Environment International 83 (2015) 146157

Contents lists available at ScienceDirect

Environment International

journal homepage: www.elsevier.com/locate/envint

Full length article

Impact of air pollutants from surface transport sources on human health:

A modeling and epidemiological approach
Preeti Aggarwal a, Suresh Jain a,b,
a
Department of Natural Resources, TERI University, 10, Institutional Area, Vasant Kunj, New Delhi 110070, India
b
Department of Energy and Environment, TERI University, 10, Institutional Area, Vasant Kunj, New Delhi 110070, India

a r t i c l e i n f o a b s t r a c t

Article history: This study adopted an integrated source-to-receptor assessment paradigm in order to determine the effects of
Received 16 February 2015 emissions from passenger transport on urban air quality and human health in the megacity, Delhi. The emission
Received in revised form 4 June 2015 modeling was carried out for the base year 2007 and three alternate (ALT) policy scenarios along with a business
Accepted 17 June 2015
as usual (BAU) scenario for the year 2021. An Activity-Structure-Emission Factor (ASF) framework was adapted
Available online 1 July 2015
for emission modeling, followed by a grid-wise air quality assessment using AERMOD and a health impact assess-
Keywords:
ment using an epidemiological approach. It was observed that a 2021-ALT-III scenario resulted in a maximum
Vehicle pollutants concentration reduction of ~24%, ~42% and ~58% for carbon monoxide (CO), nitrogen dioxide (NO2) and partic-
Air quality modeling ulate matter (PM), respectively, compared to a 2021-BAU scenario. Further, it results in signicant reductions in
Respiratory and cardiovascular mortality respiratory and cardiovascular mortality, morbidity and Disability Adjusted Life Years (DALY) by 41% and 58% on
Morbidity exposure to PM2.5 and NO2 concentrations when compared to the 2021-BAU scenario, respectively. In other
DALY words, a mix of proposed policy interventions namely the full-phased introduction of the Integrated Mass Transit
Scenario analysis System, xed bus speed, stringent vehicle emission norms and a hike in parking fees for private vehicles would
help in strengthening the capability of passenger transport to cater to a growing transport demand with a min-
imum health burden in the Delhi region. Further, the study estimated that the transport of goods would be
responsible for ~5.5% additional VKT in the 2021-BAU scenario; however, it will contribute ~49% and ~55% addi-
tional NO2 and PM2.5 concentrations, respectively, in the Delhi region. Implementation of diesel particulate lters
for goods vehicles in the 2021-ALT-IV-O scenario would help in the reduction of ~87% of PM2.5 concentration,
compared to the 2021-BAU scenario; translating into a gain of 1267 and 505 DALY per million people from expo-
sure to PM2.5 and NO2 concentrations, respectively. These ndings suggest that signicant health benets are
possible if goods transport is also included while designing strategies and polices in order to improve the overall
urban air quality and minimize health impacts in city areas.
2015 Elsevier Ltd. All rights reserved.

1. Introduction 2012). However, the situation is anticipated to deteriorate in densely

Air pollution is recognized as an important health problem affecting
millions of lives around the globe. It also adversely affects local ecolog-
ical systems, regional atmospheric chemistry as well as global climate
(Vlachokostas et al., 2010; Watson and Chow, 2007; Dockery and
Pope, 2006). The World Health Organization (WHO) identied ambient
air pollution as one of the 10 major risk factors contributing to the global
health burden (Lim et al., 2012); causing ~16% of premature deaths in
2009 (WHO, 2009) and ~3.7 million deaths in 2012. The Organisation
for Economic Cooperation and Development (OECD) in its report on
Environmental Outlook to 2050 estimated that ambient air pollution
would become the top cause of environmental mortality; resulting in
an additional ~3.6 million premature deaths by the year 2050 (OECD,
Corresponding author at: Department of Natural Resources, TERI University, 10,
Institutional Area, Vasant Kunj, New Delhi 110070, India.
E-mail addresses: sureshjain_in@yahoo.com, sureshj@teri.res.in (S. Jain).
populated urban areas and megacities of developing regions, especially
India and China (Lim et al., 2012).
The major source of air pollution in urban areas is road transport, in
addition to thermal power plants, industries and diffused biomass burn-
ing (Kumar et al., 2013; CPCB, 2010; Jain and Khare, 2010, 2008; WHO,
2005). The Auto Fuel Policy of India (MPNG, 2003) revealed that the
condition was worse in the metropolitan cities of India as compared to
many developed regions of the world. For instance, Delhi is considered
as the most motorized city of India with ~299 two-wheelers and ~162
cars for every 1000 people (MoRTH, 2011; Sahai and Bishop, 2009);
witnessing serious trafc congestion and delays (Lebel et al., 2007).
It is one of the most polluted megacities around the globe, where the
average concentration of g/m3 respirable suspended particulate
matter (RSPM) is noted to increase from 161 g/m3 in the year 2007 to
281 g/m3 in the year 2011 (Kumar et al., 2013). This exceeds the global
and national ambient air quality standards; therefore, it is responsible
for a notable amount of excess deaths and hospital admissions (CPCB,

http://dx.doi.org/10.1016/j.envint.2015.06.010
0160-4120/ 2015 Elsevier Ltd. All rights reserved.
 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157
2012; Kumar et al., 2011). Further, Sindhwani and Goyal (2014) have es-
timated the air pollution concentration from various sectors in the Delhi
region. They have reported that the transport sector contributes around
8 g/m3 PM10 concentrations for the year 2007 in the ambient environ-
ment. Several initiatives have been taken to address the issues of trafc
and air pollution in the Delhi region; for e.g., introduction of unleaded pet-
rol in 1995, and introduction of vehicle emission standards i.e., Bharat
Standard (BS) III in 2005 and BS IV in the year 2010 (Jain et al., 2014).
However, these initiatives were not able to achieve the acceptable level
of reduction in ambient air pollution (Yagi and Nagayama, 2010), as
reected in deteriorating human health due to poor air quality conditions
(Gurjar et al., 2010). These ndings strongly suggest an urgent need to in-
tegrate the assessment of environmental and social benets of proposed
transport policy interventions, prior to their implementation as a part of
development strategy (World Bank, 2013; Aunan, 1995).
1.1. Measures to mitigate vehicle emissions
Estimation of emissions from vehicular sources is the rst step to
assess the impact of vehicular use on ambient air quality and human
health. Many past studies have quantied vehicle emissions in the
Delhi region; for instance, Gurjar et al. (2004) prepared an emission in-
ventory for the years 19902000 and concluded that in the year 2001,
the transport sector contributed to 82% of nitrogen dioxide (NO2),
17% of total suspended particulate matter (TSP) and 86% of carbon
monoxide (CO) emissions, in comparison to other sources. This is attrib-
utable to old vehicle technology, diesel-based busses and secondary
emission factors collated from different sources. Recently, two studies
reported the contribution of the transport sector in overall pollutant
concentration using satellite-based emission estimation methodology
in the Delhi region. Guttikunda and Calori (2013) estimated that the
transport sector contributes to ~17% of particulate matter less than or
equal to 2.5 m (PM2.5), ~ 13% of PM less than or equal to 10 m
(PM10), ~53% of nitrogen oxides (NOx) and 18% of CO in the study re-
gion; while Sahu et al. (2011) estimated that it contributes ~ 29% and
~ 45% of PM10 and PM2.5 emissions, respectively. It was observed that
there is a tremendous change in technology and fuel-mix since 2001,
making it difcult to compare the results of Gurjar et al. (2004) with
the latter two studies. On the other hand, Guttikunda and Calori
(2013) and Sahu et al. (2011) reported the percentage (%) contribution
of transport for different boundary locations i.e., the National Capital
Territory of Delhi (NCTD) along with satellite towns around NCTD in a
former study and the Delhi National Capital Region (Delhi NCR) in a
later study, which resulted in ~ 0.15 million tons and 0.3 million tons
of PM2.5 emission, respectively. Further, these studies have used regis-
tered vehicle data for computing vehicle emissions, which may result
in uncertainties during emission modeling. Further, none of these stud-
ies have projected future transport demand (in vehicle kilometers trav-
eled, VKT or passenger kilometers, PKM).
On the other hand, it was observed that the majority of air quality
projection studies employed a deterministic modeling approach in the
Indian region. Sharma and Khare (2001) have reported that the deter-
ministic models are best suited to predict air pollution levels in the
Delhi region owing to its capability of handling temporal and spatial
variations appropriately. Further, Gokhale and Khare (2004) observed
that the causal nature as well as its ability to integrate emission in-
ventory and meteorological parameters during air quality modeling,
makes the Gaussian plume model a signicant tool for trafc emission
analyses. Similarly, Goyal and Kumar (2011) observed that the Gaussian
plume models such as CALINE-4, DFLSM and ISCST-3, have found max-
imum application for quantifying the dispersion of air pollutants from
point, area and volume sources in the Delhi region. Recently, Mohan
et al. (2011) estimated the air quality impact of residential, industrial
and transport use in Delhi using an ISC-AERMOD model and observed
that it gives a more satisfactory result when compared to the ISCST-3
model. However, no notable studies have used AERMOD to estimate
147
the impact of on-road vehicle sources on urban air quality in the Delhi
region.
In case of health impact assessment, there are notable studies which
have established an association between ambient air quality and human
health in various locales (HEI, 2004). Dockery (2001) observed that
epidemiologic studies provided the strongest evidence for establishing
a cause and effect relationship. These studies support the causal rela-
tionship of PM and other air pollutants with increased asthma cases
(Auerbach and Hernandez, 2012); and exacerbate chronic obstructive
pulmonary disease (COPD) (Eisner et al., 2010) and cardiovascular
damage (Brook et al., 2010). In the Indian context, Simon et al. (1997)
for the rst time established a causal relationship of ambient air pollu-
tion and an increase in mortality in Delhi using time-series data for
the year 19911994. Further, Pande et al. (2002) have correlated
cardio-respiratory hospital admissions with ambient air pollutants,
showing an increase of ~ 25% emergency room visits as the ambient
air pollution level increases. A similar analysis of daily hospital admis-
sions was conducted by Nidhi and Jayaraman (2007) where they re-
ported very high risk (relative risk, RR) on exposure to ambient air
pollution (1.072.82 for a 10 g/m3 increase in pollutant concentration).
However, it can be observed that inclusion of admitted patients from
Delhi and other states might have introduced structural bias during
the computation of these results.
Recently, Lahiri and Ray (2012) collected health prevalence data for
the Delhi region and correlated it with ambient air pollution to establish
an impact on morbidity, lung functions, cytology and hematology. How-
ever, the development of RR functions from Delhi and West Bengal
(outside Delhi) based prevalence rates would introduce undesirable
error and uncertainty, owing to differences in weather and air pollution
conditions prevalent in both the regions. On the other hand, very few
Delhi based studies, for instance Guttikunda (2008), Gurjar et al.
(2010) and Guttikunda and Goel (2013), have adopted global concen-
tration response functions (RR) developed by the World Health Organi-
zation (WHO, 2006) in order to quantify health effects of growing air
pollution. It is very well accepted that the use of global RR function
can aid in preliminary research and decision making, when local re-
sponse functions for a particular region are unavailable. However,
these estimates may not be accurate as global RR functions are de-
signed to capture the impact of exposure to global average air pollutant
concentrations rather than the impact of proximity and high exposure
to roadside emissions, as in case of Delhi region (Lu et al., 2015; HEI,
2010; Suresh et al., 2000; Lipfert et al., 2006a, b). Hence, it is inevitable
for the megacities with higher trafc density to study the health effects
of vehicle pollution by developing local RR functions after considering
ambient environment and health conditions (Zou et al., 2009). There-
fore, the current study estimates the impact of identied policy inter-
ventions targeting road transport on urban air quality and human
health, with the case of Delhi region.
To summarize, the current study estimates source-to-receptor
impact of transport sources in the Delhi region on ambient air quality
and public health, by adopting scenario analysis. It intends to inventorize
and characterize on-road vehicle activity in the NCTD region in order to
determine energy-emissions from passenger transport for current and fu-
ture years. Further, impact on ambient air quality would be estimated by
using dispersion modeling, extending it up to the local level. Subsequent-
ly, the human health impacts of these air pollutants would be captured by
adopting health survey and statistical regression analyses, for the pro-
posed alternate scenarios.
2. Material and methods
The primary objective of the study is to estimate the impact of vehic-
ular emissions on urban air quality and human health in the Delhi re-
gion. It can be noted that the prime focus is to study the impacts of
on-road passenger vehicle-based emissions using an integrated impact
assessment framework, as presented in Fig. 1. The study has considered
148 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157

2007 as the base year and 2021 for future projections. Four scenarios
were generated for the year 2021 with an aim to improve air quality
and reduce health impacts of passenger transport use as explained in
Table 1A. Additionally, the study has also generated ve alternate sce-
narios (as explained in Table 1B) for estimating the effect of overall
transport, as the authors observed that majority of policy interventions
were directed to control the effect of passenger vehicles only, ignoring
the impacts of goods transport. The subsequent section describes the
modeling and epidemiological approaches adopted in order to deter-
mine the impact of various policy measures in the current study region.
2.1. Site description
The National Capital Territory of Delhi (NCTD) region is one of the
largest megacities in India and accounts for around 1% of its total popu-
lation (Jain et al., 2014). It is spread over an area of 1483 km2 and is lo-
cated in the northern part of India (28.61N 77.23E) sharing its borders
with the states of Haryana in the north, west and south, and Uttar
Pradesh to its east (DoIT, 2013). It comprises of nine administrative dis-
tricts and three statutory towns, with a population density of 11,297
persons per km2 and has an ~ 31,183 km road length, which is the

Fig. 1. Framework for integrated impact assessment of surface transport sources.

 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157
Table 1A
Scenario generation for health impact assessment from surface transport in the NCTD
region.
Scenario Description and rationale
Passenger transporta
2007-BASE IMRTS Lines 1 to 3 of the Delhi Metro (part of IMRTS),
covering a distance of 65 km are taken as
operational.
Source emission BS II and BS III were introduced in the year 2003
norms and 2005, respectively, and have been duly
considered in computations.
Passenger travel Passenger travel demand distributed across
demand various travel modes such as car, motorcycle,
scooter, auto, bus and metro have been applied
based on current modal shares collected
through the eld surveys.
2021-BAU IMRTS & source No policy interventions except BS IV was
emission norms introduced in the year 2010 and BS III norms for
others
Passenger travel Existing growth rate of vehicles considered. VKT
demand is distributed across different travel modes,
using the assumptions that the distribution of
15-, 10- and 5-year old vehicles in the years
2021 will be 20, 35 and 45%, respectively. The
assumption is based on the VKT vehicle
classication data collected through the
primary survey for the 2007-REF year. The
manufacturing of two-wheelers with
two-stroke engines has been banned in India
since 2000. Hence all such types of two--
wheelers have been assumed to be phased out
by 20152016 after their operational life of 15
years.
2021-ALT-I Policy Under this scenario, the 2021-BAU was
intervention augmented by one policy intervention the
introduction of BS-V vehicle emission norms in
the year 2017.
Rationale This policy intervention has been introduced in
order to nd effects of stringent emission norms
on resulting emissions.
2021-ALT-II Policy Full phase of IMRTS implementation, which
Intervention includes 4 phases of the Delhi Metro and Bus
Rapid Transit (BRT) system, has been
considered by the year 2021 (see S.I. Table S3),
in addition to the assumptions taken in
2021-ALT-I.
Rationale This policy intervention has been introduced in
order to nd the effects of increased share of
public transport for the reduction in the use of
private on-road vehicles and resulting
emissions.
2021-ALT-III Policy Two more interventions have been added to the
Intervention 2021-ALT-II scenario, in addition to the
assumptions taken in 2021-BAU (i) bus
speed has been regulated to 25 km h1 because
of new infrastructural development for
dedicated bus corridors, and (ii) parking fees
have been hiked (from Rs 1020 to Rs 60200)
for private vehicles.
Rationale These assumptions have been made to assess
the inuence of increased share of public
transport use on VKT and air emissions
a lter (DPF) was added in HDV vehicles
The rationale for the generation of these scenarios have been explained in Aggarwal
and Jain (2014).
largest road network in India. Nearly, 7.69 million registered vehicles
exist in the NCTD region (in the year 2011), resulting in high road trafc
and air pollution, amounting to deterioration of outdoor air quality.
2.2. Emission modeling
For this study, the ASF model (Activity-Structure-Emission Factor)
was used to estimate vehicular emissions, which is adapted from the
work carried out by Khanna et al. (2011). The ASF model is a well-
structured framework (Eq. 1) which quanties air emission released
149
Table 1B
Scenario generation for health impact assessment from surface transport in the NCTD
region.
Scenario Description and rationale
Overall transport
2007-BASE-O IMRTS Lines 1 to 3 of the Delhi Metro (part of IMRTS),
covering a distance of 65 km are taken as
operational.
Source BS II and BS III were introduced in the year 2003
emission and 2005, respectively, and have been duly
norms considered in computations.
Overall The travel demand is distributed across various
travel travel modes i.e., passenger (car, motorcycle,
demand scooter, auto, bus and metro) and goods (LDV and
HDV) have been applied based on current modal
shares collected through the eld surveys.
2021-BAU-O IMRTS No policy interventions except BS IV was
introduced in the year 2010 and BS III norms for
others
Source The diesel based LDV was assumed to be
emission converted to CNG LDVs per the governmental
norms regulation in 20052007.
Overall Existing growth rate of vehicles considered. VKT
travel is distributed across different travel modes, using
demand the assumptions that the distribution of 15-, 10-
and 5-year old vehicles in the years 2021 will be
20, 35 and 45%, respectively. The assumption is
based on the VKT vehicle classication data
collected through the primary survey for the
2007-REF year. The manufacturing of
two-wheelers with two-stroke engines has been
banned in India since 2000. Hence all such types
of two-wheelers have been assumed to be phased
out by 20152016 after their operational life of 15
years.
2021-ALT-I-O Policy Under this scenario, the 2021-BAU-O was
intervention augmented by one policy intervention the
introduction of BS-V vehicle emission norms in
the year 2017.
Rationale This policy intervention has been introduced in
order to nd the effects of stringent emission
norms on resulting emissions.
2021-ALT-II-O Policy Full phase of IMRTS implementation, which
intervention includes 4 phases of the Delhi Metro and Bus
Rapid Transit (BRT) system, has been considered
by the year 2021 (see S.I. Table S3), in addition to
the assumptions taken in 2021-ALT-I-O.
Rationale This policy intervention has been introduced in
order to nd effects of increased share of public
transport for the reduction in use of private
on-road vehicles and resulting emissions.
2021-ALT-III-O Policy Two more interventions have been added to the
intervention 2021-ALT-II-O scenario (i) bus speed has been
regulated to 25 km h1 because of new
infrastructural development for dedicated bus
corridors, and (ii) parking fees have been hiked
(from Rs 1020 to Rs 60200) for private vehicles.
Rationale These assumptions have been made to assess the
inuence of increased share of public transport
use on VKT and air emissions.
2021-ALT-IV-O Policy One more intervention has been added to the
intervention 2021-ALT-III-O scenario (i) diesel particulate
Rationale These assumptions have been made to assess the
inuence of DPF on particulate matter from diesel
vehicle emissions.
(E) from vehicle use by incorporating the travel demand (A), structure
or characteristics of transport mode in-use (S) and emission factors
for specic vehicle mode and type of fuel use (F). This model has recent-
ly been used by Aggarwal and Jain (2014) in order to estimate carbon
dioxide emissions and energy demand from on-road vehicles in
the Delhi region. As suggested by Litman (2013) and observed by
Batterman et al. (2014) in Detroit (Michigan), the current study
has collected on-road vehicle activity data to make precise impact esti-
mates which are useful for future policy planning and urban transport
150 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157

infrastructure development. This vehicle characteristic information and
the demand modeling exercise are described by Aggarwal and Jain
(2014) in their latest publication on energy demand and carbon emis-
sions from on-road transport in the Delhi region and other details are pre-
sented in supplementary information (S.I.) Section S1.
2.3. Air quality modeling
In the current study, the ISC-AERMOD v.5.2 model was adopted to
estimate the pollutant concentrations from on-road transport, based
on the location of its release. This is due to its inherent improved ground
dispersion handling capability (Silverman et al., 2007; Singh et al.,
2006), better performance in complex terrain (US EPA, 2003) and
high sensitivity to small changes in wind speed (Faulkner et al., 2008).
AERMOD is an integrated system of three modules: a steady-state dis-
persion model (which is a Gaussian model and designed for short-
range ~ 50 km dispersion of air pollutants), AERMET (meteorological
data pre-processor) and AERMAP (a terrain pre-processor). It can esti-
mate emissions from point and line sources by treating them as area
sources (Cohan et al., 2011), along with improved handling of ground
dispersion (Silverman et al., 2007). Therefore, the current study con-
siders vehicles used in an area of 4 km2 as a single area source for
estimating pollutant concentrations for these specic locations. It can be
noted that vehicle activity in the grid of 2 2 km was estimated from traf-
c density assessment (using trafc count survey) and trip requirement
per road length (using GIS based road network inventorization). The
details of the emission grid development can be found in S.I. Section S2
and it can be noted that the grid-wise emission load was used as the pri-
mary input in AERMOD for predicting pollutant concentrations released
from a vehicular tailpipe only in the study region.
In case of AERMET pre-processor, the surface meteorological param-
eters namely humidity, wind speed, wind direction, sunshine hours and
precipitation (hourly-average, for the year 2007) were obtained for the
Safdarjung site (at 77.2E, 28.6N; at an elevation of 211 m) from the
India Meteorological Department (IMD), New Delhi. It was observed
that wind speed in the NCTD region was mild to moderate and mea-
sured up to 24 km/h with prominent wind direction as north and
northwest. On the other hand, the AERMAP preprocessor utilized the
United States Geological Survey (USGS) elevation data for the Datum
WGS 84 to estimate surface elevation and slope data for area source
grids in the study region. Subsequently, the output from AERMET and
AERMAP along with grid-wise emission loads was used to compute pol-
lutant concentrations (in annual average values, g/m3) from on-road
vehicular emissions for the years 2007 and 2021. These grid-wise values
were nally used to compute average pollutant concentrations (with
error value at 95 percentile) at district level using a Microsoft Excel
spreadsheet based model in order to identify areas of highest concerns
in terms of air quality (i.e., air pollution hot spots) in the NCTD region.
2.4. Health impact assessment
In order to determine the impact of air pollution released by vehicles
on human health, the current study has adopted a cross-sectional ap-
proach to collect primary health statistics for computing RR functions
for identied disorders. Lim et al. (2012) and HEI (2010) observed
that trafc emissions might cause or aggravate respiratory disorders
(ICD Codes A15A16, C30C39, J00J98) and cardiovascular disorders
(ICD Codes I00I99); therefore, the current research has collected
primary health statistics on mortality and morbidity for these disor-
ders in the study region. In addition, the disability-adjusted life-
years (DALYs) on exposure to vehicular pollution was also computed
by adopting discounting and age-weighting approach as proposed by
Prss-stn et al. (2003). The equations and assumptions for computing
mortality, morbidity and DALY for respiratory and cardiovascular disor-
ders have been given in Fig. 2.
Further, the study developed primary RR functions for respiratory
and cardiovascular impact on exposure to ambient NO2 and PM2.5 con-
centrations in the study region. It can be noted that a cross-sectional
epidemiological approach was adopted for developing mortality and
morbidity RR function (as explained by Jekel et al., 2007). Further,
Poisson (mortality) and logistic (morbidity) regression approaches
have been adopted for developing RR functions for 10 g/m3 increase
in NO2 and PM2.5 concentrations as summarized in Table 2. The details
for the development of RR functions for mortality and morbidity associ-
ated with respiratory and cardiovascular disorder in Delhi have been
provided in the S.I. Section S3.
3. Result and discussion
Table 3 summarizes the urban air quality and health impact of on-
road vehicular use in the study region. Section 3.1 describes the impacts
of vehicular pollution on ambient air quality in the year 2007 and 2021. In
Sections 3.2 and 3.3, the impact of vehicular pollutants on human health
have been discussed Delhi specic RR functions were developed for the
identied disorders (Section 3.2); and number of mortality cases, morbid-
ity cases and DALY on exposure to PM2.5 and NO2 concentrations have
been discussed in Section 3.3. Subsequently, air quality and health impact
of goods vehicles have been discussed in Section 3.4 in order to identify
their contribution to transport impacts, with the context of study region.
3.1. Urban air quality impact
Fig. 3 illustrates the spatial distribution (isopleths) of CO, NO2 and
PM2.5 concentrations from on-road passenger vehicles used in the
Delhi region in the year 2007 and 2021 under various policy scenarios.
It was observed that there is a pattern in spatial variations of pollutant
concentrations owing to vehicle count/volume, major trafc intersec-
tions, prevalent wind direction and uneven surface conditions.
In the year 2007, the modeled annual-average concentrations for CO,
NO2 and PM2.5 were 30 g/m3 (zonal concentrations ranged between 67
and 21 g/m3), 10 g/m3 (20 to 6 g/m3) and 0.45 g/m3 (1.3 to 0.3 g/
m3), respectively. Further, it was observed that Central, East, North-East,
New Delhi and West zones were the most affected regions due to on-
road vehicular pollution (24-hour concentration is depicted in S.I.
Fig. S4ac). The air pollution hotspots due to passenger vehicles are
Janakpuri in the west district (due to CO, NO2 and PM2.5), Ashram
Chowk in the south-east district (due to CO, NO2 and PM2.5), ITO in
the central district (for CO and PM2.5), Shahdra in the north-east district
(due to CO, NO2 and PM2.5), Kashmiri Gate in the north district (for CO
and PM2.5) and Madhuban Chowk in the north-west district (due to
NO2). The hotspots identied by the current study are in concurrence
with the results reported in the source apportionment study of Delhi in
the year 20062007 (CPCB, 2010). Further, it was estimated that ~86%
of the total VKT demand was from private vehicles, which resulted in
~85% of CO, ~44% of NO2 and ~63% of PM2.5 emissions of the total estimat-
ed emission load (for details, refer to S.I. Sub-section 1.2). These air pollut-
ants resulted in the deterioration of urban air quality in the study region.
In the 2021-BAU scenario, it was observed that passenger transport
will continue to affect ambient air quality in the NCTD region
resulting in 24 g/m3 for CO (Fig. 3a), 13 g/m3 for NO2 (Fig. 3b) and
0.36 g/m3 for PM2.5 (Fig. 3c), annually. It was estimated that CO and
PM2.5 concentrations decreased by ~19% and ~22%, respectively, com-
pared to the 2007-BASE year. It can be attributed to change in vehicle
technology and emission norms including BS V introduction, removal
of 2-stroke scooters and vehicles older than 15 years by the year 2021.
On the other hand, the concentration of NO2 was observed to increase
by ~33%, compared to the 2007-BASE year, owing to an ~37% increase
in NO2 emissions in the year 2021 (see S.I. Sub-section 1.2). The high
NO2 emission is due to a change in engine technology from 2-stroke
scooters (two-wheelers) to 4-stroke, which results in a 92% increase
in NO2 emissions from two wheelers only compared to 2007. This was
 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157 151

Equation 1: ASF model for modeling vehicle emission

=
10
Where, A is the transport activity (VKT in km/year; computed from demand modelling exercise), S is the modal structure (computed
from extensive vehicle user survey), E is the emission factor adopted to compute vehicle emission (g/km, collected from vehicle
profiling study conducted by Automotive Research Association of India); while, k represents the mode-type, j represents vehicle
technology (2/4 stroke) and i represents the fuel type (gasoline/diesel/CNG).

Equation 2: Relative risk function (RRi) and its 95% confidence interval (C.I.)
= exp .. and
. .= 1.96 . . ; + 1.96 . .

Where; is the change in pollutant concentration; is the regression coefficient describing relationship of air pollutant and health
impact; S. E. represents the standard error (which is calculated from the regression model)

Equation 3: Number of Excess Cases (N)

= 1 1
Where; is change in regression coefficient; p is probability of exposure; AP is apportioned population exposed to concentration

Equation 4: DALY calculation (which is equal to YLL + YLD)

1
= 1 ; =

Where, N is the number of deaths, L is the standard life expectancy at the age of death (in case of mortality) and duration of disability
(in case of morbidity), r is discount rate (standard r is assumed to be 0.03), I is number of incident cases and DW is disability weight (it
is assumed to be 1).

Fig. 2. Equation used in health impact assessment from vehicle emission sources.

further supplemented by an increase in CNG auto-rickshaws and busses, in PM2.5 concentrations were observed to be highest among the other
which added to ~21% NO2 emissions from public transport, when com- air pollutants i.e., ~ 38%, ~ 46% and ~ 58% for 2021-ALT-I, 2021-ALT-II
pared to the 2007-BASE year (details can be found in S.I. S1.2). and 2021-ALT-III scenarios, respectively, compared to the 2021-BAU
On analyzing the impact of alternate policy scenarios (2021-ALT-I to scenario as shown in Fig. 3c. These reductions in NO2 and PM2.5 concen-
2021-ALT-III), it was found that the reduction of emission load from trations were due to a shift of private vehicle users to public transport
passenger vehicles would directly improve ambient air quality (as and introduction of the stringent emission norm BS V in the year 2017
shown in Table 3 and Fig. 3). In the case of CO, it would decrease by (Table 1A).
~13% and ~24% in 2021-ALT-II and 2021-ALT-III scenarios, respectively, This scenario analysis proves that 2021-ALT-III is the best alternate
compared to the 2021-BAU scenario (Fig. 3a). However, the 2021-ALT-I scenario in order to meet the future travel demand with negligible air
scenario did not show any change because the BS V emission norm does quality impact in the NCTD region. Similar reductions were reported
not focus on CO reduction (see S.I. Table S2 for details). Similarly, in the by Chen et al. (2009) in Shanghai, where pollution control measure
case of NO2 concentrations, there would be a decrease of ~ 38%, ~ 45% as a policy intervention (scenario II) would reduce PM2.5 concentration
and ~42% under 2021-ALT-I, 2021-ALT-II and 2021-ALT-III scenarios, re- from passenger transport by ~17%, as compared to the baseline scenar-
spectively, compared to the 2021-BAU scenario; however, the decrease io. In the case of transport cities of Pakistan (Rawalpindi and Islamabad),
in NO2 concentration was observed to be less under the 2021-ALT-III Shabbir and Ahmad (2010) estimated that improvement of public
scenario due to a shift of passengers from the Delhi Metro, cars and transport would reduce PM2.5 emission by ~25%, compared to a baseline
two wheelers to CNG busses as shown in Fig. 3b. Further, the reductions scenario of the year 2030. This would, consequently, contribute to im-
provement in the overall quality of life in urban areas owing to reduced
Table 2 congestion, lesser exposure to air pollution and improved transport ac-
RR for mortality and morbidity due to respiratory and cardiovascular disorder in NCTD cessibility due to introduction of mass transit systems such as IMRTS
region. (Jain et al., 2014).
Health impact Pollutant RRa Condence intervalb Baseline
incidencec 3.2. Health impact of passenger transport
Respiratory mortality NO2 1.00016 0.9997154 1.00061 147
PM2.5 1.12824 1.1279848 1.128498 The cross-sectional health surveys were used to develop primary
Cardiovascular mortality NO2 1.00100 1.0002164 1.001783 325 health risk functions for the NCTD region in order to assess health
PM2.5 1.01502 1.0147869 1.015248 impact due to exposure to NO2 and PM2.5 concentrations, as discussed
Respiratory morbidity NO2 0.99505 0.9941494 0.995956 20,465
in Section 3.2.1. Subsequently, the excess cases of cardio-respiratory
PM2.5 1.00685 1.0066255 1.007083
Cardiovascular NO2 1.06138 1.061094 1.061658 2533 (respiratory and cardiovascular) disorder based mortality, morbidity
morbidity PM2.5 1.04560 1.0453666 1.045842 and DALY owing to vehicular NO2 and PM2.5 exposure is discussed in
a
10 g/m3 increment in NO2 and PM2.5 concentration.
Section 3.2.1. It is important to note that these cases have been devel-
b
95% C.I. oped using annual NO2/PM2.5 concentrations released from vehicular
c
B.I. is per 100,000. exhaust use under the proposed alternate scenarios.
152 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157

Table 3
Air quality and health impact of surface transport in NCTD region.

Modes of transport Policy scenarios

Passenger 2007-BASE 2021-BAU 2021-ALT-I 2021-ALT-II 2021-ALT-III

NO2 Conc. Annual 10 (3, 14) 13 (4, 19) 8 (3, 12) 7 (2, 10) 8 (3, 11)
(g/m3) 24 h 35 (23, 43) 45 (30, 56) 28 (18, 35) 25 (16, 31) 26 (17, 32)
Mortality Respiratory 4 (1, 15) 8 (1, 29) 5 (1, 18) 4 (1, 16) 4 (1, 16)
Cardiovascular 24 (5, 43) 47 (10, 83) 29 (6, 52) 26 (6, 46) 27 (6, 48)
Morbidity Respiratory
Cardiovascular 24,213 (24,108, 46,232 (46,034, 29,114 (28,987, 25,831 (25,719, 27,119 (27,001,
24,317) 46,430) 29,240) 25,944) 27,237)
DALY Respiratory 112 (1, 418) 215 (1, 806) 134 (1, 500) 118 (1, 443) 124 (1, 465)
Cardiovascular 25,721 (25,076, 49,124 (47,886, 30,923 (30,150, 27,434 (26,750, 28,803 (28,084,
26,365) 50,362) 31,695) 28,118) 29,522)
PM2.5 Conc. Annual 0.46 (0.15, 0.69) 0.36 (0.11, 0.54) 0.22 (0.07, 0.33) 0.15 (0.06, 0.30) 0.15 (0.05, 0.23)
(g/m3) 24 h 1.58 (0.95, 2.03) 1.19 (0.71, 1.53) 0.75 (0.43, 0.95) 0.66 (0.38, 0.84) 0.51 (0.30, 0.65)
Mortality Respiratory 136 (102, 169) 155 (117, 193) 96 (72, 119) 85 (64, 105) 65 (49, 82)
Cardiovascular 17 (17, 17) 19 (19,19) 12 (12, 12) 10 (10, 11) 8 (8, 8)
Morbidity Respiratory 1073 (716, 1430) 1225 (817, 1632) 754 (503, 1005) 667 (445, 889) 516 (344, 688)
Cardiovascular 866 (862, 871) 989 (984, 994) 609 (606, 612) 539 (536, 542) 417 (415, 419)
DALY Respiratory 5210 (1382, 6610) 5949 (1578, 7547) 3663 (971, 4648) 3243 (860, 4115) 2510 (665, 3185)
Cardiovascular 1370 (1359, 3810) 1564 (1551, 4349) 963 (954, 2678) 852 (845, 2371) 660 (654, 1835)

Overall 2007-BASE-O 2021-BAU-O 2021-ALT-I-O 2021-ALT-II-O 2021-ALT-III-O 2021-ALT-IV-O

NO2 Conc. Annual 16 (5, 24) 26 (9, 37) 25 (8, 36) 16 (5, 23) 17 (6, 24) 17 (6, 24)
(g/m3) 24 h 58 (37, 72) 88 (58, 109) 85 (56, 105) 55 (36, 68) 57 (37, 71) 57 (37, 71)
Mortality Respiratory 6.6 (1, 25) 14.8 (1, 56) 14.3 (1, 54) 9.3 (1, 35) 9.7 (1, 36) 9.7 (1, 36)
Cardiovascular 40 (6, 72) 91 (15, 162) 88 (15, 156) 57 (10, 102) 59 (10, 106) 59 (10, 106)
Morbidity Respiratory
Cardiovascular 39,581 (29,197, 86,761 (67,641, 83,875 (65,389, 56,077 (43,447, 57,960 (44,996, 57,960 (44,996,
39,782) 87,208) 84,308) 56,375) 58,267) 58,267)
DALY Respiratory 186 (1, 698) 419 (1, 1570) 404 (1, 1514) 263 (1, 987) 273 (1, 1022) 273 (1, 1022)
Cardiovascular 42,068 (30,371, 92,278 (70,370, 89,202 (68,026, 59,598 (45,194, 61,603 (46,806, 61,603 (46,806,
43,170) 94,751) 91,589) 61,172) 63,230) 63,230)
PM2.5 Conc. Annual 1.20 (0.39, 1.80) 0.68 (0.22, 1.02) 0.53 (0.17, 0.80) 0.35 (0.11, 0.53) 0.28 (0.09, 0.42) 0.09 (0.03, 0.13)
(g/m3) 24 h 4.15 (2.49, 5.33) 2.29 (1.32, 2.90) 1.80 (1.04, 2.28) 1.18 (0.68, 1.50) 0.94 (0.55, 1.20) 0.28 (0.17, 0.36)
Mortality Respiratory 355 (328, 539) 291 (268, 442) 229 (207, 341) 150 (137, 226) 120 (110, 182) 37 (27, 112)
Cardiovascular 44 (30, 55) 36 (24, 45) 28 (19, 34) 19 (13, 23) 15 (10, 18) 5 (3, 11)
Morbidity Respiratory 2814 (2300, 4593) 2299 (1876, 3748) 1807 (1446, 2889) 1187 (957, 1911) 950 (769, 1537) 293 (237, 473)
Cardiovascular 2268 (1555, 2791) 1855 (1264, 2280) 1458 (993, 1758) 959 (653, 1164) 767 (522, 936) 237 (172, 576)
DALY Respiratory 13,618 (12,197, 11,153 (9972, 8768 (7692, 13,338) 5767 (5092, 8834) 4614 (4095, 7106) 1426 (1055, 4380)
21,114) 17,285)
Cardiovascular 3589 (2451, 4432) 2935 (1992, 3620) 2306 (1565, 2791) 1516 (1029, 1847) 1213 (823, 1485) 375 (271, 915)

3.2.1. Primary RR function for the Delhi region
Table 2 summarizes the health risk associated on exposure to ambi-
ent air pollutant concentrations in the context of the NCTD region. The
estimated RR functions for respiratory mortality was 1.00016 for NO2
and 1.12824 for PM2.5 concentrations; while in the case of cardiovascu-
lar mortality, RR was 1.00100 for NO2 and 1.01502 for PM2.5. It can be
noted that exposure to PM2.5 concentrations will contribute to a huge
health burden from cardio-respiratory mortality; however, the number
of cases may be lower for PM2.5 owing to its lower quantum, when
compared to NO2 concentrations. Beelen et al. (2008) have reported a
similar impact of high PM2.5 concentrations in the Dutch region
where 10 g/m3 increases in PM2.5 concentrations would contribute to
1.22 RR and 1.04 RR for cardio-respiratory disorders, respectively. On
the other hand, it is important to note that exposure to PM2.5 would
have a higher cardiovascular morbidity in the present study region, com-
pared to respiratory morbidity (Table 2). Similar observations have been
made by the Global Burden of Disease study (Lim et al., 2012).
Concurrently, the widely adopted WHO's RR functions may appear
lower in magnitude, when compared to the primary RR functions devel-
oped by the current study. For instance, the global RR functions for re-
spiratory mortality due to exposure to PM2.5 concentration were ~11%
lower as compared to RR functions developed in the current study
(Table 2). The details related to global RR functions have been provided
in S.I. S3.5. This is due to the higher exposure to ambient air pollution in
the study region, as compared to other megacities and regions around
the globe. Similar results were reported by Zhang et al. (2014) for
Chinese cities, where exposure to high PM2.5 concentrations resulted
in 1.24 RR for every 10 g/m3 increase in case of cardiovascular mortal-
ity. It is important to note that very few studies have quantied health
risk (RR) associated with NO2 exposure. For instance, HEI (2010) report-
ed that Hong Kong, China (Wong et al., 1999, 2002) and Seoul, South
Korea (Ha et al., 2003) based studies have a positive association of
NO2 and health cases, which further can be related to high motor vehicle
use in both the cities.
3.3. Health burden of PM2.5 and NO2 exposure
The estimated decrease in PM2.5 and NO2 concentrations in the pro-
posed alternate scenarios would translate into reduced health burden,
as presented in Table 3. In 2021-BAU, there is an increase of ~14% and
~ 93% in cardio-respiratory mortality and morbidity cases due to a
change in vehicular PM2.5 and NO2 concentrations, respectively, when
compared to the 2007-BASE year. In alternate scenarios, mortality and
morbidity from respiratory and cardiovascular disorders would
decrease by ~ 38%, ~ 45% and ~ 58% under 2021-ALT-I, 2021-ALT-II,
2021-ALT-III scenarios, respectively, compared to the 2021-BAU scenar-
io due to a decrease in PM2.5 concentrations.
It can be noted that the health burden reduced in line with the re-
duction of ambient air quality impact as depicted in Fig. 3. The change
in modal structure (shift from private to public vehicles), stringent
vehicle technology (BS VI to BS V emission norm), dedicated bus infra-
structure (bus speed regulation and increase in CNG use) and pricing
mechanism (hike in parking fee for private vehicles) are the primary
reasons behind change in concentration and cardio-respiratory impact
 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157
a) CO concentration in 2007 and 2021
b) NO2 concentration in 2007 and 2021
c) PM2.5 concentration in 2007 and 2021
2007-BASE 2021-BAU 2021-ALT-I
Fig. 3. Isopleths of air pollutants from passenger transport in the NCTD region.
in the study region. The detailed emission modeling results have been
presented in the Sub-section S1.2 of the S.I. document, with the reasons
behind changes in emission load under different scenarios. Among the
three alternate scenarios, 2021-ALT-III emerges as the best-estimate
scenario with the lowest impact on cardio-respiratory mortality. In
case of PM2.5 concentration, the annual average is reported to decrease
by 58%, when compared to the 2021-BAU scenario, which results in the
saving of ~4 lives per million people in the NCTD region. On the other
hand, the NO2 concentration is estimated to decrease by 42%, which re-
sults in saving of one life per million people in the 2021-ALT-III scenario,
compared to the 2021-BAU scenario. Woodcock et al. (2009) have esti-
mated similar health benets on the introduction of lower air emitting
vehicles and restricting vehicle use, with examples of the London and
Delhi regions. Similar results have been reported by Rojas-Rueda et al.
(2013) for the Barcelona region, where a decrease in car trips resulted
in a reduction of nearly two cases of respiratory disorders and seven
cases of cardiovascular diseases. In the case of a developing country con-
text, Chen et al. (2009) reported similar results for Shanghai (China)
where ~ 25% reduction in exposure to PM concentrations would help
in avoiding 2452 hospital admissions (~ morbidity) in the year 2020.
These observations make it important to reduce private vehicle use in
urban areas and as strongly emphasized by the current study.
Further, the reduced health burden was quantied as DALYs saved
on implementation of alternate scenarios with estimated reductions in
air pollution exposure in the NCTD region. It was observed that DALYs
153
2021-ALT-II 2021-ALT-III
annually would reduce from 6164 (in 2021-BAU) to 2634 (in 2021-
ALT-III) in the case of respiratory disorders and 50,688 (in 2021-BAU)
to 29,463 (in 2021-ALT-III) in the case of cardiovascular disorders.
This would result in cumulative reduction of ~44% DALYs from vehicular
air pollutants, when compared to the 2021-BAU scenario, which is near-
ly of the same levels of health burden as estimated for the 2007-BASE
year. Similar results have been reported by Rao et al. (2013) where
they found that ~50% decrease in outdoor PM2.5 concentrations reduced
5% DALYs associated with cardiovascular, respiratory and lung cancer at
the global level in the year 2030, compared to the 2005 scenario. Fur-
ther, Dhondt et al. (2013) have reported a gain of ~1650 DALYs on re-
duction of ~ 23% car usage in Flanders and Brussels (Belgium), which
was almost double the DALYs saved due to respiratory disorders in
the current study. This clearly shows that adoption of proposed policy
interventions i.e., increase in public transport use, adoption of stringent
emission norms and hike in parking price as described in the 2021-ALT-
III scenario would aid in reducing air quality as well as improving life
expectancy in the context of the NCTD region.
3.4. Health impact of goods transport
The authors observed that goods transport contributes to ~ 5.5% of
VKT demand of the total on-road vehicles in the NCTD region in the
year 2007; however, it contributed to ~ 23%, ~ 48% and ~ 62% of
the total CO, NO2 and PM2.5 emissions load, respectively (details are
154 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157
provided in S.I. Section 1.7). It can be noted that overall transport would
contribute to 4.15 g/m3 PM2.5 concentration in the year 2007; whereas,
a recent study by Sindhwani and Goyal (2014) estimated that the trans-
port sector in the Delhi region contributes around 8 g/m3 PM10 concen-
trations for the year 2007. The air quality estimate is comparable to the
current study, if the impact of the assumed transport activity data is
considered. This study has adopted on-road vehicle information along
with transport demand model based VKM information to predict emis-
sion load; while, Sindhwani and Goyal (2014) have adopted vehicle
population data, with assumed growth rate and average VKM per
mode information collected from secondary sources. This clearly indi-
cates that a rise in vehicle activity would aid in the deterioration of
urban air quality in the study region. Further, under the 2021-BAU-O
scenario, the goods transport would contribute to 27%, 49% and 55% of
the total CO, NO2 and PM2.5 emission loads, respectively (details are pro-
vided in S.I. S1 Table S3); contributing to more than half of PM2.5 emis-
sions compared to passenger vehicles in the year 2021. This is due to
high PM2.5 and low CO emitting goods transport (primarily diesel and
CNG goods vehicle), when compared to majority of gasoline based
passenger vehicles in the NCTD region. Similar results were observed
in a study on diesel operated trucks conducted by Goyal (2007),
where ~50% of the total PM10 emission was attributed to diesel based
HDV vehicles in the Delhi region. Therefore, it can be concluded that
goods transport would continue to emit pollutants, higher than passenger
a) CO concentration in 2007 and 2021
b) NO2 concentration in 2007 and 2021
c) PM2.5 concentration in 2007 and 2021
2007-BASE-O 2021-BAU-O 2021-ALT-I-O
Fig. 4. Isopleths of air pollutants from overall transport in the NCTD region.
transport, making it necessary to look for feasible solutions to mitigate
their impacts.
Fig. 4 shows the isopleths of air pollutant concentration from overall
transport use in the NCTD region, where it was found that air pollutant
hot-spots are similar to the assessment made in Section 3.1, owing to
private and goods transport activities for the years 2007 and 2021.
Four alternate scenarios (2021-ALT-I-O to 2021-ALT-IV-O) were pro-
posed with an aim to improve the air quality and health impact from
the surface transport system as explained in Table 1B. It is important
to note that dispersion patterns are similar to passenger vehicles (see
Fig. 3); however, the quantum of pollutant concentration has almost
doubled owing to the huge contribution from goods transport use. The
analysis shows that 2021-ALT-IV-O would be the best estimate scenario
with minimum air quality and health burden in the NCTD region as
summarized in Table 2.
It can be observed that under the 2021-ALT-IV-O scenario, ambient
air quality would improve by ~23% for CO, ~35% for NO2 and ~87% for
PM2.5, compared to the 2021-BAU-O scenario (Fig. 4ac). This would
translate in ~87% and ~35% reduction in PM2.5 and NO2 based cases of
cardio-respiratory mortality and morbidity, respectively, compared to
the 2021-BAU-O scenario. Further, the DALYs due to exposure to vehicu-
lar PM2.5 and NO2 concentrations would reduce from 0.11 million (2021-
BAU-O) to 0.07 million (2021-ALT-IV-O), which would save almost elev-
en times more DALYs compared to interventions proposed under the
2021-ALT-II- 2021-ALT-III- 2021-ALT-IV-
O O O
 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157
2021-ALT-III scenario in the case of passenger transport. Similar results
have been reported by Cesaroni et al. (2012) while considering policy in-
terventions for goods transport in Rome. They found that a 24% decrease
in PM2.5 concentration contributed to gain 3.4 days per person. In addi-
tion, the results of the current study are in agreement with a recent policy
analysis by Guttikunda and Mohan (2014), who recommended that
adoption of BS V, removal of older vehicles and improvement of the
goods vehicles would minimize vehicle exhaust emission and its impact
on cities of India.
4. Conclusions
Growing travel demand and increasing dependence on private
vehicles would continue to pollute the ambient environment, increase
the economic burden of oil dependence and deteriorate the overall
quality of life in urban areas. A similar condition has been observed in
the case of the Delhi (NCTD) region, where rapid urbanization and in-
crease in ownership of vehicles are noted to have a damaging impact
on health and welfare of its residents. In the present study, an integrat-
ed source-to-receptor assessment paradigm was adopted to showcase
how an effective roadmap can be developed with an aim to minimize
negative externalities and maximizing benets of mobility at the city
level, especially in the context of a developing region.
The current study, for the rst-time, has collected primary vehicle
and health statistics for the National Capital territory of Delhi, including
vehicle characteristics, road length inventory, vehicle density and prev-
alence of cardio-respiratory disorder in the Delhi region. This adds value
to the accumulating body of trafc impact assessment studies in the
Indian context, where a majority of them have not moved beyond reg-
istration or PUC information available with respective departments.
Further, the air quality assessment in the Delhi region is unique because
it estimates pollutant concentration from on-road vehicles at the local
level, by quantifying gridded emissions load and vehicle density surveys
through the AERMOD dispersion model. Further, correlation between
ambient air pollution and cardio-respiratory disorders in Delhi has
been developed as a primary RR function for excess cases of mortality
and morbidity on exposure to localized vehicular NO2 and PM2.5 con-
centrations in the study region.
The extensive modeling and scenario analysis exercise results of the
current study show that maximum health benets (in terms of mortal-
ity, morbidity and DALY) and improvement in ambient air quality
would occur when private vehicle users would shift to public transport
such as the Delhi Metro and busses, the parking fee for private vehicles
will increase and stringent vehicular emission norms would be intro-
duced in the NCTD region (as in case of 2021-ALT-III scenario). Many re-
cent reports suggest that the regulatory bodies in India are considering
the implementation of the suggested policies in order to reduce the
dependence on private vehicles and their impacts of air quality. For in-
stance, the IMRTS project of Delhi is proceeding in-line with the project
plan (DMRC, 2015), while a recent media report suggests that the
Government of India is very keen on introducing stringent vehicle emis-
sion norms (BS V & BS VI) to reduce associated emissions (Sethi, 2015).
On the inclusion of goods transport with DPF tted HDV's, the overall
health impact in terms of DALY was observed to reduce by ~ 87% in
the case of exposure to PM2.5 concentrations. The proposed interven-
tions would improve the state of the urban transport system as it
would aid in the reduction of congestion and delays, trafc noise and
potential trafc injuries, as well as mitigate greenhouse gas emissions
owing to a reduction in VKT demand and emissions in the near future.
This was reported in a recent study by Aggarwal and Jain (2014)
where a similar policy intervention in the Delhi region would reduce en-
ergy demand and CO2 emissions from on-road vehicles by ~ 17% and
~14%, respectively, compared to the 2021-BAU scenario. These results
are in-line with the conclusion of Olowoporoku et al. (2010), who indi-
cated that there is a growing need for a policy shift from an air quality
155
hotspot management approach to a more holistic management of ve-
hicular pollution within wider spatial administrative areas.
There are a few methodological limitations of the current study.
Firstly, the vehicle emission factors of ARAI are weighted-average fac-
tors, which do not consider the impact of speed variation, congestion
and driver behavior during emission quantication. The study could
not identify any PM2.5 emission factor for Indian vehicles, and therefore
adopted a fuel-based PM2.5/PM10 ratio to estimate PM2.5 emissions load
in the Delhi region. This shows that there is a growing need to improve
existing emission factors by considering the above city's limitations, in
order to estimate an accurate level of air pollutant concentrations
from Indian vehicles. Secondly, the weather condition of year 2007
was used to predict air quality impact in the year 2021. There is develop-
ment in the weather simulation research, which can help in predicting
future weather impact on air quality conditions at an urban scale. The
third major limitation of the current study was the adoption of prevalence
survey and logistic regression to the compute morbidity based impact in
the Delhi region. In initial stages of the study, it was identied that health
statistics in Delhi were dispersed (with individual facility), incomplete
(temporally, may vary among facility) and imprecise in nature. Therefore,
the current study conducted extensive health surveys in Delhi to estimate
the prevalence of health disorders; however, it was concluded that there
is a huge scope for future work in terms of creating and maintaining accu-
rate health databases at the city and national levels.
Conclusively, this study provides useful information to policy
makers, planners, epidemiologists and governmental agencies. The pro-
posed integrated source-to-receptor assessment paradigm and sce-
nario based analysis proves to be a one-of-its kind framework, which
can aid in maximizing the benets of surface transport in the context
of urban areas. With the inclusion of goods, it can be concluded that
the growing impact of vehicle use on urban air quality and health in
major parts of the world can be reduced in an exponential manner. Sce-
nario analysis emerges as an important tool to predict the impact of pro-
posed transport policy interventions on future air quality and human
health in urban areas, as explained with the case of the Delhi region.
This will help us to ensure that urban areas can continue to develop
and provide a healthy life, when their travel requirements are met
through an appropriate selection of policy interventions as shown
with the case of the Delhi region. The ndings of the current study
would provide useful information for researchers, policy makers and
infrastructure planning organizations in urban areas of developing
regions of the world.
Acknowledgements
The authors greatly acknowledge the University Grant Commis-
sion (UGC-Ref. No.: 3208/NET-DEC. 2010) and TERI University for
their nancial assistance and TERI for infrastructural support. We
would also like to acknowledge all the students who had contribut-
ed in data collection using questionnaire surveys for vehicle charac-
terization, distribution of vehicles in the Delhi region and health
status at household levels. The editors and four anonymous referees
are gratefully acknowledged. The authors also thank Mr. Neeraj
Sharma (TERI University) for the assistance in language editing.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.envint.2015.06.010.
References
Aggarwal, P., Jain, S., 2014. Energy demand and CO2 emissions from urban on-road
transport in Delhi: current and future projections under various policy measures.
J. Clean. Prod. http://dx.doi.org/10.1016/j.jclepro.2014.12.012.
Auerbach, A., Hernandez, M.L., 2012. The effect of environmental oxidative stress on air-
way inammation. Curr. Opin. Allergy Clin. Immunol. 12 (2), 133139.
156 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157
Aunan, K., 1995. Exposure-response Functions for Health Effect of Air Pollutants Based on
Epidemiological Findings Prepared by the, Center for International Climate and Envi-
ronmental Research, Oslo (Available http://www.cicero.uio.no/media/81.pdf, ISSN:
0804-4562).
Batterman, S., Burke, J., Isakov, V., Lewis, T., Mukherjee, B., Robins, T., 2014. A comparison
of exposure metrics for trafc-related air pollutants: application to epidemiology
studies in Detroit, Michigan. Int. J. Environ. Res. Public Health 11 (9), 95539577.
Beelen, R., Hoek, G., Piet, A., Brandt van der, Goldbohm, R.A., Fischer, C.P., Schouten, L.J.,
Armstrong, B., Brunekreef, B., 2008. Long-term exposure to trafc-related air pollu-
tion and lung cancer risk. Epidemiol. 19, 702710.
Brook, R.D., Rajagopalan, S., Pope, C.A., Brook, J.R., Bhatnagar, A., Diez-Roux, A.V., HolguinF,
Hong Y., Luepker, R.V., Mittleman, M.A., Peters, A., Siscovick, D., Smith, S.C., Whitsel, L.,
Kaufman, J.D., Iii, C.A.P., 2010. Particulate matter air pollution and cardiovascular
disease: an update to the scientic statement from the American Heart Association.
Circulation 121, 23312378.
Cesaroni, G., Boogaard, H., Jonkers, S., Porta, D., Badaloni, C., Cattani, G., Hoek, G., 2012.
Health benets of trafc-related air pollution reduction in different socioeconomic
groups: the effect of low-emission zoning in Rome. Occup. Environ. Med. 69 (2),
133139.
Chen, C.H., Kan, H.D., Huang, C., Li, L., Zhang, Y.H., Chen, R.J., Chen, B.H., 2009. Impact of
ambient air pollution on public health under various trafc policies in Shanghai,
China. Biomed. Environ. Sci. 22 (3), 210215.
Cohan, A., Wu, J., Dabdub, D., 2011. High-resolution pollutant transport in the San Pedro
Bay of California. Atmos. Pollut. Res. 2 (3), 237246.
CPCB, 2010. Air quality monitoring, emission inventory and source apportionment study
for Indian cities. National Science Report of Central Pollution Control, India (Available
http://cpcb.nic.in/Source_Apportionment_Studies.php).
CPCB, 2012. Epidemiological study on effect of air pollution on human health (Adults) in
Delhi. Central Pollution Control Board HMS/01/2012. Government of National Capital
Territory of Delhi (NCTD), India.
Dhondt, S., Kochan, B., Beckx, C., Lefebvre, W., Pirdavani, A., Degraeuwe, B., Bellemans, T.,
Int Panis, L., Macharis, C., Putman, K., 2013. Integrated health impact assessment of
travel behaviour: model exploration and application to a fuel price increase. Environ.
Int. 51, 4558.
DMRC, 2015. Current Status of the Delhi Metro Network. (DMRC website: accessed
bhttp://www.delhimetrorail.com/whatnew_details.aspx?id=WqrZpjgGVHolldN on
02-06-2015).
Dockery, D.W., 2001. Epidemiologic evidence of cardiovascular effects of particulate air
pollution. Environ. Health Perspect. 109, 483486.
Dockery, D.W., Pope, C.A., 2006. Health effects of ne particulate air pollution: lines that
connect (2006 critical review). J. Air Waste Manag. Assoc. 56, 709742.
DoIT, 2013. Environmental concerns. Economic Survey of Delhi 201213. Department of
Planning, Government of NCT of Delhi, India.
Eisner, M.D., Anthonisen, N., Coultas, D., Kuenzli, N., Perez-Padilla, R., Postma, D., Romieu,
I., Silverman, E.K., Balmes, J.R., 2010. An ofcial American Thoracic Society public
policy statement: novel risk factors and the global burden of chronic obstructive
pulmonary disease. Am. J. Respir. Crit. Care Med. 182 (5), 693718.
Faulkner, W.B., Shaw, B.W., Grosch, T., 2008. Sensitivity of two dispersion models
(AERMOD and ISCST3) to input parameters for a rural ground-level area source.
J. Air Waste Manag. Assoc. 58 (10), 12881296.
Gokhale, S., Khare, M., 2004. A review of deterministic, stochastic and hybrid vehicular
exhausts emission models. Int. J. Transp. Manag. 2 (2), 5974.
Goyal, P., 2007. A Detailed Study to Ascertain the Effect of Diesel Operated Trucks,
Tempos, Three Wheelers and other Commercial Vehicles on the Ambient Air Quality
of Delhi. Department of Environment, Govt. of NCT of Delhi.
Goyal, P., Kumar, A., Mathematical modeling of air pollutants: an application to Indian
urban city, air quality-models and applications. Prof. Dragana Popovic (Ed.); ISBN:
978-953-307-307-1, InTech; 2011.
Gurjar, B.R., van Aardenne, J.A., Lelieveld, J., Mohan, M., 2004. Emission estimates and
trends (19902000) for megacity Delhi and implications. Atmos. Environ. 38
(33), 56635681.
Gurjar, B.R., Jain, A., Sharma, A., Agarwal, A., Gupta, P., Nagpure, A.S., Lelieveld, J.,
2010. Human health risks in megacities due to air pollution. Atmos. Environ.
44, 46064613.
Guttikunda, S.K., 2008. Estimating health impacts of urban air pollution. SIM-air Working
Paper Series 06 p. 2008.
Guttikunda, S.K., Calori, G., 2013. A GIS based emissions inventory at 1 km 1 km spatial
resolution for air pollution analysis in Delhi, India. Atmos. Environ. 7, 101111.
Guttikunda, S.K., Goel, R., 2013. Health impacts of particulate pollution in a
megacityDelhi, India. Environ. Dev. 6, 820.
Guttikunda, S.K., Mohan, D., 2014. Re-fueling road transport for better air quality in India.
Energy Policy 68, 556561.
Ha, E.-H., Lee, J.-T., Kim, H., Hong, Y.-C., Lee, B.-E., Park, H.-S., Christiani, D.C., 2003. Infant
susceptibility of mortality to air pollution in Seoul, South Korea. Pediatrics 111,
284290.
HEI. Health effects of outdoor air pollution in developing countries of Asia: a literature
review. Prepared by the HEI International Scientic Oversight Committee, Health
Effects Institute; Special Report 15; 2004.
HEI. Trafc-related air pollution: a critical review of the literature on emissions, exposure,
and health effects. HEI panel on the health effects of trafc-related air pollution. Pre-
pared by the Health Effects Institute, HEI special report 17; 2010.
Jain, S., Khare, M., 2008. Urban air quality in mega cities: a case study of Delhi City using
vulnerability analysis. Environ. Monit. Assess. 136 (3), 257265.
Jain, S., Khare, M., 2010. Adaptive Neuro-Fuzzy Modelling for Prediction of Ambient CO
Concentration at Urban Intersections and Roadways. Air Qual. Atmos. Health 3,
203212.
Jain, S., Aggarwal, P., Kumar, P., Singhal, S., Sharma, P., 2014. Identifying public preferences
using multi-criteria decision making for assessing the shift of urban commuters from
private to public transport: a case study of Delhi. Transport. Res. F: Trafc Psychol.
Behav. 24, 6070.
Jekel, J.F., Katz, D.L., Elmore, J.G., Wild, D., 2007. Epidemiology, Biostatistics and Preventive
Medicine. Elsevier Health Sciences.
Khanna, P., Jain, S., Sharma, P., Mishra, S., 2011. Impact of increasing mass transit share on
energy use and emissions from transport sector for National Capital Territory of
Delhi. Transp. Res. D Transp. Environ. 16 (1), 6572.
Kumar, P., Gurjar, B., Nagpure, A., Harrison, R., 2011. Preliminary estimates of nanoparticle
number emissions from road vehicles in megacity Delhi and associated health im-
pacts. Environ. Sci. Technol. 13, 55145521.
Kumar, P., Jain, S., Gurjar, B.R., Sharma, P., Khare, M., Morawska, L., Britter, R., 2013. Can a
Blue Sky return to Indian megacities? Atmos. Environ. 71, 198201.
Lahiri, T., Ray, M.R., 2012. Epidemiological study on effect of air pollution on human
health (adults) in Delhi. Environmental Health Management Series: EHMS/01/2012.
Central Pollution Control Board, New Delhi.
Lebel, L., Garden, P., Banaticla, M.R.N., Lasco, R.D., Contreras, A., Mitra, A.P., Sharma, C.,
Nguyen, H.T., Ooi, G.L., Sari, A., 2007. Integrating Carbon Management into the Devel-
opment Strategies of Urbanizing Regions in Asia Implications of Urban Function,
Form, and Role. J. Indus. Eco. 11, 6181.
Lim, S.S., Vos, T., Theo, V., Abraham, D.F., Goodarz, D., Kenji, S., Heather, A.-R., AlMazroa,
Mohammad A., et al., 2012. A comparative risk assessment of burden of disease
and injury attributable to 67 risk factors in 21 regions, 19902010: a systematic anal-
ysis for the Global Burden of Disease Study 2010. Lancet 380, 22242260.
Lipfert, F.W., Wyzga, R.E., Baty, J.D., Miller, J.P., 2006a. Trafc density as a surrogate mea-
sure of environmental exposures in studies of air pollution health effects: long-term
mortality in a cohort of US veterans. Atmos. Environ. 40, 154169.
Lipfert, F.W., Baty, J.D., Miller, J.P., Wyzga, R.E., 2006b. PM2.5 constituents and related air
quality variables as predictors of survival in a cohort of U.S. military veterans. Inhal.
Toxicol. 18, 646657.
Litman, T., 2013. Generated Trafc and Induced Travel Implications for Transport Plan-
ning. Victoria Transport Policy Institute (Available http://www.vtpi.org/gentraf.pdf).
Lu, F., Xu, D., Cheng, Y., Dong, S., Guo, C., Jiang, X., Zheng, X., 2015. Systematic review and
meta-analysis of the adverse health effects of ambient PM2.5 and PM10 pollution in
the Chinese population. Environ. Res. 36, 196204.
Mohan, M., Bhati, S., Rao, A., 2011. Application of air dispersion modelling for exposure
assessment from particulate matter pollution in mega city Delhi. Asia Pac. J. Chem.
Eng. 6, 8594.
MoRTH, 2011. Road Transport Year Book (20072009) (Volume I). Ministry of Road
Transport and Highway, Government of India.
MPNG, 2003. Auto Fuel Policy of India Prepared by, Government of India (Available
http://petroleum.nic.in/docs/autopol.pdf).
Nidhi, Jayaraman, G., 2007. Air quality and respiratory health in Delhi. Environ. Monit. As-
sess. 135, 313325.
OECD, 2012. OECD Environmental Outlook to 2050: The Consequences of Inaction Key
Facts and Figures Prepared by. OECD.
Olowoporoku, D., Hayes, E., Longhurst, J., Parkhurst, G., 2010. The rhetoric and realities of
integrating air quality into the local transport planning process in English local au-
thorities. J. Environ. Manag. 101, 2332.
Pande, J.N., Bhatta, N., Biswas, D., Pandey, R.M., Ahluwalia, G., Siddaramaiah, N.H.,
Khilnani, G.C., 2002. Outdoor air pollution and emergency room visits at a hospital
in Delhi. Indian J. Chest Dis. Allied Sci. 44 (1), 1320.
Prss-stn, A., Mathers, C., Corvaln, WA., Introduction and methods: assessing the
environmental burden of disease at national and local levels. WHO Environmental
Burden of Disease series, No. 1. Prepared by, World Health Organization, 2003.
Rao, S., Pachauri, S., Dentener, F., Kinney, P., Klimont, Z., Riahi, K., Schoepp, W., 2013.
Better air for better health: forging synergies in policies for energy access, climate
change and air pollution. Glob. Environ. Chang. 23 (5), 11221130.
Rojas-Rueda, D., de Nazelle, A., Teixid, O., Nieuwenhuijsen, M.J., 2013. Health impact as-
sessment of increasing public transport and cycling use in Barcelona: a morbidity and
burden of disease approach. Prev. Med. 57 (5), 573579.
Sahai, S.N., Bishop, S., 2009. Bus System Reform in Delhi. Publication of DIMTS, India.
Sahu, S.K., Beig, G., Parkhi, N.S., 2011. Emissions inventory of anthropogenic PM2.5
and PM10 in Delhi during Commonwealth Games 2010. Atmos. Environ. 45 (34),
61806190.
Sethi, K., 2015. Government Wants BS V and VI Norms to be Implemented Sooner Than
Planned April 27, (Accessed bhttp://auto.ndtv.com/news/government-wants-bs-v-
and-vi-norms-to-be-implemented-sooner-than-planned-758506N, 03-06-2015).
Shabbir, R., Ahmad, S.S., 2010. Monitoring urban transport air pollution and energy de-
mand in Rawalpindi and Islamabad using leap model. Energy 35 (5), 23232332.
Sharma, P., Khare, M., 2001. Modelling of vehicular exhausts a review. Transp. Res. D
Transp. Environ. 6 (3), 179198.
Silverman, K., Tell, J., Sargent, E., 2007. Comparison of the industrial source complex and
AERMOD dispersion models: case study for human health risk assessment. J. Air
Waste Manag. Assoc. 57, 14391446.
Simon, N.B., Alberini, A., Sharma, P.K., 1997. The Health Effects of Air Pollution in Delhi,
India. No. 1860. World Bank, Development Research Group.
Sindhwani, R., Goyal, P., 2014. Assessment of trafc-generated gaseous and particu-
late matter emissions and trends over Delhi (20002010). Atmos. Pollut. Res. 5,
438446.
Singh, R.B., Desloges, C., Sloan, J.J., 2006. Application of a microscale emission factor model
for particulate matter to calculate vehicle-generated contributions to ne particulate
emissions. J. Air Waste Manag. Assoc. 56 (1), 3747.
Suresh, Y., Sailaja Devi, M.M., Manjari, V., Das, U.N., 2000. Oxidant stress, antioxidants, and
nitric oxide in trafc police of Hyderabad, India. Environ. Pollut. 109, 321325.
 P. Aggarwal, S. Jain / Environment International 83 (2015) 146157
US EPA, 2003. Comparison of regulatory design concentrationsAERMOD vs. ISCST3,
CTDMPLUS, ISC-PRIME. Staff Report. EPA-454/R-03-002. U.S. Environmental Protec-
tion Agency, Ofce of Air Quality Planning and Standards, Emissions Monitoring
and Analysis Division, Research Triangle Park, North Carolina.
Vlachokostas, C., Nastis, S.A., Achillas, C., Kalogeropoulos, K., Karmiris, I., Moussiopoulos,
N., Chourdakis, E., Banias, G., Limperi, N., 2010. Economic damages of ozone air pollution
to crops using combined air quality and GIS modelling. Atmos. Environ. 44 (28),
33523361.
Watson, J.G., Chow, J.C., 2007. Receptor models for source apportionment of suspended
particles. In: Murphy, B.L., Morrison, R.D. (Eds.), Introduction to Environmental
Forensics, 3rd ed. Academic Press, Burlington, pp. 273310.
WHO, 2005. Ambient Air Quality Guidelines Prepared by. World Health Organization.
WHO, 2006. Air Quality Guidelines for Particulate Matter, Ozone, Nitrogen Dioxide and
Sulfur Dioxide: Global Update 2005: Summary of Risk Assessment Prepared by,
World Health Organization (Available http://www.euro.who.int/__data/assets/pdf_
le/0005/78638/E90038.pdf).
WHO, 2009. Global Health Risks: Mortality and Burden of Disease Attributable to Selected
Major Risks Prepared by. World Health Organization.
Wong, T.W., Lau, T.S., Yu, T.S., Neller, A., Wong, S.L., Tam, W., Pang, S.W., 1999. Air pollu-
tion and hospital admissions for respiratory and cardiovascular diseases in Hong
Kong. Occup. Environ. Med. 56, 679683.
157
Wong, T.W., Tam, W.S., Yu, T.S., Wong, A.H.S., 2002. Associations between daily mortal-
ities from respiratory and cardiovascular diseases and air pollution in Hong Kong,
China. Occup. Environ. Med. 59, 3035.
Woodcock, J., Edwards, P., Tonne, C., Armstrong, B.G., Ashiru, O., Banister, D., Roberts, I.,
2009. Public health benets of strategies to reduce greenhouse-gas emissions:
urban land transport. Lancet 374, 19301943.
World Bank, 2013. Diagnostic Assessment of Select Environmental Challenges in India
Prepared by, World Bank (Available http://documents.worldbank.org/curated/en/
docsearch?query=P101805, 2013).
Yagi, S., Nagayama, K., 2010. Opinions of intermodal transfer functions of urban railway
systems: a case study of Delhi Metro. Trans Res Board 89th Annual Meeting;
Washington, pp. 103937.
Zhang, L.W., Chen, X., Xue, X.D., Sun, M., Han, B., Li, C.P., Tang, N.J., 2014. Long-term
exposure to high particulate matter pollution and cardiovascular mortality: a
12-year cohort study in four cities in northern China. Environ. Int. 62, 4147.
Zou, B., Wilson, J.G., Zhan, F.B., Zeng, Y., 2009. Air pollution exposure assessment methods
utilized in epidemiological studies. J. Environ. Monit. 11, 475490.