Analysis of AQI and weather data

to obtain useful insights for

different cities
By
Muskan Verma (2019UCE1525)
Nitish Sharma (201UCE1198)
Vaishali Rawat (2019UCE1766)
Submitted
In partial fulfillment of the requirements in the degree of
Bachelor of Technology (Civil Engineering)

Department of Civil Engineering

Malaviya National Institute of Technology, Jaipur

CERTIFICATE
This is to certify that the project report entitled
"Analysis of AQI and weather data to obtain
useful insights for different cities"which is being
submitted by Vaishali Rawat (2019UCE1766),
Muskan Verma (2019UCE1525), Nitish Sharma
(2019UCE1198) for the partial fulfillment of the
requirements of the degree of B.Tech. (Civil
Engineering) to the Department of Civil
Engineering,Malaviya National Institute of
Technology Jaipur has been carried out under my
supervision and guidance. This work is approved for
submission.

Dr. Ruchi Sharma

Supervisor

Dept. of Civil Engg.

MNIT, Jaipur

ACKNOWLEDGEMENT
We would like to express our sincere gratitude to our
supervisor Dr.Ruchi Sharma for their invaluable
support and guidance throughout the completion of
this report.
We are also grateful to our project coordinators
Dr.Urmila Brighu and Dr.Manoj Diwakar for
their insightful feedback and suggestions, which
helped us to refine our ideas and improve the quality
of our work.
We are grateful to the Department of Civil
Engineering, MNIT Jaipur, for allowing us to
execute this project, which is an integral part of the
curriculum in B. Tech. Program at the Malaviya
National Institute of Technology Jaipur.

Date: Vaishali Rawat

MuskanVerma
Nitish Sharma

ABSTRACT
TABLE OF CONTENTS
 Particulars Page no.
Certificate

Acknowledgements

Abstract

Table of content

1.Introduction

1.1 Air pollution

1.2 AQI

1.3 Heatmap

1.4 Histogram

1.5 Pearson correlation

2. Literature review

Research gaps

3. Methodology

Chapter1
 Introduction
1.1 Air pollution
Air pollution has been a problem for a long time. Any
toxic substance, including chemicals, particulate matter, or
biological materials that is present in sufficient quantities
to such an extent to change the physical and chemical
properties of the air, endanger people or other living
things, harm the environment or man made environment,
or any combination these, is regarded as air pollution.
Urban air pollution take place due to anthropogenic
emission sources, which includes domestic fuel
combustion, industries, and vehicles. Also some natural
phenomena such as volcanic eruption, airborne germs and
viruses, wildfires, earthquakes, and landslides are also the
contributing factors to air pollution.
Additionally, local meteorological factors like
temperature, relative humidity, wind speed, solar
radiation and pressure also have potential to affect the
concentration of pollutants. Any area's pollution levels and
air quality are determined by the source of emissions,
weather patterns, and other factors.These factors can have
a significant impact on human health, climate change,
agricultural crops, ecosystems, and forest species.
According to the World Air Quality Report 2021 released
by IQAir, a Swiss firm, India is the fifth most polluted
country among 117 countries, regions and territories
around the world, assessed.
The US environmental protection agency (EPA) has
defined six common air pollutants as criteria pollutants
which includes.
1. Ozone (O3): Ozone gas is formed by the reaction of
sunlight with NOx and volatile organic compounds
(VOCs) which is emitted by cars, factories, and other
sources. Ozone can cause respiratory problems and can
damage crops and other vegetation.
2. Carbon monoxide (CO):This gas is produced by
incomplete combustion of fossil fuels, wood, and other
organic materials. The persons exposed to high levels of
CO can have symptoms of headache, dizziness, nausea,
and even death.
3. Nitrogen oxides (NOx): Combustion processes taking
place in vehicles and power plants leads to the production
in oxides of nitrogen. These oxides contribute to the
formation of smog and acid rain and can cause respiratory
problems.
4. Sulfur dioxide (SO2): This gas is the result of the
burning of fossil fuels, particularly coal. It is also a
contributing factor of the formation of acid rain and may
cause respiratory problems.
5. Particulate matter (PM):Volatile organic compounds
(VOCs) and tiny particles and air droplets in the air
produces a complex mixture of particulate matter . It can
also impact crops and vegetation.
6. Lead (Pb): It is a heavy metal that can be found in
gasoline and other industrial products. Its exposure to
children may have neurological problems in them.
In this perspective, the present study comes therefore to
establish a relationship between meteorological
parameters (temperature, humidity, wind speed and solar
radiation, pressure, rainfall) and criteria air pollutants
concentration levels in 4 Indian cities (Jaipur, Bengaluru,
Guwahati, Srinagar).

1.2 AQI
Air Quality Index (AQI) is a numerical value that
represents the air quality of a particular area. It’s use lies
in providing information to the public regarding the level
of air pollution and its potential health effects.
In India, the Central Pollution Control Board (CPCB) is
responsible for presenting AQI values to the public and for
monitoring of air quality in major cities across the country
It also makes available real-time air quality data to public
through the National Air Quality Index (NAQI) website
and mobile app. Under Swachh Bharat Abhiyan, the
National Air Quality Index was launched in September
2014 in New Delhi to measure India’s air quality index.
The CPCB formed an expert group of medical
professionals, air quality experts, academia, advocacy
groups , SPCBs. Also, a technical study was awarded to
IIT Kanpur . In 2014 , the IIT Kanpur and the expert group
recommended an AQI scheme for India .
Few instruments that are used for measuring air quality
are:
 PCE-RCM 05
 PCE-RCM 8
 PCE- HFX 100
A real time data tracking system for continuous
monitoring is available in six Indian cities: Ahmadabad,
New Delhi, Kolkata ,Mumbai, and Pune.

CENTRAL POLLUTION CONTROL BOARD’s AIR

QUALITY STANDARDS:

AIR QUALITY INDEX CATEGORY

0-50 Good
51-100 Satisfactory
101-200 Moderate
201-300 Poor
301-400 Very poor
401-500 Severe
 Chapter 2
Literature review
S.No. Author and year Major Conclusions
1. Sigamani et al., Air quality prediction with
2022 influence of influence of
meteorological parameters using
machine learning model for IOT
application
 This study found that random
forest, support vector machine
were effective in predicting AQI.
Random Forest was most
accurate with accuracy of 90%.
 This model showed that as the
 temperature increased, AQI also
increased, while higher wind
speed and humidity were
associated with lower AQI.
2. Hui DI et al., Correlation analysis of AQI
2019 characteristics and
meteorological conditions in
heating season This study
showed that AQI was positively
correlated with PM2.5,NO2 and SO2
concentrations, while negatively
correlated with temperature and
wind speed. this study showed that
significant factor affecting AQI
during the heating season in Beijing
was PM2.5 followed by temperature
and wind speed.
3. Z Poorolaie et Determination of Air Quality
al., 2022 Index and its relationship with
Meteorological parameters in city
of Mashhad, Iran This study
found that the AQI was highest
during summer months, the main
pollutant being PM10, SO2 AND
N02. AQI was positively correlated
with temperature and relative
humidity and negatively with wind
speed.
4. Nairita Sarkar et Air Quality Index prediction
al., 2022 using an effective hybrid deep
learning model In this study the
author combined convolutional
neural network (CNN)and long-
short term memory
(LSTM)networks to build a hybrid
model to predict AQI. In this study
sensitivity analysis was also
 performed to identify the most
significant air pollutants affecting
AQI.
5. Shrabanti dutta et Urban air quality assessment and
al., 2021 inferring the association between
different factors: A comparative
study among Delhi, Kolkata and
Chennai Megacity of India In this
study statistical analysis techniques
such as principal component
analysis and cluster analysis were
used to identify the most significant
air pollutants and their sources in
the cities . Delhi had the highest
level of air pollution with
particulate matter being the main
source.At Kolkata and Chennai
major sources of pollution were
NO2 And SO2 respectively.
6. Munoth et GIS - based air quality index
al.,2021 spatial model for Indian cities In
this study GIS techniques were used
to develop a spatial model that
incorporates the spatial and
temporal variations of air pollution
in different parts of India. With the
sensitivity analysis the most
significant pollutant found was
particulate matter.
7. Bhat et al., 2021 Data analytics based statistical
analysis of air pollution in the
major cities of Karnataka
This study found that levels of
PM2.5 were higher than the levels of
PM10 in most of the cities. The
authors used time-series analysis to
investigate the trends and patterns
 of air pollutant levels over time.
8. kumar et al., Air pollution prediction with
2022 machine learning: A case study of
Indian cities This study found
that random forest algorithms
outperformed the other algorithms
in predicting PM2.5, while SVR was
found to be the most accurate in
predicting NO2 .
9. T Tharani et al., Spatial distribution analysis of air
2021 pollutants and the impact of
meteorological factors In this
study spatial distribution analysis
was determined by inverse distance
method(IDW) interpolation
technique using ArcGIS 10.5. this
study showed positive correlation
between concentration of air
pollutants and temperature,wind
speed and turbulence and negative
correlation with precipitation.
10. Deng et al., 2020 Correlation and principal
component regression analysis for
studying air quality and
meteorological elements in
Wuhan, china In This study
principal component regression
(PCR) technique was used to
predict the daily AQI based on the
previous day's AQI. The
atmospheric pressure was
negatively correlated with AQI
while temperature, relative
humidity ,precipitation and wind
speed were negatively correlated .
11. Pang et al., 2023 Air Quality-Meteorology
Correlation Modeling Using
 Random Forest and Neural
Network This study showed that
both random forest and neural
network were able to effectively
model the correlation between air
quality and meteorological factors .
The random forest model
outperformed the neural network in
terms of prediction accuracy.
12. Fernando et al., Prediction of Air Quality Index in
2022 Colombo In this study PM2.5
concentration in Colombo had been
predicted using four correlated air
pollutant concentration such as
SO2,NO2,PM2.5&PM10 . Random
Forest was recognized as best
appropriate prediction model with
over 85% accuracy.
13. Miao et al.,2022 Spatiotemporal heterogeneity
analysis of air quality in the
Yangtze River Delta, China In
this study the primary pollutants for
each metropolitan area had been
identified using Gray relational
analysis(GRA)method. Also spatial
heterogeneity of the effects of 6 air
influencing factors on AQI had
been analyzed using geographically
weighted regression(GWR) .
14. Zhang et al., A Multi- step Prediction Method
2022 of Urban Air Quality Index Based
on meteorological Factors
Analysis In this the proposed
method included four steps-data
preprocessing, feature selection,
model training and prediction.The
results showed that method
 proposed in this chapter has better
accuracy over classical ARIMA
method
15. Lodhi et al., 2022 Comparative Analysis of
Forecasting Models for Air
Quality Index Prediction In this
study the LSTM(Long Short Term
Memory) Model,SARIGMA model
have been used to forecast AQI.
And the sub index calculation is
used in LSTM model for AQI
calculation .
16. Liu et al., 2022 Analysis and Forecast of Beijing’s
Air Quality Index Based on
ARIMA Model and Neural
Network Model The author found
that both ARIMA and NN model
can effectively predict AQI in
Beijing. Although NN model
performed slightly better than
ARIMA Model.

RESEARCH GAPS
1.Lack of Interpretability: Many studies do not provide
insights into the underlying mechanisms of the
relationship between AQI and weather parameters.
2.Limited Sample Size: The study focuses on a limited
number of monitoring stations in a specific city.
3.In many studies all the parameters of meteorology is not
being considered which may have a significant impact on
AQI.
4.No study is being carried out to obtain the correlation of
different cities of 4 distinct corners of the country (north,
south, east, west)

METHODOLOGY

Step 1: Selection of cities

The below 4 cities has been selected as our area of study

in India-

1. Srinagar (North)
2. Jaipur (West)
3. Bengaluru (South)
4. Guwahati (East)

1. Srinagar :
 Srinagar is the largest city in Jammu and
Kashmir. It is located at 34.0837° N latitude and
74.7973° E longitude.
 It has a humid sub-tropical climate, with mild to
warm summers and cold winters.
 The summer season lasts from June to
September, with average temperatures ranging
from 16°C to 30°C.
 The winter season lasts from December to
February, with average temperatures ranging
from -2°C to 7°C.
 The city experiences heavy snowfall during the
winter season. The monsoon season lasts from
July to September, with an average annual
rainfall of about 650 mm.

2. Jaipur :
 Jaipur is the capital and largest city of the Indian
state of Rajasthan. It is located in the
northwestern part of India. Jaipur's latitude and
longitude are 26.9124° N, 75.7873° E,
respectively.
 Jaipur has a semi-arid climate with hot summers
and mild winters.
 The city experiences three distinct seasons -
summer, monsoon, and winter. Summer starts
 from March and lasts until June, with
temperatures ranging from 30°C to 45°C.
 Monsoon starts in July and lasts until
September, bringing relief from the summer heat
with occasional heavy rainfall.
 Winter starts in December and lasts until
February, with temperatures ranging from 10°C
to 27°C.
 January is the coldest month of the year, with an
average temperature of 15°C.
In conclusion, Jaipur's climate, weather, topography,
and location make it a unique and interesting place to
study.

3. Bengaluru :
 It is located in the southeastern part of the state,
at an altitude of 920 meters (3,020 feet) above
sea level. It's latitude and longitude are 12.97° N
and 77° 34' E respectively.
 It has a tropical climate, characterized by warm
to hot temperatures throughout the year and
distinct wet and dry seasons.
 The city experiences three different seasons -
summer, monsoon, and winter.
 Summer starts from March and lasts until May,
with temperatures ranging from 23°C to 35°C.
  Monsoon starts in June and lasts until
September, with heavy rainfall and occasional
thunderstorms.
 Winter starts in December and lasts until
February, with temperatures ranging from 12°C
to 28°C.

4. Guwahati :
 Guwahati is a city located in the Indian state of
Assam. It has a latitude of 26.1440° N and
longitude of 91.7362° E.
 It has a humid sub-tropical climate with hot
summers and mild winters.
 The monsoon season lasts from June to
September, during which the city experiences
heavy rainfall. The average annual rainfall in
Guwahati is around 1,750 mm.
 The temperature can range from 6°C in winter to
36°C in summer.

Step 2: Data Acquisition & Resolution

 The data has been downloaded from CPCB (Central
Pollution Control Board) website.
 Last 3 years data (2020-2022) has been used for the
analysis.
 Resolution of the data is 24 hours.
Step-3: Data Cleansing
Data cleaning involves correcting or deleting inaccurate,
damaged, improperly formatted, duplicate, or insufficient
data from a dataset.
These are following steps for data cleansing –
 Data is imported using pandas library
 The code checks Data Frame for missing values and
handling them depending on how many missing
values are present.
 If more than 10% of the total values in the Data
Frame are missing, then missing values are replaced
with the mean value of each column.
 If less than 10% of the total values are missing, the
rows containing missing values are dropped from the
Data Frame.
 Heat map has been used for the representation of
missing values.

Heat map: A heat map is a graphical representation of

data that uses color-coded squares to represent values.
Step-4: AQI Calculation
 The formula to calculate AQI is the same as per the
Indian CPCB and US-EPA.
 The AQI is calculated using the equations for each
parameters separately.
 Worst sub-index is the AQI.
 A sub-index is a linear function (two different yet
related notions) of the concentration of pollutants.

Ip = [IHi – ILo / BPHi – BPLo] (Cp – BPLo) + ILo

Where,
 Ip =index of pollutant p
 Cp = truncated concentration of pollutant p
  BPHi = concentration breakpoint i.e. greater than or
equal to Cp
 BPLo = concentration breakpoint i.e. less than or
equal to Cp
 IHi = AQI value corresponding to BPHi
 ILo = AQI value corresponding to BPLo

Step-5: Techniques Used

1. Histogram:
 A histogram can be used to analyze the distribution of
meteorological parameters in a particular area over a
given period of time.

 By creating a histogram of meteorological parameters,

it is possible to visualize the frequency and distribution
of different levels of pollution.

 The histogram plots the data into bins, which are

defined intervals or ranges of values, and counts the
number of data points that fall within each bin.

 The height of each bar on the histogram represents the

frequency of data points in that bin.
2. Pearson correlation:
 Pearson correlation is a statistical measure that
describes the strength and direction of the linear
relationship between two variables.
 In the context of AQI, Pearson correlation can be used
to examine the relationship between meteorological
parameters and their impact on air quality.
 It is independent of the unit of measurement of the
variables.
 For example, suppose the unit of measurement of one
variable is in years while the unit of measurement of the
second variable is in kilograms. In that case, even then,
the value of this coefficient does not change.
 The Pearson Correlation Coefficient formula is as
follows:
Where,

 r = Pearson Coefficient
 n= number of pairs of the stock
 ∑xy = sum of products of the paired stocks
 ∑x = sum of the x scores
 ∑y= sum of the y scores
 ∑x2 = sum of the squared x scores
 ∑y2 = sum of the squared y scores