CHAPTER 1

INTRODUCTION
This section provides you with a brief scope and overview of all the inclusions in the
report of the project. The major purpose of this documentation has also been described
along with goals and visions
Air quality index (AQI) is a numerical scale used for reporting day to day air quality
with regard to human health and the environment. The daily results of the index are
used to convey to the public an estimate of air pollution level. An increase in air
quality index signifies increased air pollution and severe threats to human health. In
most cases, AQI indicates how clear or polluted the air in our surrounding is, and the
associated health risks it might present. The AQI centers on the health effects that
may be experienced within a few days or hours after breathing polluted air.

AQI calculations focus on major air pollutants including: particulate matter, ground-
level ozone, sulfur dioxide (SO2), nitrogen dioxide (NO2), and carbon monoxide
(CO). Particulate matter and ozone pollutants pose the highest risks to human health
and the environment. For each of these air pollutant categories, different countries
have their own established air quality indices in relation to other nationally set air
quality standards for public health protection.

1.1 Origin and Concepts of Air Quality Index

In addition to land and water, air is the prime resource for sustenance of life. With the
technological advancements, a vast amount of data on ambient air quality is generated
and used to establish the quality of air in different areas. The large monitoring data
result is in encyclopaedic volumes of information that neither gives a clear picture to a
decision maker nor to a common man who simply wants to know how good or bad the
air is? One way to describe air quality is to report of daily levels of urban air
pollution is important to those who suffer from illnesses caused by exposure to air
pollution, the issue of air quality communication should be addressed in an effective
manner. Further, the success of a nation to improve air quality depends on the support
of its citizens who are wellinformed about local and national air pollution problems
and about the progress of mitigation efforts.

To address the above concerns, the concept of an Air Quality Index (AQI) has been
developed and used effectively in many developed countries for over last three
decades .An AQI is defined as an overall scheme that transforms weighted values of
individual air pollution related parameters (SO2, CO, visibility, etc.) into a single
number or set of numbers. There have not been significant efforts to develop and use
AQI in India, primarily due to the fact that a modest air quality monitoring
programme was started only in 1984 and public awareness about air pollution was
almost non existent. The challenge of communicating with the people in a
comprehensible manner has two dimensions:
(i) Translate the complex scientific and medical information into simple and precise
knowledge

1
(ii) Communicate with the citizens in the historical, current and futuristic sense.
Addressing these challenges and thus developing an efficient and comprehensible
AQI scale is required for citizens and policy makers to make decisions to prevent and
minimize air pollution exposure and ailments induced from the exposure.

1.2 Applications of Air Quality Index

Ott (1978) has listed the following six objectives that are served by an AQI:
 Resource Allocation: To assist administrators in allocating funds and determining
priorities. Enable evaluation of trade-offs involved in alternative air pollution control
strategies.

 Ranking of Locations: To assist in comparing air quality conditions at different

locations/cities. Thus, pointing out areas and frequencies of potential hazards.

 Enforcement of Standards: To determine extent to which the legislative standards

and existing criteria are being adhered. Also helps in identifying faulty standards and
inadequate monitoring programs.

 Trend Analysis: To determine change in air quality (degradation or improvement)

which have occurred over a specified period. This enables forecasting of air quality
(i.e., tracking the behaviour of pollutants in air) and plan pollution control measures.

 Public Information: To inform the public about environmental conditions (state of

environment). It’s useful for people who suffer from illness aggravated or caused by
air pollution. Thus it enables them to modify their daily activities at times when they
are informed of high pollution levels.

 Scientific Research: As a means for reducing a large set of data to a comprehendible

form that gives better insight to the researcher while conducting a study of some
environmental phenomena. This enables more objective determination of the
contribution of individual pollutants and sources to overall air quality. Such tools
become more useful when used in conjunction with other sources such as local
emission surveys.

1.3 Project Overview

In the past, AQI has been based on maximum sub-index approach using five
parameters i.e. suspended particulate matter (SPM), SO2 CO, PM10, and NO2
(Sharma 2001). However, the calculated AQI was always dominated by sub-index of
SPM due to lack of data availability for other pollutants. Recently, Indian Institute of
Tropical Meteorology (IITM), Pune has evolved an AQI, which provides sub-index
for PM10, PM2.5, O3, NO2, and CO (Beig et al, 2010), and has applied to continuous
air quality monitoring network. The IITM-AQI describes air quality in terms of very
unhealthy, very poor, poor (unhealthy for sensitive groups), moderate and good.
The revised CPCB air quality standards necessitate that the concept of AQI in India is
examined afresh. The revised National Ambient Air Quality Standards (CPCB 2009)
are notified for 12 parameters – PM10, PM2.5, NO2, SO2, CO, O3, NH3, Pb, Ni, As,

2
Benzo(a)pyrene, and Benzene. Although AQI is usually based on criteria pollutants
(i.e. PM10, PM2.5, SO2, NO2, CO and O3), a new approach to AQI which considers
as many pollutants from the list of notified pollutants as possible is desirable.
However, the selection of parameters primarily depends on AQI objective,data
availability, averaging period, monitoring frequency, and measurement methods.
While PM10, PM2.5, NO2, SO2 and Pb have 24-hourly as well annual average
standards, Ni, As, benzo(a)pyrene, and benzene have only annual standards and CO
and O3 have short-term standards (01 and 08 hourly average). PM10, PM2.5, SO2,
NO2, CO, and O3 are measured on a continuous basis at many air quality stations Pb,
Ni, As, Benzo(a)pyrene, if monitored, use manual systems. To get an updated AQI at
short time intervals, ideally eight parameters (PM10, PM2.5, NO2, SO2, CO, O3 and
Pb) for which, short-term standards are prescribed should, be measured on a
continuous basis.
It is seen that multiple agencies propose AQI schemes which may provide varying air
quality assessments, e.g. air quality may be termed as ‘good’ by one scheme and
‘poor’ by the other; this may be very confusing to general public. There is a need to
devise a uniform and efficient AQI scheme which provides information about every
pollutant and generates an overall index and be unique for the entire country.
In view of the above background, Central Pollution Control Board (CPCB) has
initiated this project on National Air Quality Index to strengthen air quality
information dissemination system for larger public awareness and their participation
on air quality management. An expert group was constituted with members drawn
from academia, medical fraternity, research institutes, MoEF&CC, advocacy groups,
SPCBs and CPCB. The group was mandated to deliberate, discuss and devise
consensus on the AQI system that is appropriate for Indian conditions. The technical
study was assigned to IIT Kanpur on grant-in-aid basis.
1.4 Project Objective

The project aims to achieve the following:

(i) Inform public regarding overall status of air quality through a summation
parameter that is easy to understand.
(ii) Inform citizens about associated health impacts of air pollution exposure in Delhi.
The overall objective of the project can be stated as under:
“To adopt/develop an Air Quality Index (AQI) based on national air quality
standards, health impacts and monitoring programme which represents perceivable air
quality for general public in easy to understand terms and assist in data interpretation
and decision making processes related to pollution mitigation measures.”
1.5 Scope Of Work

The scope of the work is summarized below:

(i) Review of available AQIs including international practices.
(ii) Suggest health impact threshold limits for eight parameters for which short-term
air quality standards are prescribed.

3
(iii) Develop a uniform AQI considering objectives, health impacts, air quality
standards, existing and future monitoring scenario including parameters, method and
frequency of measurements, and other relevant aspects.
(iv) Suggest qualitative description of air quality and associated likely health impacts
for different AQI values.
(v) Evaluate proposed AQI with data .
(vi) Develop web-based system for dissemination of AQI to public using current and
historical air quality database.
(vii) Finalize AQI and dissemination system in consultation with leading air quality
experts, medical professionals working in the field of air pollution health impacts,
State Pollution Control Boards and other stakeholders.

4
 CHAPTER 2

2. LITERATURE REVIEW

2.1 Air Quality Index: A Review

An air quality index is defined as an overall scheme that transforms the weighed
values of individual air pollution related parameters (for example, pollutant
concentrations) into a single number or set of numbers (Ott, 1978). The result is a set
of rules (i.e. most set of equations) that translates parameter values into a more simple
form by means of numerical manipulation.

FIG 1-Formation of an Aggregated Air Quality Index

If actual concentrations are reported in μg/m3 or ppm (parts per million) along with
standards, then it cannot be considered as an index. At the very last step, an index in
any system is to group specific concentration ranges into air quality descriptor
categories.
2.2 Structure Of an Index

Primarily two steps are involved in formulating an AQI:

(i) formation of sub-indices (for each pollutant) and
(ii) aggregation of sub-indices to get an overall AQI.
Formation of sub-indices (I1, I2,...., In) for n pollutant variables (X1, X2...., Xn) is
carried out using sub-index functions that are based on air quality standards and
health effects. Mathematically;
[1] Ii=f (Xi), i=1, 2,...,n
Each sub-index represents a relationship between pollutant concentrations and health
effect. The functional relationship between sub-index value (Ii) and pollutant
concentrations (Xi) is explained later in the text.
Aggregation of sub-indices, Ii is carried out with some mathematical function
(described below) to obtain the overall index (I), referred to as AQI.
5
 [2] I=F (I1,I2,....,In)
The aggregation function usually is a summation or multiplication operation or simply
a maximum operator.
2.3 Sub-Indices(Step 1)

Sub-index function represents the relationship between pollutant concentration Xi and

corresponding subindex Ii. It is an attempt to reflect environmental consequences as
the concentration of specific pollutant changes. It may take a variety of forms such as
linear, non-linear and segmented linear. Typically, the I-X relationship is represented
as follows:
[3] I = αX + β
Where, α =slope of the line, β = intercept at X=0.
The general equation for the sub-index (Ii) for a given pollutant concentration (Cp); as
based on ‘linear segmented principle’ is calculated as:
[4] Ii = [{(IHI - ILO)/(BHI -BLO)} * (Cp-BLO)]+ ILO
where,
BHI= Breakpoint concentration greater or equal to given concentration.
BLO= Breakpoint concentration smaller or equal to given concentration.
IHI =AQI value corresponding to BHI
ILO = AQI value corresponding to BLO
Ip = Pollutant concentration
2.3.1 Aggregation of Sub-indices(Step 2)

Once the sub-indices are formed, they are combined or aggregated in a simple
additive form or weighted additive form:
Weighted Additive Form
[5] I = Aggregated Index = ∑wiIi (For i= 1, …..,n)
where, ∑wi = 1
Ii= sub-index for pollutant i
n = number of pollutant variables
wi = weightage of the pollutant
Root-Sum-Power Form (non-linear aggregation form)
[6] I = Aggregated Index = [∑Iip](1/p)
where, p is the positive real number >1.
Root-Mean-Square Form

6
 [7] I = Aggregated Index = {1/k (I12 + I22 + …… +
Ik2}^0.5
Min or Max Operator (Ott 1978)
[8] I = Min or Max (I1, I2, I3, ..., In)
2.4 Indices in the Literature

i) Green Index(GI)

One of the earliest air pollution indices to appear in literature was proposed by Green
(1966). It included just two-pollutant variables - SO2 and COH (Coefficient of Haze).
The equations to calculate the subindices were:
I SO2 = 84 *X0.431
I COH = 26.6 *X0.576
Where, ISO2 = Sulphur dioxide sub-index
ICOH= Coefficient of Haze Sub-index
X = Observed pollutant concentration
The Green Index is computed as the arithmetic mean of the two sub-indices:
GI = 0.5 * (ISO2 + ICOH)
The above equations are obtained from the break point concentration.
Table 2.1 Break Point Concentration of Green Index

As the index did not include any other pollutants besides SO2 and SPM, it had limited
applications. It is applicable in colder seasons only. It is also subjected to eclipsing
and ambiguity phenomena (arithmetic mean weighted as linear sum). This index was
intended more as a system for triggering control actions during air pollution episodes
than a means for reporting air quality data to the public.
ii) Fenstock Air Quality Index(AQI)

Fenstock (1969) proposed an index to assess the relative severity of air pollution and
applied it to assess AQI of 29 U.S cities. This was the first index to estimate air
pollutant concentrations from the data on source emissions and meteorological
conditions in each city:
AQI = Wi Ii

7
where, Wi = weightages for CO, TSP and SO2
Ii= estimated sub-indices for CO, TSP and SO2
This index is applicable to square urban area with wind always parallel to one side for
uniform meteorological conditions under neutral stability with continuous source
distributed uniformly. This AQI is not used for daily air quality reports but for
estimating overall air pollution potential for a metropolitan area.
iii) Ontario API

Shenfeld (1970) developed Ontario Air Pollution Index in Canada. This index was
intended to provide the public with daily information about air quality levels and to
trigger control actions during air pollution episodes. It includes two pollutants
variables:
API = 0.2 (30.5 COH + 126 SO2) 1.35
Both COH and SO2 (in ppm) are 24 hour running averages.

iv) Oak Ridge Air Quality Index

Oak Ridge National Laboratory published the ORAQI in 1971. It was based on the
24-hour average concentrations of the following five pollutants:
1. SO2
2. NO2
3. PM
4. CO
5. Photochemical Oxidants

8
The sub-index is calculated as the ratio of the observed pollutant concentration to its
respective standard. As reported by Babcock and Nagda (1972), the ORAQI
aggregation function was a non-linear function:
ORAQI = {5.7 ∑ Ii}1.37
where, Ii= (X/Xs)i
X = Observed pollutant concentration
Xs = Pollutant Standard
I = Pollutant
The standards for the pollutants used in developing ORAQI

The constants are so selected that the ORAQI = 10 when all concentrations are at
their naturally occurring or backgrounds levels and ORAQI = 100 when all
concentrations are at their standards.
Although well-defined descriptors are given, its developers imply no correlation with
health effects. It is subjected to eclipsing and ambiguity. It is also difficult to explain
to public and involves complex calculations.
v) Greater Vancouver Air Quality Index

The GVAQI is based on Canadian Federal Government air quality objectives that are
designed to protect public health and environment. The index includes the following
pollutants:
1. SO2

9
2. NO2
3. O3
4. TSP
5. COH
6. PM10
GVAQI values are divided into ranges. The federal Desirable, Acceptable and
Tolerable air quality objectives levels are assigned GVAQI values of 25, 50 and 100
respectively. Intermediate values can be obtained by extrapolation. Each range is
associated with descriptor categories. The break point concentrations used to find
GVAQI

The overall GVAQI value is determined by calculating a sub-index for each pollutant
measurement and averaging time. Each sub-index is calculated by straight-line
extrapolation of the break point concentrations corresponding to GVAQI values of 25,
50 and 100 respectively. The maximum sub-index is reported as the GVAQI, based
on the assumption that the combined effect of a number of air pollutants is related to
the highest concentrations relative to air quality objectives. The particular pollutant
responsible for the maximum Sub-Index is called the “Index pollutant”. It is reported
with the GVAQI when the index value is greater than 25. Each GVAQI range is

10
associated with descriptor categories, general health effects and cautionary
statements.

vi) Most Undesirable Respirable Contaminants Index

MURC was published in 1968 (taken from Ott, 1978). This was routinely used in the
city of Detroit to report air quality data to the public and was broadcast between 8:30
A.M. and 9.00 A.M. each day on local radio stations. MURC is based on just one
pollutant variable, coefficient of Haze (COH)
MURC = 70X0.7
where, X= COH units
This equation is obtained such that COH values ranging from 0.3 – 2.15 give MURC
values ranging from 30 – 120 approximately. Five different descriptors are reported
for varying ranges of the MURC index.

The function was so chosen to reflect a good average approximation of the actual
weight of SPM in the atmosphere as measured by high volume sampler. However, for
MURC values higher than 120, the correlation with SPM concentration does not hold.
2.5 Current status of AQI Application in India

There have not been significant efforts to develop and use AQI in India, primarily due
to the fact that the National Air Quality Monitoring Programme has started only in
1984. Although NEERI, Nagpur started monitoring programme in 10 cities in 1978
and Bombay Municipal Corporation even before 1978, attempts were not made to use
AQI for data interpretation and public broadcasting. Agharkar (1982) reviewed
available AQIs and compared Air Quality status of the city of Bombay with its
suburbs. Although many technical papers proposing specific indices appeared in
international literature, no detailed study was undertaken to use an index in India.

11
A recent study to define Air Quality Index in India has been taken up by Beig et al
(2010) which includes air quality forecasting and named the system as SAFAR
(System of Air Quality-Weather Forecasting and Research). This study considered
correlation analysis of long term air quality data of different pollutants and health data
for two cities, Chennai and Delhi. The shortcoming of this study was that it accounted
health data only for two cities whereas for an ideal AQI representative of a country,
one needs to account health data for as many cities and towns as possible.
2.6 Eclipsing and Ambiguity

Two important characteristics, eclipsing and ambiguity are common to many indices
and are significant to interpret any index in the right perspective. This could be best
illustrated by a simple aggregation of two indices as in situation presented below:
Example: Let I= I1 + I2 and if I1> 100, I2> 100 indicate that the concentration of
each pollutant is greater than the ‘standard’. Question arises whether ‘I’ combined in
this manner reflect properly the meaning implied in each index? It is possible to have
combinations of I1 and I2 such that I = 100, yet I1<100 and I2<100. each pollutant
being within the prescribed standards but for e.g. if I1 = 70, I2 = 70; I =140. This
gives an impression that combined Index, I > 100, i.e. pollution standards are
violated, when they are actually not. Such a situation is called as ‘ambiguous region’.
In this region, Index I exaggerates pollution status i.e. Over-estimation of pollution
level. In case of more than 2 sub-indices I will be greater than 100, if each sub-index
is slightly more than 100/n without violating standards.
Now, let I = 0.5(I1 + I2). Effect of this on I=100, is to move the line parallel to itself
without changing its slope . If I2 = 60 and I1 = 120; I = 90. Hence, though the
standards are violated for I2 the combined index underestimates the pollution. This is
known as “Eclipsing” .These two characteristics of index (Ambiguity and Eclipsing)
are serious problems of additive and multiplicative indices. There is a significant
difference between air quality perceived by index and actual air quality. Therefore,
new indices which have been proposed are not of additive or multiplicative type; but
based on Maximum operator approach as it removes Ambiguity and Eclipsing.

FIG 2 Ambiguity characteristic of Indices

12
 FIG 3 Eclipsing characteristic of Indices

For the proposed AQI, a maximum operator system is selected:

AQI = Max(I1, I2, I3, ..., In)
There are two reasons for adopting a maximum operator:
• Free from eclipsing and ambiguity (Ott 1978)
• Health effects of combination of pollutants (synergistic effects) are not known and
thus a healthbased index cannot be combined or weighted
Maximum operator approach has been adopted by Sharma et al. (2001, 2002 and
2003) for development of an AQI scale for IIT-Kanpur and for entire India
respectively.

13
 CHAPTER 3

3. DEVELOPMENT AND DISSEMINATION OF AQI

3.1 Indian Air Quality Index (IND-AQI): Proposed System

Air quality standards are the basic foundation that provides a legal framework for air
pollution control.
An air quality standard is a description of a level of air quality that is adopted by a
regulatory authority as enforceable. The basis of development of standards is to
provide a rational for protecting public health from adverse effects of air pollutants, to
eliminate or reduce exposure to hazardous air pollutants, and to guide national/ local
authorities for pollution control decisions. With these objectives, CPCB notified
(http:// www.cpcb.nic.in) a new set of Indian National Air Quality Standards
(INAQS) for 12 parameters [carbon monoxide (CO) nitrogen dioxide (NO2), sulphur
dioxide (SO2), particulate matter (PM) of less than 2.5 microns size (PM2.5), PM of
less than 10 microns size (PM10), Ozone (O3), Lead (Pb), Benzo(a)Pyrene (BaP),
Benzene (C6H6), Arsenic (As), and Nickel (Ni)] . The first eight parameters have
short-term (1/8/24 hrs) and annual standards (except for CO and O3) and rest four
parameters have only annual standards.
Pollutant So2 NO2 PM2.5 PM10 03 CO Pb
Average 8 8 8 8 8 8 8 8 8
Time(hr)
Standard 80 80 60 100 180 100 4 2 1
Indian air quality standard

As stated in Chapter 1, an AQI scheme transforms weighted values of individual air

pollutant concentrations into a single number or set of numbers. It is important that an
AQI system should initially build on the sacrosanct AQS and then embark on
pollutant dose-response relationships to describe air quality in simple terms which
clearly relate to health impacts. While complexity in building the AQI is inevitable,
simplicity in AQI description is essential for general public to understand the air
pollution, possibly take actions to protect themselves and for policy makers to take
quick and broad decisions to improve air quality.
The objective of an AQI is to quickly disseminate air quality information (almost in
real-time) that entails the system to account for pollutants which have short-term
impacts. It is equally important that most of these pollutants are measured
continuously through an online monitoring network. Consequently, in the proposed
AQI system, the following pollutants are considered CO, NO2, SO2, PM2.5, PM10,
O3 and Pb. It is recognized that air concentrations of Pb are not known in real-time
and cannot contribute to real-time AQI but its consideration in AQI calculation of past
days will help in scrutinizing the status of this important toxic.
It is described in Chapter 2 that the aggregation function, F of combining sub-indices
of individual pollutants is delicate as most indices suffer from ambiguity and

14
eclipsing. For the proposed AQI, a maximum operator system has been adopted which
is free from ambiguity and eclipsing, as shown below:
AQI=Max (I1,I2,I3,...,In)
The operational scheme of AQI system based of maximum operator (i.e. maximum
subindex being the overall index). To present status of the air quality and its effects on
human health, the following description categories have been adopted for IND-AQI

These categories/AQI ranges should map to key references (breakpoints) of

concentration of each pollutants through a segmented linear or a nonlinear function.

15
 FIG 4 Overall AQI system

3.2 Air Quality Monitoring and AQI Considerations

The air quality monitoring network in India can be classified as (i) online and (ii)
manual. The pollutant parameters, frequency of measurement and monitoring
methodologies for two networks are very different. The AQI system for these
networks could be at variance, especially for reporting and completeness in terms of
parameters.
(i) Online Monitoring network: These are automated air quality monitoring stations
which record continuous hourly, monthly or annually averaged data. In India, ~ 40
automatic monitoring stations are operated (continuous stations in Delhi), where
parameters like PM10, PM2.5, NO2, SO2, CO, O3, etc. are monitored continuously.
Data from these stations are available almost in real-time. Thus such networks are
most suitable for computation of AQI sub-indices, as information on AQI can be
generated in real time. For AQI to be more useful and effective, there is a need to set
up more online monitoring stations for continuous and easy availability of air quality
data for computation of AQI for more Indian cities.

16
 FIG 5 Online monitoring station (ITO, New Delhi)

(ii) Manual: The manual stations involve mostly intermittent air quality data
collection, thus such stations are not suitable for AQI calculation particularly for its
quick dissemination. In India, air quality is being monitored manually at 573 locations
under National Air Monitoring Programme (NAMP). In most of these manually
operated stations, only three criteria pollutants viz. PM10, sulphur dioxide (SO2) and
nitrogen dioxide (NO2) are measured, at some stations PM2.5 and Pb are also
measured. The monitoring frequency is twice a week. Such manual networks are not
suitable for computing AQI, as availability of monitored data could have a lag of 1-3
days and sometimes not available at all. However, some efforts are required to use the
information in some productive manner. Historical AQIs on weekly basis can be
calculated and used for data interpretation and ranking of cities or towns for further
prioritization of actions on air pollution control.
3.3 Computation and Basis of Sub-index Breakpoints

Segmented linear functions are used for relating actual air pollution concentration
(Xi) (of each pollutant) to a normalized number referred to as sub-index (Ii). While
AQI system is not complex in understanding, to arrive at breakpoints which will relate
to AQI description is of paramount significance. Consequences of inappropriate
adoption of breakpoints could be far reaching; it may lead to incorrect information to
general public (on health effects) and decisions taken for pollution control may be
incorrect.
The basis for linear functions (for this study) to relate air quality levels to AQI
requires careful consideration. Services of practicing doctors and experts in this field
(see Appendix 1) have proved very useful. In this study, in addition to dose response
relationship, the breakpoints adopted by other countries/agencies (USEPA 2014; U.K.
2013; Malaysia 2013; GVAQI 2013; Ontario 2013) have been examined for using
these in INDAQI.

17
It is important that an AQI system should build on AQS and pollutant dose-response
relationships to describe air quality in simple terms which clearly relates to health
impacts. The first step for arriving at breakpoints for each pollutant is to consider
attainment of INAQS (Table 3.1). The index category is classified as ‘good’ for
concentration range up to half of INAQS (for example, for SO2 AQI=0-50 for
concentration range of 0-40 μg/m3) and as ‘satisfactory’ up to attainment of INAQS
(i.e. SO2 range 41-80μg/m3 linearly maps to AQI=51-100). To arrive at breakpoints
for other categories (for each pollutant), we require a thorough research/review of
dose response relationships, which is described here.
3.3.1 Carbon Mono-oxide (CO)

Carbon monoxide (CO) is an important criteria pollutant which is ubiquitous in urban

environment. CO production mostly occurs from sources having incomplete
combustion. Due to its toxicity and appreciable mass in atmosphere, it should be
considered as an important pollutant in AQI scheme.
CO rapidly diffuses across alveolar, capillary and placental membranes.
Approximately 80-90% of absorbed CO binds with Hb to from Carboxyhaemoglobin
(COHb), which is a specific biomarker of exposure in blood. The affinity of Hb for
CO is 200-250 times than that of oxygen. In patients with hemolytic anemia, the CO
production rate was 2–8 times higher and blood COHb concentration was 2–3 times
higher than in normal person (WHO 2000). The initial symptoms of CO poisoning
may include headache, dizziness, drowsiness, and nausea. These initial symptoms
may advance to vomiting, loss of consciousness, and collapse if prolonged or high
exposures are encountered and may lead to Coma or death if high exposures continue.
A US study estimated that 6 per cent of the congestive heart failures and
hospitalizations in the cities were related to an increase in CO concentration in
ambient atmosphere (WHO 2000). Reduction in the ability of blood to transport
oxygen leads to tissue hypoxia. The body compensates for this stress by increasing
cardiac output and the blood flow to specific areas, such as the heart and brain. As the
level of COHb in the blood increases, the person suffers from effects which become
progressively more serious. CO has both 1 hr and 8 hr standard. Figure 3.3 shows air
pollution level and percent of COHb. The symptoms associated with various percent
blood saturation levels of COHb.

18
FIG 6 CO Concentration and COHb level in Blood

FIG 7 Symptoms Based on COHb Level

19
 Table 3.3 Breakpoint for CO

3.3.2 Nitrogen Dioxide (NO2)

The major source of NO2 is combustion processes. An appreciable quantity of NO2 is

present in rural and urban environments. Further, NO2 is showing alarmingly high
increasing trend in Indian cities due to increase in number of vehicles. On inhalation,
70–90% of NO2 can be absorbed in the respiratory tract of humans, and physical
exercise increases the total percentage absorbed (Miller et al., 1982). NO2 exposure
can cause decrement in lung function (i.e. increased airway resistance), increased
airway responsiveness to broncho-constrictions in healthy subjects at concentration
exceeding 1 ppm (WHO 2000). Below 1 ppm level, there are evidences of change in
lung volume, flow volume, characteristics of lung or airway resistance in healthy
persons. It has been established that continuous exposure with as little as 0.1 ppm
NO2 over a period of one to three years, increases incidence of bronchitis,
emphysema and have adverse effect on lung performance (WHO 2000). Exposure to
excessive NO2, affects the defence mechanism leaving the host susceptible to
respiratory illness.
Chronic exposure of NO2 may lead to chronic lung disease and variety of
structural/morphological changes in lung epithelium conducting airways and air -gas
exchange region. Exposure to high levels (>1.0 ppm) of NO2 causes Eustachian of
bronchiolar and alveolar epithelium, inflammation of epithelium and definite
emphysema (WHO 2000).
Normal healthy people exposed at rest or with light exercise for less than 2 hours to
concentrations of more than 4700μg/m3 (2.5ppm) experience pronounced decrements
in pulmonary function; generally, such people are not affected at less than 1880μg/m3
(1ppm). One study showed that the lung function of people with chronic obstructive
pulmonary disease is slightly affected by a 3.75-hour exposure to 560μg/ m3
(0.3ppm). A wide range of findings in asthmatics has been reported; one study
observed no effects from a 75-minute exposure to 7520μg/m3 (4ppm), whereas others

20
showed decreases in FEV1 (forced expiration volume in one second) after 10 minutes
of exercise during exposure to 560μg/m3 (0.3ppm).
For acute exposures, only very high concentrations (1990 μg/m3; > 1000 ppb) affect
healthy people. Asthmatics and patients with chronic obstructive pulmonary disease
are clearly more susceptible to acute changes in lung function, airway responsiveness
and respiratory symptoms. Given the small changes in lung function (< 5% drop in
FEV1 between air and nitrogen dioxide exposure) and changes in airway
responsiveness reported in several studies, 375–565 μg/m3 (0.20 to 0.30 ppm) is a
clear lowest-observedeffect level. A 50% margin of safety is proposed because of the
reported statistically significant increase in response to a bronchoconstrictor
(increased airway responsiveness) with exposure to 190 μg/m3 and a metaanalysis
suggesting changes in airway responsiveness below 365 μg/m3 (WHO 2000).
After giving due consideration to INAQS for NO2, two categories good (Sub-Index:
0-50) and satisfactory (51-100), the breakpoint concentration are fixed as 40μg/m3
and 80μg/m3. Various studies reported that the small change in lung function (< 5%
drop in FEV1 between air and nitrogen dioxide exposure) and changes in airway
responsiveness gives 375–565μg/m3 (0.20 to 0.30 ppm), as the lowest-observed-
effect level. Therefore, breakpoints of 280μg/m3 for poor, 400 μg/m3for very poor
and 400+ μg/m3 for severe category are adopted. For moderately-polluted category an
intermediate value of 180 μg/m3(between 80 and 280 μg/m3) has been adopted. It
may be noted that minor tweaking has been done with breakpoints so that these also
corroborate with international breakpoints adopted by other countries. The details of
proposed break points for IND-AQI and breakpoints of USEPA, China and EU.

Table 3.4 Breakpoint for NO2

21
3.3.3 Particulate Matter (PM): PM10 and PM2.5

PM levels in Indian cities are about 4-5 times higher than in the US cities (WRI,
1996). These high PM levels may have severe impact on public health. The sixteen-
year long survey by Dockery et al. (1994) has revealed that there is a strong
correlation between ambient PM concentrations and increase in mortality and
hospitalizations due to respiratory diseases. Several epidemiological studies (Pope,
1989; Schwartz, 1996) have linked PM10 (aerodynamic diameter ≤ 10 μm) and
PM2.5 with significant health problems, including: premature mortality, chronic
respiratory disease, emergency visits and hospital admissions, aggravated asthma,
acute respiratory symptoms, and decrease in lung function. PM2.5 is of specific
concern because it contains a high proportion of various toxic metals and acids, and
aerodynamically it can penetrate deeper into the respiratory tract.
A HEI study, (Wichmannet al., 2000) reported that the concentration of both ultrafine
(PM<0.1) and fine particles (PM0.1-2.5) was associated with increased daily
mortality. Lippmann et al. (2000) reported that four of five size fractions (PM40
PM10-40 PM10 PM2.5-10 PM2.5) were associated with increased in morbidity and
mortality. The largest particle size fraction (10 μm – 40 μm) was not associated with
increased morbidity and mortality. However, Castillejos et al (2000) in Mexico City
and Ostro et al. (2000) in western United States have found health effects being
associated with the coarse fraction as well but studies (Schwartz et al., 1996)
conducted in other parts of the United States and in Canada have reported that effects
of fine particles are predominant.
Major concerns for human health from exposure to PM10 include effects on
breathing, respiratory symptoms, decrease in pulmonary function and damage to lung
tissue, cancer, and premature death. An association between elevated PM10 levels and
hospital admissions for pneumonia, bronchitis, and asthma was observed by Pope
(1989). Long-term particulate exposure was associated with an increase in risk of
respiratory illness in children (Dockery et al., 1989). Statistically significant
relationships were observed between TSP levels and forced vital capacity (FVC) and
FEV1 (Chestnut et al., 1991). Ostro (1993) has reported a series of studies that
observed associations between daily changes in particulate pollution and daily
mortality.
Prospective cohort studies by Pope et al. (1995) observed 30 to 50% increase in lung
cancer rates associated with exposure to respiratory particles. Associations between
mortality risk and air pollution were strongest for respiratory particles and sulfates
(Pope et al., 1995). PEFR (peak expiratory flowrate) and respiratory symptoms were
strongly associated with PM10 levels and marginally with ozone levels (Romieu et
al., 1996). Increase in PM concentration correlated with increase in mortality and
morbidity rates. An increase of 10μg/m3 of PM10 levels resulted in a 3-6 % increase
in visits for asthma people and a 1-3 % increase in visits for upper respiratory diseases
not with asthma to hospitals. The findings are consistent with the result of previous
studies of particulate pollution in other urban areas and provide evidence that the
coarse fraction of PM10 may affect the health of working people (Gordian et al.,
1996). A study in six US cities has shown that there is an association between fine
particulate matter (PM2.5) primarily from combustion sources and daily mortality
(Schwartz et al., 1996). Combustion particles in the fine fraction from mobile and coal

22
combustion sources, both not fine crustal particles, are associated with increase in
mortality (Laden et al., 2000).
Sharma et al. (2004) through a study in Kanpur reported that mean PEF (L/min)
values of a cohort (of over 100 subjects) decrease with the increase in PM10 and/or
PM2.5. The findings of the study can be summarized as under:
(i) The correlation (negative) between mean ΔPEF (i.e. deviation in PEF) of a day
(no. of days of sampling = 39) and four indicators of PM levels (PM10, PM2.5, PM10
(one-day lag) and PM2.5 (one-day lag)) was found to be statistically significant (p <
0.05). It showed that as the pollution level increases the lung function in terms of
PEFR reduces/deteriorates. The negative correlation with PM10 (one day lag) and
PM2.5 (one-day lag) also suggested that PM pollution may have sustained effect on
PEFR value due to pollution level of previous day.
(ii) PM10 and PM2.5 correlate with ΔPEF, PM10 and their concentration levels are
better indicator to reflect changes in PEFR values. This suggests that the deposition of
larger particles (PM10) takes place in upper part of respiratory system that activates
mucus secretion resulting is constriction of airways and thus lowering PEFR value.
The fine particles impact the pulmonary region (lower respiratory system), which are
known to cause long-term chronic effects.
(iii) FEV1, PEFR and FVC are the key lung function parameters that reflect health
impact of air pollution (Bates, 2002). The deviations found in FEV1 and FVC are: (a)
FEV1 -0.30 L (at VikasNagar (VN): PM10: 300μg/m3), -0.31 (at Juhi Colony (JC):
PM10: 300 μg/m3) and -0.18 L (IIT Kanpur (IITK): PM10: 185 μg/m3 IITK) and (b)
FVC -0.42 L (VN), -0.40 (JC) and -0.27 L (IITK).
It is evident from the above discussion that both PM10 and PM2.5 have specific
health impacts and both of these pollutants should be considered for AQI.
PM10

WHO (2005) suggests that there is no threshold for particulate concentration below
which there is no harmful effect. At the same time, high PM10 background
concentration in India cannot be disregarded which is reflected in relatively high level
of INAQS for PM10; Sharma (2009) has estimated background concentration of
PM10 as 35 μg/m3. For PM10, in view of no specific studies in India, it is proposed
that the breakpoints proposed by USEPA may be adopted after accounting for
INAQS.

23
 Table 3.5 Breakpoint for
PM10
PM2.5

Sharma (2009) has estimated background concentration of PM2.5 as 17-28 μg/m3.

The background concentration in Europe and the US is very low (< 5 μg/m3).
Therefore, for lower concentration range, it is not reasonable to follow the breakpoints
as proposed by US or EU. With due regard to INAQS (which accounts for
background pollution), the first two categories, Good and Satisfactory, the breakpoints
are kept as 30 and 60 μg/m3.
As per HEI Global Burden of disease report (2013), till 90μg/m3 the relative risk of
Ischemic Heart Disease increase and then more or less it plateaus off, therefore the
next break point for category moderately polluted is kept as 90 μg/m3. For PM2.5, in
view of no specific studies in India, it is proposed that the breakpoints proposed by
USEPA may be adopted. Beyond first three categories, the breakpoints proposed by
USEPA and China are adopted.

24
 Table 3.6 Breakpoint for PM2.5

3.3.4 Ozone

Ozone, a secondary pollutant formed in the atmosphere, has serious health impacts.
Ozone is a strong oxidant, and it can react with a wide range of cellular components
and biological materials. Ozone can aggravate bronchitis, heart disease, emphysema,
asthma and reduce lung capacity. Irritation can occur in respiratory system, causing
coughing, and uncomfortable sensation in chest (WHO, 2000). It can reduce lung
function and can make breathing difficult. Ozone makes people more sensitive to
allergens, which are the most common triggers for asthma attacks, thus it can
aggravate asthma, when ambient ozone levels are high. Also, asthmatics are more
severely affected by the reduced lung function and irritation in the respiratory system.
Ozone can inflame and damage lung cells. Within few days of ozone exposure the
damaged cells are replaced and the old cells shed (WHO 2000). Ozone may aggravate
chronic lung diseases such as emphysema and bronchitis and reduce the immune
system’s ability to fight off bacterial infections in the respiratory system.
For 1–3 hours of ozone exposure in healthy subjects during moderate-to-heavy
exercise (ventilation > 45 litres/minute), changes in pulmonary function have been
reported for the following tests (lowest-observedeffect levels under conditions of
strenuous exercise) (McDonnell et al., 1983 and Gong et al., 1986):
• Forced expiratory volume in 1 second (FEV1) (240 μg/m3)
• Airway resistance (360 μg/m3)
• Forced vital capacity (FVC) (240 μg/m3)
• Increased respiratory frequency (400 μg/m3).

25
For 4–8 hours of ozone exposure in healthy adults doing moderate exercise, the
following changes in pulmonary function tests have been reported (Horstman et al.,
1990) with given concentrations.
• FEV1, 160 μg/m3
• Airway resistance, 160 μg/m3
• FVC, 200 μg/m3
• Increased airway responsiveness, 160 μg/m3.

After giving due consideration to INAQS for ozone, for two categories - Good
(subindex 0-50) and Satisfactory (51-100), the breakpoint concentrations are fixed as
50 μg/m3and 100 μg/m3. It can be seen that 180, 250 and 320 μg/m3 (8-hour
concentration) cause important health endpoints leading to 2, 4 and 8 fold
inflammatory changes in population .
With these endpoints, the proposed breakpoints are: moderately polluted at 200 μg/m3
poor at 250 μg/m3and 1-hr concentration break points for very poor is taken as 750
and for severe it is taken as 750+ μg/m3 (this concentration will nearly match to 350
μg/m3of 8-hr average concentration). AQI breakpoints for various categories for
ozone along with breakpoints of other countries.

26
3.3.5 Sulphur Dioxide (SO2)

SO2 is soluble in aqueous media and affects mucous membranes of the nose and
upper respiratory tract. Reduction in mean lung function values among groups of
healthy individual have been observed for 10minute exposures at 4000 ppb (11 440
μg/m3) (Linn et al. 1984) and at 5000 ppb (14 300 μg/m3) (Lawther et al., 1975). No
significant changes in group mean lung function value have been seen below 1000
ppb (2860 μg/m3) even during exercise.
Asthmatic people appear to respond in a similar way to normal subjects, with
development of bronchoconstriction, but at lower concentrations. Several studies
(Linn et al., 1984) have shown fairly large changes in mean values of lung function
indices with 600 ppb (1716 μg/m3) and heavy exercise. Linn et al. (1984) examined
the dose–response relationship of change in mean FEV1 with increasing
concentrations of SO2 with exercise in patients with moderate or severe asthma.
Overall, the mean response at 400 ppb (1144 μg/m3) has been definite though small,
at around 300-ml fall in mean values and at 200 ppb (572 μg/ m3) changes were
negligible. Hence, from the information published hitherto, it can be concluded that
the minimum concentration evoking changes in lung function in exercising asthmatics
is of the order of 400 ppb (1144 μg/m3).
SO2 breakpoints :The first step is the attainment of INAQS (Table 3.1). The index
category for SO2 is classified as ‘good’ for concentration range 0-40μg/m3 (half of
27
INAQS for SO2) for AQI range 0-50 and as ‘satisfactory’ from 4180μg/m3 for AQI
range 51-100. For the third sub-index range 101–200, violations of USEPA standards
are examined. The INAQS for SO2 (80μg/m3) is more stringent than the USEPA
standard (377μg/m3, USEPA 2014). In other words, the built-in safety factor is higher
for the Indian standard. The USEPA standard (and discussions above) suggests that
for SO2 levels up to 365μg/m3, the air quality is acceptable from a public health point
of view. Thus, for SO2 levels between 81 and 365μg/m3, the corresponding sub-index
value has been taken to vary linearly between 101 and 200, and the AQI category for
SO2 is classified as ‘moderately polluted’. In absence of any other pollutant health
criteria in India the rest of the categorization of AQI is based on the USEPA federal
episode criteria and significant harm level (USEPA 1998) and studies of Lawther et
al., 1975) and Linn et al. (1983 and 1984). Table 3.9 shows proposed SO2
breakpoints.

Table 3.81 Breakpoint for SO2

3.3.6 AQI Breakpoints for Pb

It is to be noted that most of the countries have taken only six pollutants (described
above) for formulation of AQI. An attempt has been made to propose breakpoints for
Pb as the pollutants also have short-term standards of 24-hr. However, Pb levels can
be utilized in calculation of AQI of past days to assess impact of lead pollution.
Inhalation of high levels of Pb causes irritation to the nose, throat and respiratory
tract. Increased inhalation may result in cough and an increased respiratory rate as
well as respiratory distress. An association has been reported between exposure to
ammonia and cough, phlegm, wheezing, and asthma at high concentration.
28
Pb is a toxic metal and its exposure through all routes result in increased blood lead
level. At lower concentrations, the blood lead level is proportional to air concentration
(after accounting for all resulting exposure routes). For example, 1 μg/m3 of annual
lead level will result in 5μg/dL(on an average) of blood lead level (WHO 2000). The
effect of blood level above 10μg/dL is seen in haematological changes in sensitive
population, therefore, at moderate pollution level the break point is proposed at
2μg/m3. At 20μg/ dL blood lead level the effects become more prominent and this
corresponds to break point of 4 μg/m3 but to account for factor of safety, next break
point is kept at 3.0 μg/m3 (and not at 4 μg/m3) and if the lead concentration in air is
more than 3.5 μg/m3 the AQI category will be severe.
AQI Category PB ug/m3
Good(0-50) 0.5
Satisfactory(51-100) 1.0
Moderately Polluted(101-200) 2.0
Poor (201-300) 3.0
Very poor(301-400) 3.5
Severe(401-500) 3.5+
TABLE 3.9

It have presented basis of AQI breakpoints for eight pollutant parameters considered
for AQI and these are summarized with colour scheme to represent the AQI bands. It
shows health statements for every AQI category for people to understand health
effects and protect themselves from these effects

29
3.4 Interpretation of Air quality using IND-AQI: an example

An exampele of AQI calculation and description for Delhi (online air quality
monitoring network) and Kanpur (manual network) is presented here for two seasons,
monsoon and winter. The sub-index (Ip) for a given pollutant concentration (Cp), as
based on ‘linear segmented principle’ is calculated as:
I p= [{(IHI - ILO)/ (BHI -BLO)} * (Cp-BLO)] + ILO
BHI= Breakpoint concentration greater or equal to given conc.
BLO= Breakpoint concentration smaller or equal to given conc.
IHI = AQI value corresponding to BHI
ILO = AQI value corresponding to BLO
Finally; AQI = Max (Ip) (where; p= 1,2,...,n; denotes n pollutants)
AQI of Delhi

AQI has been calculated for July (clean period) and November (highly polluted
period) for monitoring stations Anand Vihar, Rk Puram ,Punjabi Bagh.

JULY AQI

The AQI for CO and O3 has been calculated for running 8-hr averages. This will give
23 AQI values, here maximum and minimum AQI of CO and O3 are presented. It can

30
be seen that for most pollutants air quality is good/satisfactory. It is PM10 which is in
moderately polluted category.

November AQI

The AQI for CO and O3 has been calculated for running 8-hr averages. This will give
23 AQI values; here maximum and min AQI of CO and O3 are presented. It can be
seen that for most pollutants air quality is good/satisfactory. It is PM10 and PM2.5
which suggest AQI to be in Severe category

31
From the above interpretaion of air Quality index for Delhi responsible parameter for
pollution is PM10 and PM2.5. In Monsoon the responsible parameter for pollution in
Anand Vihar and Panjabi Baag is PM10 with moderate pollution, R K Puram with
PM2.5 responsible parameter is satisfactory or moderate polluted. In winters Anand
Vihar and R K Puram has very severe PM10 index, whereas Panjabi Baag has very
severe PM2.5 index.

32
3.5 Web-based AQI Dissemination

The AQI system should have web-based AQI dissemination which should be
designed for online calculation and display of nation-wide AQI. The website should
render a quick, simple and an elegant looking response to an AQI query. The other
features of the website should include reporting of pollutant responsible for index,
pollutants exceeding the standards and health effects.
The first functionality of the website is taken as AQI query which is presented in
using three steps on the AQI website. It shows AQI of past 48 hours on time scale.
The last AQI is based on 24-hr running average (8-hr running average for CO and
O3).

33
 FIG 8 Web-based AQI Query: Reporting and Display

As a second part of the functionality, the website can also render menu-based AQI
query by searching through states and cities:

34
 FIG 9 Menu-based AQI Query and display

Technologies for Website

 Front-End / GUI
 HTML 5
 Java-Scripts & CSS
 Self-Developed java script Library for primary functionalities of the website.
 Google Maps Library for map - https://developers.google.com/maps/web/
 Google Charts Library for graph charts - https://developers.google.com/chart/
 Bootstrap for GUI - http://getbootstrap.com/
 just Gage Library for solid gauges - http://justgage.com/
 Bootstrap sortable for sorting the rankings -
https://github.com/drvic10k/bootstrap-sortable
 Bootstrap date picker for selecting date graphically - http://bootstrap-
datepicker.readthedocs. org/en/release/
 Middle-Ware
 Apache Server
 Play framework with Java
 Database – MySQL

35
 CHAPTER 4

4. TOOLS AND TECHNOLOGIES

We are using python programming language, data science and its different modules
i.e numpy,pandas,matplotlib,seaborn etc
4.1 PYTHON:
Python is an interpreted, high-level, general-purpose programming language. Created
by Guido van Rossum and first released in 1991, Python's design philosophy
emphasizes code readability with its notable use of significant whitespace. Its
language constructs and object-oriented approach aim to help programmers write
clear, logical code for small and large-scale projects
Python is dynamically typed and garbage-collected. It supports multiple programming
paradigms, including procedural, object-oriented, and functional programming.
Python is often described as a "batteries included" language due to its
comprehensive standard library.
Python was conceived in the late 1980s as a successor to the ABC language.
Python 2.0, released in 2000, introduced features like list comprehensions and
a garbage collection system capable of collecting reference cycles. Python 3.0,
released in 2008, was a major revision of the language that is not
completely backward-compatible, and much Python 2 code does not run unmodified
on Python 3.

36
The Python 2 language, i.e. Python 2.7.x, is "sunsetting" on January 1, 2020 (after
extension; first planned for 2015), and the Python team of volunteers will not fix
security issues, or improve it in other ways after that date. With the end-of-life, only
Python 3.5.x and later will be supported.
Python interpreters are available for many operating systems. A global community of
programmers develops and maintains CPython, an open source reference
implementation. A non-profit organization, the Python Software Foundation, manages
and directs resources for Python and CPython development.

4.2 DATA SCIENCE :Data science is a multi-disciplinary field that uses scientific
methods, processes, algorithms and systems to extract knowledge and insights from
structured and unstructured data. Data science is the same concept as data
mining and big data: "use the most powerful hardware, the most powerful
programming systems, and the most efficient algorithms to solve problems".
Data science is a "concept to unify statistics, data analysis, machine learning and their
related methods" in order to "understand and analyze actual phenomena" with data. It
employs techniques and theories drawn from many fields within the context
of mathematics, statistics, computer science, and information science. Turing
award winner Jim Gray imagined data science as a "fourth paradigm" of science
(empirical, theoretical, computational and now data-driven) and asserted that
"everything about science is changing because of the impact of information
technology" and the data deluge. In 2015, the American Statistical
Association identified database management, statistics and machine learning,
and distributed and parallel systems as the three emerging foundational professional
communities.
In 2012, when Harvard Business Review called it "The Sexiest Job of the 21st
Century", the term "data science" became a buzzword. It is now often used
interchangeably with earlier concepts like business analytics, business
intelligence, predictive modeling, and statistics.Even the suggestion that data science
is sexy was paraphrasing Hans Rosling, featured in a 2011 BBC documentary with the
quote, "Statistics is now the sexiest subject around." Nate Silver referred to data
science as a sexed up term for statistics. In many cases, earlier approaches and
solutions are now simply rebranded as "data science" to be more attractive, which can
cause the term to become "dilute[d] beyond usefulness." While many university
programs now offer a data science degree, there exists no consensus on a definition or
suitable curriculum contents. To the discredit of the discipline, however, many data-
science and big-data projects fail to deliver useful results, often as a result of poor
management and utilization of resources.
4.2.1 Numpy: Numpy(Numerical Python) is a linear algebra library in Python. It is a
very important library on which almost every data science or machine learning Python
packages such as SciPy (Scientific Python), Mat−plotlib (plotting library), Scikit-learn,
etc depends on to a reasonable extent.

NumPy is very useful for performing mathematical and logical operations on Arrays. It
provides an abundance of useful features for operations on n-arrays and matrices in
Python.

37
 This course covers basics things to know about NumPy as a beginner in Data science.
These includes how to create NumPy arrays, use broadcasting, access values, and
manipulate arrays. More importantly, you will learn NumPy’s benefit over Python
lists, which include: being more compact, faster access in reading and writing items,
being more convenient and more efficient.

In this, we will be using the Jupyter notebook as our editor.

Installing NumPy

If you have Anaconda, you can simply install NumPy from your terminal or command
prompt 810using: { conda install numpy}

If you do not have Anaconda on your computer, install NumPy from your terminal
using:

{pip install numpy}

4.2.2 PANDAS:

The pandas package is the most important tool at the disposal of Data Scientists and
Analysts working in Python today. The powerful machine learning and glamorous
visualization tools may get all the attention, but pandas is the backbone of most data
projects.
Pandas has so many uses that it might make sense to list the things it can't do instead
of what it can do.

This tool is essentially your data’s home. Through pandas, you get acquainted with
your data by cleaning, transforming, and analyzing it.

For example, say you want to explore a dataset stored in a CSV on your computer.
Pandas will extract the data from that CSV into a DataFrame — a table, basically —
then let you do things like:

 Calculate statistics and answer questions about the data, like

o What's the average, median, max, or min of each column?
o Does column A correlate with column B?
o What does the distribution of data in column C look like?
 Clean the data by doing things like removing missing values and filtering rows or
columns by some criteria
 Visualize the data with help from Matplotlib. Plot bars, lines, histograms, bubbles,
and more.
 Store the cleaned, transformed data back into a CSV, other file or database
Before you jump into the modeling or the complex visualizations you need to have a
good understanding of the nature of your dataset and pandas is the best avenue
through which to do that.

38
Install and import PANDAS

Pandas is an easy package to install. Open up your terminal program (for Mac users)
or command line (for PC users) and install it using either of the following commands:

{conda install pandas}

OR

{pip install pandas}

Alternatively, if you're currently viewing this article in a Jupyter notebook you can
run this cell:
{!pip install pandas}
The ! at the beginning runs cells as if they were in a terminal.
To import pandas we usually import it with a shorter name since it's used so much:
{ Import pandas as pd }
4.2.3 MATPLOTLIB:

Matplotlib is a 2-D plotting library that helps in visualizing figures. Matplotlib

emulates Matlab like graphs and visualizations. Matlab is not free, is difficult to scale
and as a programming language is tedious. So, matplotlib in Python is used as it is a
robust, free and easy library for data visualization.
Installing Matplotlib

Type !pip install matplotlib in the Jupyter Notebook or if it doesn’t work in cmd type conda
install -c conda-forge matplotlib . This should work in most cases.

Matplotlib comes with a wide variety of plots. Plots helps to understand trends,
patterns, and to make correlations. They’re typically instruments for reasoning about
quantitative information. Some of the sample plots are covered here.

1> BARPLOT

39
2>HISTOGRAM

Importing Matplotlib

Just as we use the np shorthand for NumPy and the pd shorthand for Pandas, we will
use some standard shorthands for Matplotlib imports:

import matplotlib as mpl

import matplotlib.pyplot as plt

The plt interface is what we will use most often, as we shall see throughout this
chapter.

4.3.4 SEABORN:

Seaborn is a Python data visualization library based on matplotlib. It provides a high-

level interface for drawing attractive and informative statistical graphics.

Keys Features

 Seaborn is a statistical plotting library

 It has beautiful default styles

 It also is designed to work very well with Pandas dataframe objects.

Installing and getting started

To install the latest release of seaborn, you can use pip:

40
  pip install seaborn

It’s also possible to install the released version using conda:

 conda install seaborn

41
 CHAPTER 5

5. DESIGN AND IMPLEMENTATION

5.1 FLOW CHART:

The aim of this study is to present a new model for AQI forecasting using
collaborative multiple city air quality data as input. The structured flowchart
representation of the proposed model is clearly shown below in Fig

In order to forecast the Air Quality Index (AQI) values for city A, we take into
account the air quality data of neighboring cities B and C along with city A. We take
3 cases—firstly we take air quality data of City A and City B as input; secondly we
take air quality data of City A and City C as input and finally we take air quality data
of all 3 cities as input in order to forecast the AQI values for city A. After this we split
the three data sets into training and testing data. Then we develop SVR machine
learning algorithm on all the three training data sets. Finally, Forecasting is carried

42
out on the testing dataset based on the developed model on the training set. Since the
number of input variables or features are increasing when we take into consideration
the information from more than one city, the training complexity increases that leads
to larger time to train the data. After the training the number of support vectors are
selected and at the time of testing the complexity is linear on the number of the
support vectors and linear on the number of features.
5.2 SOURCE CODE:

43
44
45
46
47
48
49
50
 CHAPTER 6

CONCLUSIONS AND SUMMARY

The revised air quality standards (CPCB, 2009) necessitate that the concept of AQI in
India is examined afresh. An AQI system based on maximum operator function
(selecting the maximum of sub-indices of various pollutants as overall AQI) is
adopted. Ideally, eight parameters (PM10, PM2.5, NO2, SO2, CO, O3and Pb) having
short-term standards should be considered for near real-time dissemination of AQI. It
is recognized that air concentrations of Pb are not known in real-time and cannot
contribute to AQI. However, its consideration in AQI calculation of past days will
help in scrutinizing the status of this important toxic.
A scientific basis in terms of attainment of air quality standards and dose-response
relationships of various parameters have been derived and used in arriving at
breakpoint concentrations for each AQI category .
It is proposed that for continuous air quality stations, AQI is reported in near real-time
for as many parameters as possible. For manual stations, the daily AQI is reported
with a lag of one week to ensure manual data are scrutinized and available for AQI.
A web-based AQI dissemination system is developed for quick, simple and an elegant
looking response to an AQI query. The other features of the website include reporting
of pollutants responsible for index, pollutants exceeding the standards and health
effects

Awareness of daily levels of air pollution is important to the citizens, especially for
those who suffer from illnesses caused by exposure to air pollution. Further, success
of a nation to improve air quality depends on the support of its citizens who are well-
informed about local and national air pollution problems and about the progress of
mitigation efforts. Thus, a simple yet effective communication of air quality is
important. The concept of an air quality index (AQI) that transforms weighted values
of individual air pollution related parameters (e.g. SO2, CO, visibility, etc.) into a
single number or set of numbers is widely used for air quality communication and
decision making in many countries.
After reviewing literature (on AQI), air quality monitoring procedures and protocols,
Indian National Air Quality Standards (INAQS), and dose-response relationships of
pollutants, an AQI system is devised. The AQI system is based on maximum operator
of a function (i.e. selecting the maximum of subindices of individual pollutants as an
overall AQI). The objective of an AQI is to quickly disseminate air quality
information (almost in real-time) that entails the system to account for pollutants
which have short-term impacts. Eight parameters (PM10, PM2.5, NO2, SO2, CO, O3,
NH3, and Pb) having short-term standards have been considered for near real-time
dissemination of AQI. It is recognized that air concentrations of Pb are not known in
real-time and cannot contribute to AQI. However, its consideration in AQI calculation
of past days will help in scrutinizing the status of this important toxic. The proposed
index has six categories with elegant colour scheme, as shown below

51
A scientific basis in terms of attainment of air quality standards and dose-response
relationships of various pollutant parameters have been derived and used in arriving at
breakpoint concentrations for each AQI category.
It is proposed that for continuous air quality stations, AQI is reported in near real-time
for as many parameters as possible. For manual stations, the daily AQI is reported
with a lag of one week to ensure manual data are scrutinized and available for AQI.
AQIs must be identified if these are from continuous or manual station to maintain
uniformity and clarity on sources of data. A web-based AQI dissemination system is
developed for quick, simple and elegant looking response to an AQI query. The other
features of the website include reporting of pollutant responsible for index, pollutants
exceeding the standards and health effects.

52
 CHAPTER 7

REFRENCES

 Agency for Toxic Substances and Disease Registry (ATSDR). 2007.

Toxicological profile for Lead. Atlanta, GA: U.S. Department of Health and
Human Services, Public Health Service.
 ATS (1991). “Lung Function Testing: Selection of Reference Values and
Interpretative Strategies”, Am Rev Respir Dis, 144, 1202-1219.
 Bates, D.V. (1999).“Introduction” in Air pollution and Health (Eds S.T.
Holgate, J.M. Samet, H.S. Koren, R.L. Maynard). Academic Press, New York.
 Beig G., Ghude S. D., Deshpande A., Scientific Evaluation of Air Quality
Standards and Defining Air Quality Index for India, 2010; Indian Institute of
Tropical Meteorology-Pune; ISSN 0252-1075.
 Ott, W.R. 1978. Environmental indices theory and practice.AnnArborScience
Publishers Inc., Ann Arbor, Mich. 48106.
 Pope, C.A. III (1989). “Respiratory Disease Associated with Community Air
Pollution and A Steel Mill - Utah Valley”, Am. J. Public Health, 79, 623-628.
 Pope, C.A. III (1991). “Respiratory Hospital Admissions Associated with
PM10 Pollution in Utah, Salt Lake, and Cache Valleys”, Arch. Environ.
Health, 46, 90-97.
 Pope, C.A. III, Dockery, D.W. (1992). “Acute Health Effects of PM10 on
Symptomatic and Asymptomatic Children”, 145, 1123-1128.
 Pope, C.A. III, Dockery, D.W. (1999). “Epidemiology of Particle Effects, Air
Pollution and Health”, Academic Press, 673-705.
 Pope, C.A. III, Hill, R.W., Villegas, G.M. (1999). “Particulate Air Pollution
and Daily Mortality on Utah’s Wasatch Front”, Environ. Health Perspect.,
107, 567-573.
 Romieu, I., Meneses, F., Ruiz, S., Sienra, J.J., Huerta, J., White, M.C., Etzel,
R.A. (1996). “Effects of Air Pollution on the Respiratory Health of Asthmatic
Children Living in Mexico City”, Am. J. Respir. Crit. Care Med., 154, 300-
307.
 Schwartz, J. (1996). “Air Pollution and Hospital Admissions for Respiratory
Disease”. Epidemiology, 7, 2028.
 Sharma M. (2009) Review of National Air Quality Criteria/Standards. Report
submitted to Central Pollution Control Board, New Delhi
 Sharma, M., Aggrawal, S, Bose P. (2002). Meteorology-based Forecasting of
Air Quality Index using Neural Network. Proceedings of International
Conference ICONIP’02-SEAL’02-FSKD’02 on Neural Network, Singapore,
November, 2002.
 Sharma, M., Maheshwari, M., and Pandey, R. 2001.Development of air
quality index for data interpretation and public information. IIT-Kanpur,
Report submitted to Central Pollution Control Board Delhi.
 Sharma, M.,Maheshwari, M., Sengupta, B., Shukla B.P. (2003). Design of a
website for dissemination of air quality index in
IndiaEnvironmentalModelling& Software 18 (2003) 405–411.
 Shenfeld, L. 1970. Note on Ontario’s air pollution index and alert system. J.
Air Pollut. Control Assoc.20 (9): 622.

53
 UK Air Quality Index, 2013 Revised; Committee on the Medical Effects of
Air Pollutants (COMEAP), Department of Environment, Food & Rural
Affairs, UK
 USEPA, Air Quality Index: A Guide to Air Quality and Your Health. February
2014, EPA-456/F-14-002
 USEPA. 1976. Federal register, Vol. 41 No 174- Tuesday September7, 1976.
 Vedal, S., Schenker, M.B., Munoz, A., Samet, J.M., Batterman, S., Speizer,
F.E.(1987). “Daily Air Pollution Effects on Children’s Respiratory Symptoms
and Peak Expiratory Flow”, Am. J. Public Health, 77, 694-698.
 WHO (2000) Air quality guidelines for Europe. Copenhagen, World Health
Organization Regional Office for Europe, 2000 (WHO Regional Publications,
European Series, No. 91).

54