP, Kerminen
V-M, Lehtinen KEJ, Viisanen Y (2012) One year of Raman-lidar
measurements in Gual Pahari EUCAARI site close to New Delhi
in India: seasonal characteristics of the aerosol vertical structure.
Atmos Chem Phys 12:4513–4524
Krishnan R, Sabin TP, Ayantika DC, Kitoh A, Sugi M, Murakami H,
Turner AG, Slingo JM, Rajendran K (2012) Will the South Asian
monsoon overturning circulation stabilize any further? Clim
Dyn. doi:10.1007/s00382-012-1317-0
Kuhlmann J, Quaas J (2010) How can aerosols affect the Asian
summer monsoon? Assessment during three consecutive pre-
monsoon seasons from CALIPSO satellite data. Atmos Chem
Phys 10:4673–4688
K-1 Model Developers (2004) K-1 coupled GCM (MIROC) descrip-
tion, K-1 Technical Report 1. Hasumi H, Emori S, (eds),
University of Tokyo, Tokyo, 34 pp
Lau KM, Kim MK, Kim KM (2006) Asian summer monsoon
anomalies induced by aerosol direct forcing: the role of the
Tibetan Plateau. Clim Dyn 26:855–864
Lau KM, Kim MK, Kim KM, Lee WS (2010) Enhanced surface
warming and accelerated snow melt in the Himalayas and
Tibetan Plateau induced by absorbing aerosols. Environ Res Lett
5. doi:10.1088/1748-9326/5/2/025204
Lawrence MG, Lelieveld J (2010) Atmospheric pollutant outflow
from southern Asia: a review. Atmos Chem Phys
10:11017–11096
Levy RC, Remer LA, Mattoo S, Vermote E, Kaufman YJ (2007)
Second-generation operational algorithm: retrieval of aerosol
properties over land from inversion of Moderate Resolution
Imaging Spectroradiometer spectral reflectance. J Geophys Res
112:D13211. doi:10.1029/2006JD007811
Levy RC, Remer LA, Kleidman RG, Mattoo S, Ichoku C, Kahn R,
Eck TF (2010) Global evaluation of the Collection 5 MODIS
dark-target aerosol products over land Atmos. Chem Phys
10:10399–10420. doi:10.5194/acp-10-10399-2010
Lodhi NK, Beegum SN, Singh S, Kumar K (2013) Aerosol
climatology at Delhi in the western Indo-Gangetic Plain:
Microphysics, long-term trends, and source strengths. J Geophys
Res 118: doi:10.1002/jgrd.50165
Lu Z, Zhang Q, Streets DG (2011) Sulfur dioxide and primary
carbonaceous aerosol emissions in China and India, 1996–2010.
Atmos Chem Phys 11:9839–9864
Manly BFJ (1986) Multivariate Statistical Methods: A primer.
Chapman & Hall, London
Manoj MG, Devara PCS, Safai PD, Goswami BN (2011) Absorbing
aerosols facilitate transition of Indian monsoon breaks to active
spells. Clim Dyn 37:2181–2198
Mishra AK, Shibata T (2012) Synergistic analyses of optical and
microphysical properties of agricultural crop residue burning
aerosols over the Indo-Gangetic Basin (IGB). Atmos Environ
57:205–218
Misra AS, Tripathi SN, Kaul D, Welton E (2012) Study of MPLNET
derived aerosol climatology over Kanpur, India, and validation
of CALIPSO Level 2 Version 3 Backscatter and Extinction
products. J Atmos Ocean Technol 29:1285–1294. doi:10.1175/
JTECH-D-11-00162.1
Monahan EC, Spiel DE, Davidson KL (1986) A model of marine
aerosol generation via whitecaps and wave disruption. In:
Monahan E, Niocaill GM, Reidel D (eds) Oceanic whitecaps.
Norwell, Mass., USA, pp 167–174
Nakajima T, Tsukamoto M, Tsushima Y, Numaguti A, Kimura T
(2000) Modeling of the radiative process in an atmospheric
general circulation model. Appl Opt 39:4869–4878
Nastos PT (2012) Meteorological patterns associated with intense
Saharan dust outbreaks over Greece in winter. Adv Meteor ID
828301. doi:10.1155/2012/828301
Pathak B, Bhuyan PK, Biswas J, Takemura T (2013) Long term
climatology of particulate matter and associated microphysical
and optical properties over Dibrugarh, North-East India and
inter-comparison with SPRINTARS simulations. Atmos Environ
69:334–344
Prasad AK, Singh RP (2007a) Changes in aerosol parameters during
major dust storm events (2001–2005) over the Indo-Gangetic
Plains using AERONET and MODIS data. J Geophys Res
112:D09208. doi:10.1029/2006JD007778
Prasad AK, Singh RP (2007b) Comparison of MISR-MODIS aerosol
optical depth over the Indo-Gangetic basin during the winter and
summer seasons (2000–2005). Rem Sens Environ 107:109–119
Ramanathan V, Chung C, Kim D, Bettge T, Buja L, Kiehl JT,
Washington WM, Fu Q, Sikka DR, Wild M (2005) Atmospheric
brown clouds: impacts on South Asian climate and hydrological
cycle. PNAS 102:5326–5333. doi:10.1073/pnas.0500656102
D. G. Kaskaoutis et al.
123
Page 19
Reddy MS, Venkataraman C (2002) Inventory of aerosol and sulphur
dioxide emission from India: II-biomass combustion. Atmos
Environ 36:699–712
Remer LA, Kleidman RG, Levy RC, Kaufman YJ, Tanre D, Mattoo
S, Martins JV, Ichoku C, Koren I, Yu H, Holben BN (2008)
Global aerosol climatology from the MODIS satellite sensors.
J Geophys Res 113: D14S07. doi:10.1029/2007JD009661
Rotstayn LD, Lohmann U (2002) Tropical rainfall trends and the
indirect aerosol effect. J Clim 15:2103–2116
Sekiguchi M, Nakajima T (2008) A k-distribution-based radiation
code and its computational optimization for an atmospheric
general circulation model. J Q Spectrosc Radiat Transf
109:2779–2793
Sharma S (1996) Applied Multivariate Techniques. John Wiley,
New York
Sharma AR, Kharol SK, Badarinath KVS, Singh D (2010) Impact of
agriculture crop residue burning on atmospheric aerosol load-
ing—a study over Punjab State, India. Ann Geophys 28:367–379
Sharma D, Singh M, Singh D (2012) Impact of post-harvest biomass
burning on aerosol characteristics and radiative forcing over
Patiala, North–West region of India. J Instit Eng 8:11–24
Shi Y, Zhang J, Reid JS, Hyer EJ, Eck TF, Holben BN, Kahn RA
(2011) A critical examination of spatial biases between MODIS
and MISR aerosol products—application for potential AERON-
ET deployment. Atmos Meas Tech 4:2823–2836
Singh RP, Dey S, Tripathi SN, Tare V, Holben BN (2004) Variability
of aerosol parameters over Kanpur, northern India. J Geophys
Res 109:D23206. doi:10.1029/2004JD004966
Smirnov A, Holben BN, Eck TF, Dubovik O, Slutsker I (2000) Cloud
screening and quality control algorithms for the AERONET data
base. Rem Sens Environ 73:337–349
Srivastava R, Ramachandran S (2012) The mixing state of aerosols
over the Indo-Gangetic Plain and its impact on radiative forcing.
Q J Royal Meteorol Soc. doi:10.1002/qj.1958
Srivastava AK, Tiwari S, Devara PCS, Bisht DS, Srivastava MK,
Tripathi SN, Goloub P, Holben BN (2011) Pre-monsoon aerosol
characteristics over the Indo-Gangetic Basin: implications to
climatic impact. Ann Geophys 29:789–804
Srivastava AK, Singh S, Tiwari S, Bisht DS (2012a) Contribution of
anthropogenic aerosols in direct radiative forcing and atmo-
spheric heating rate over Delhi in the Indo-Gangetic Basin.
Environ Sci Pollut Res 19:1144–1158
Srivastava AK, Tripathi SN, Dey S, Kanawade VP, Tiwari S (2012b)
Inferring aerosol types over the Indo-Gangetic Basin from ground
based sunphotometer measurements. Atmos Res 109:64–75
Srivastava AK, Singh S, Tiwari S, Kanawade VP, Bisht DS (2012c)
Variation between near-surface and columnar aerosol character-
istics during the winter and summer at Delhi in the Indo-
Gangetic Basin. J Atmos Sol-Terr Phys 77:57–66
Streets DG, Bond TC, Carmichael GR, Fernandes SD, Fu Q, He D,
Klimont Z, Nelson SM, Tsai NY, Wang MQ, Woo J-H, Yarber
KF (2003) An inventory of gaseous and primary aerosol
emissions in Asia in the year 2000. J Geophys Res
108(D21):8809. doi:10.1029/2002JD003093
Sugar CA, James GM (2003) Finding the number of clusters in a
dataset: an information-theoretic approach. J Am Stat Assoc
98:750–763
Takemura T, Okamoto H, Maruyama Y, Numaguti A, Higurashi A,
Nakajima T (2000) Global three-dimensional simulation of
aerosol optical thickness distribution of various origins. J Geo-
phys Res 105:17853–17873
Takemura T, Nozawa T, Emori S, Nakajima TY, Nakajima T (2005)
Simulation of climate response to aerosol direct and indirect
effects with aerosol transport radiation model. J Geophys Res
110:D02202. doi:10.1029/2004JD005029
Takemura T, Egashira M, Matsuzawa K, Ichijo H, O’ishi R, Abe-
Ouchi A (2009) A simulation of the global distribution and
radiative forcing of soil dust aerosols at the Last Glacial
Maximum. Atmos Chem Phys 9:3061–3073
Tripathi SN, Dey S, Chandel A, Srivastva S, Singh RP, Holben B
(2005) Comparison of MODIS and AERONET derived aerosol
optical depth over the Ganga basin, India. Ann Geophys
23:1093–1101
Verma S, Venkataraman C, Boucher O (2012) Attribution of aerosol
radiative forcing over India during the winter monsoon to
emissions from source categories and geographical regions.
Atmos Environ 45:4398–4407
Vijayakumar K, Devara PCS, Simha CP (2012) Aerosol Features
during Drought and Normal Monsoon Years: a Study Under-
taken with Multi-Platform Measurements over A Tropical Urban
Site. Aeros Air Qual Res 12:1444–1458
Wang C, Kim D, Ekman AML, Barth MC, Rasch PJ (2009) Impact of
anthropogenic aerosols on Indian summer monsoon. Geophys
Res Lett 36:L21704. doi:10.1029/2009GL040114
Watanabe M, Suzuki T, O’ishi R, Komuro Y, Watanabe S, Emori S,
Takemura T, Chikira M, Ogura T, Sekiguchi M, Takata K,
Yamadaki D, Tokohata T, Nozawa T, Hasumi H, Tatebe H, Kimoto
M (2010) Improved climate simulation by MIROC 5: mean states,
variability, and climate sensitivity. J Clim 23:6312–6335
Synoptic weather conditions
123