Air_Pollution_EDA(2015-2022)¶

Background¶

Air pollution is widely believed as world's greatest environmental health threat.

Among the 30 most polluted cities in the world, 21 are located in India ,as reported by Swiss organisation, IQAir (based on PM 2.5 concentration.)

Objective¶

In this analysis we will be analysing the pollution level of indian cities over the last 7 years.

The aim is to analyze trends , pinpoint the possible causes and deriving insights from the large amount of granular data , relating to the concentration of air pollutants.

Eventually , we would propose solutions, based on the insights from data.

About the dataset¶

The dataset consist of daily air quality informations (i.e. particulate matter PM2.5,PM10, sulfur dioxide, nitrogen dioxide, carbon monoxide, and ozone etc.) of 26 major cities across india , It is extracted from cpcb.nic.in .

Steps Taken¶

Importing libraries¶

In [1]:

import numpy as np 
import pandas as pd 
import os

import matplotlib.pyplot as plt
%matplotlib inline
import seaborn as sns
sns.set()
import pycountry
import plotly.express as px
from plotly.offline import init_notebook_mode, iplot 
import plotly.graph_objs as go
import plotly.offline as py
from plotly.offline import download_plotlyjs,init_notebook_mode,plot,iplot
!pip install chart_studio
import chart_studio.plotly as py
import cufflinks
cufflinks.go_offline()
cufflinks.set_config_file(world_readable=True, theme='pearl')

import folium
from folium import Choropleth, Circle, Marker
from folium import plugins
from folium.plugins import HeatMap, MarkerCluster

!pip install bar_chart_race
import bar_chart_race as bcr
from IPython.display import HTML

plt.rcParams['figure.figsize'] = 8, 5
plt.style.use("fivethirtyeight")

for dirname, _, filenames in os.walk('/kaggle/input'):
    for filename in filenames:
        print(os.path.join(dirname, filename))

import warnings
warnings.filterwarnings('ignore')

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import datetime
import missingno as msno

Requirement already satisfied: chart_studio in c:\users\ankit sharma\miniconda3\lib\site-packages (1.1.0)
Requirement already satisfied: requests in c:\users\ankit sharma\miniconda3\lib\site-packages (from chart_studio) (2.28.1)
Requirement already satisfied: retrying>=1.3.3 in c:\users\ankit sharma\miniconda3\lib\site-packages (from chart_studio) (1.3.4)
Requirement already satisfied: six in c:\users\ankit sharma\miniconda3\lib\site-packages (from chart_studio) (1.16.0)
Requirement already satisfied: plotly in c:\users\ankit sharma\miniconda3\lib\site-packages (from chart_studio) (5.13.0)
Requirement already satisfied: tenacity>=6.2.0 in c:\users\ankit sharma\miniconda3\lib\site-packages (from plotly->chart_studio) (8.1.0)
Requirement already satisfied: charset-normalizer<3,>=2 in c:\users\ankit sharma\miniconda3\lib\site-packages (from requests->chart_studio) (2.0.4)
Requirement already satisfied: certifi>=2017.4.17 in c:\users\ankit sharma\miniconda3\lib\site-packages (from requests->chart_studio) (2022.12.7)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in c:\users\ankit sharma\miniconda3\lib\site-packages (from requests->chart_studio) (1.26.13)
Requirement already satisfied: idna<4,>=2.5 in c:\users\ankit sharma\miniconda3\lib\site-packages (from requests->chart_studio) (3.4)

Requirement already satisfied: bar_chart_race in c:\users\ankit sharma\miniconda3\lib\site-packages (0.1.0)
Requirement already satisfied: matplotlib>=3.1 in c:\users\ankit sharma\miniconda3\lib\site-packages (from bar_chart_race) (3.6.3)
Requirement already satisfied: pandas>=0.24 in c:\users\ankit sharma\miniconda3\lib\site-packages (from bar_chart_race) (1.5.2)
Requirement already satisfied: cycler>=0.10 in c:\users\ankit sharma\miniconda3\lib\site-packages (from matplotlib>=3.1->bar_chart_race) (0.11.0)
Requirement already satisfied: pillow>=6.2.0 in c:\users\ankit sharma\miniconda3\lib\site-packages (from matplotlib>=3.1->bar_chart_race) (9.4.0)
Requirement already satisfied: fonttools>=4.22.0 in c:\users\ankit sharma\miniconda3\lib\site-packages (from matplotlib>=3.1->bar_chart_race) (4.38.0)
Requirement already satisfied: kiwisolver>=1.0.1 in c:\users\ankit sharma\miniconda3\lib\site-packages (from matplotlib>=3.1->bar_chart_race) (1.4.4)
Requirement already satisfied: packaging>=20.0 in c:\users\ankit sharma\miniconda3\lib\site-packages (from matplotlib>=3.1->bar_chart_race) (22.0)
Requirement already satisfied: python-dateutil>=2.7 in c:\users\ankit sharma\miniconda3\lib\site-packages (from matplotlib>=3.1->bar_chart_race) (2.8.2)
Requirement already satisfied: contourpy>=1.0.1 in c:\users\ankit sharma\miniconda3\lib\site-packages (from matplotlib>=3.1->bar_chart_race) (1.0.7)
Requirement already satisfied: numpy>=1.19 in c:\users\ankit sharma\miniconda3\lib\site-packages (from matplotlib>=3.1->bar_chart_race) (1.23.5)
Requirement already satisfied: pyparsing>=2.2.1 in c:\users\ankit sharma\miniconda3\lib\site-packages (from matplotlib>=3.1->bar_chart_race) (3.0.9)
Requirement already satisfied: pytz>=2020.1 in c:\users\ankit sharma\miniconda3\lib\site-packages (from pandas>=0.24->bar_chart_race) (2022.7)
Requirement already satisfied: six>=1.5 in c:\users\ankit sharma\miniconda3\lib\site-packages (from python-dateutil>=2.7->matplotlib>=3.1->bar_chart_race) (1.16.0)

Data Cleaning¶

I start by importing the dataset into Jupyter notebook. After checking its dimensions and composition, I check for mixed-type data, missing values, and duplicates.

In [3]:

pwd = os.getcwd()

In [4]:

Imported_data = pd.read_csv( pwd + "/city_data.csv")
Imported_data

Out[4]:

	City	Date	PM2.5	PM10	NO	NO2	NOx	NH3	CO	SO2	O3	Benzene	Toluene	Xylene	AQI	AQI_Bucket
0	Ahmedabad	1/1/2015	NaN	NaN	0.92	18.22	17.15	NaN	0.92	27.64	133.36	0	0.02	0	NaN	NaN
1	Ahmedabad	2/1/2015	NaN	NaN	0.97	15.69	16.46	NaN	0.97	24.55	34.06	3.68	5.5	3.77	NaN	NaN
2	Ahmedabad	3/1/2015	NaN	NaN	17.4	19.3	29.7	NaN	17.4	29.07	30.7	6.8	16.4	2.25	NaN	NaN
3	Ahmedabad	4/1/2015	NaN	NaN	1.7	18.48	17.97	NaN	1.7	18.59	36.08	4.43	10.14	1	NaN	NaN
4	Ahmedabad	5/1/2015	NaN	NaN	22.1	21.42	37.76	NaN	22.1	39.33	39.31	7.01	18.89	2.78	NaN	NaN
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
54169	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
54170	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
54171	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
54172	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
54173	NaN	NaN	NaN	.	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

54174 rows × 16 columns

We can see in the above dataframe that , the rows in the end has null values only. So first step is to get rid of these null values.

In [5]:

Imported_data = Imported_data[Imported_data['City'].notna()]
Imported_data

Out[5]:

	City	Date	PM2.5	PM10	NO	NO2	NOx	NH3	CO	SO2	O3	Benzene	Toluene	Xylene	AQI	AQI_Bucket
0	Ahmedabad	1/1/2015	NaN	NaN	0.92	18.22	17.15	NaN	0.92	27.64	133.36	0	0.02	0	NaN	NaN
1	Ahmedabad	2/1/2015	NaN	NaN	0.97	15.69	16.46	NaN	0.97	24.55	34.06	3.68	5.5	3.77	NaN	NaN
2	Ahmedabad	3/1/2015	NaN	NaN	17.4	19.3	29.7	NaN	17.4	29.07	30.7	6.8	16.4	2.25	NaN	NaN
3	Ahmedabad	4/1/2015	NaN	NaN	1.7	18.48	17.97	NaN	1.7	18.59	36.08	4.43	10.14	1	NaN	NaN
4	Ahmedabad	5/1/2015	NaN	NaN	22.1	21.42	37.76	NaN	22.1	39.33	39.31	7.01	18.89	2.78	NaN	NaN
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
53263	Visakhapatnam	6/2/2023	130.17	369.34	118.16	63.8	130	8.13	1.5	24.67	13.77	12.09	44.85	4.61	NaN	NaN
53264	Visakhapatnam	7/2/2023	None	None	None	None	None	None	None	None	None	None	None	None	NaN	NaN
53265	Visakhapatnam	8/2/2023	52.6	154.07	30.98	31.22	39.93	5.77	1.32	19.69	15	3.46	10.85	0.98	NaN	NaN
53266	Visakhapatnam	9/2/2023	57.74	147.25	29.47	31.34	40.67	5.84	1.07	26.98	13.99	3.65	21.75	0.8	NaN	NaN
53267	Visakhapatnam	10/2/2023	60.09	153.13	30.36	32.69	42.08	4.54	0.98	28.22	14.05	2.99	24.62	0.96	NaN	NaN

53268 rows × 16 columns

Data prepration¶

In order to prepare data for analysis . 'None' value is set to NULL. Also , It is crucial to convert all the pollutant concentration to numeric values , to avoid any errors while plotting these values later on.

In [6]:

Imported_data = Imported_data.replace('None', np.NaN)

In [7]:

Imported_data['PM2.5'] = pd.to_numeric(Imported_data['PM2.5'])
Imported_data['PM10'] = pd.to_numeric(Imported_data['PM10'])
Imported_data['NO2'] = pd.to_numeric(Imported_data['NO2'])
Imported_data['CO'] = pd.to_numeric(Imported_data['CO'])
Imported_data['SO2'] = pd.to_numeric(Imported_data['SO2'])
Imported_data['O3'] = pd.to_numeric(Imported_data['O3'])
Imported_data['Benzene'] = pd.to_numeric(Imported_data['Benzene'])
Imported_data['Toluene'] = pd.to_numeric(Imported_data['Toluene'])
Imported_data['Xylene'] = pd.to_numeric(Imported_data['Xylene'])
Imported_data['NOx'] = pd.to_numeric(Imported_data['NOx'])
Imported_data['NO'] = pd.to_numeric(Imported_data['NO'])
Imported_data['NH3'] = pd.to_numeric(Imported_data['NH3'])

Date column is converted to Date Time format with the help of datetime library.

In [9]:

Imported_data['Date'] = pd.to_datetime(Imported_data['Date'])

In [10]:

print(f"The available data is between {Imported_data['Date'].min()} and {Imported_data['Date'].max()}")

The available data is between 2015-01-01 00:00:00 and 2023-12-01 00:00:00

Copy of the imported dataset is created , to keep the imported dataset in original format .

In [11]:

Imported_data_modified = Imported_data.copy()

Handling Missing infromation¶

To visually understand , the proporation of rows with missing information , The missingno library can be used.missingno is a third-party Python library used for visualizing missing data in datasets

In [18]:

Imported_data_modified.drop(columns= ["AQI" , "AQI_Bucket"] , inplace= True)

In [19]:

plt.style.use('seaborn-white')
msno.matrix(Imported_data_modified, );

Points to note:¶

We can clearly see that Xylene information is largely missing .
Missing information indicates the installed sensor's inability to measure pollutant's concentration.

In order to dig a little deeper , We can use heatmap for understanding the quantity of missing information , for each of the given pollutant. Heatmap is created by summing up the number of null values and converting them into percentages.

In [21]:

def missing_values_table(df):
    mis_val = df.isnull().sum()
    mis_val_percent = 100 * df.isnull().sum() / len(df)
    mis_val_table = pd.concat([mis_val,mis_val_percent] , axis = 1)
    mis_val_table_ren_columns = mis_val_table.rename(
    columns = {0 : 'Missing Values' , 1 : '% of Total Values'})

    mis_val_table_ren_columns = mis_val_table_ren_columns[
        mis_val_table_ren_columns.iloc[:,1] != 0 ].sort_values(
        '% of Total Values' , ascending = False).round(1)

    print('Your selected dataframe has' + str(df.shape[1]) + 'columns.\n' 
    'There are ' + str(mis_val_table_ren_columns.shape[0]) + ' columns that have missing values.')

    return mis_val_table_ren_columns

missing_values = missing_values_table(Imported_data_modified)
missing_values.style.background_gradient(cmap = 'YlOrBr')

Your selected dataframe has14columns.
There are 12 columns that have missing values.

Out[21]:

	Missing Values	% of Total Values
Xylene	33578	63.000000
Toluene	17023	32.000000
NH3	13809	25.900000
PM10	13056	24.500000
Benzene	9112	17.100000
O3	6798	12.800000
PM2.5	6762	12.700000
NOx	5781	10.900000
NO2	5677	10.700000
NO	5307	10.000000
SO2	5163	9.700000
CO	3244	6.100000

As expected Xylene's information is largely missing. The major pollutants like PM2.5 and PM10 concentration is within permissible limit.

Now , we can proceed further to analyze the dataset.

Correlation in the data¶

In correlation analysis, I use pandas to create a correlation matrix for all the numerical variables in the dataset. This will indicate the level of interdependence between the variables.

The volatile organic compounds Benzene , Toluene and Xylene are grouped together and represented as 'BTX'.

In [23]:

Imported_data_modified['BTX'] = Imported_data_modified['Benzene']+Imported_data_modified['Toluene']+Imported_data_modified['Xylene']
Imported_data_modified['Particulate_Matter'] = Imported_data_modified['PM2.5']+Imported_data_modified['PM10']
Imported_data_modified['Nitrogen Oxides'] = Imported_data_modified['NO']+Imported_data_modified['NO2']+Imported_data_modified['NOx']
Imported_data_modified.drop(['Benzene','Toluene','Xylene','PM2.5','PM10','NO','NO2','NOx'],axis=1,inplace=True)

plt.figure(figsize=(5,4))
sns.heatmap(Imported_data_modified.corr(),cmap='seismic',annot=True);

The heatmap shows that there are only moderate relationships (like - BTX with other pollutants , Particulate matter with Nitrogen Oxides). It is important to note that- There is no strong relationships (i.e cofficent above 0.75).

Exploratory Data analysis¶

I choose to use visual tools, charts, and graphs, to do the analysis.

We will be analysing the data by plotting it over the entire 7 year duration .

It would help us to identify trends , In the pollutants concentration.
It is imperative to test the Hypothesis that - "Air pollution is associated with the type of climate and specificity of particular months."

Let's now analyse the data to see what patterns and insights we can uncover from it.

In [12]:

pollutants = ['PM2.5' , 'PM10' , 'NO2' , 'CO' , 'SO2' , 'O3']

In [14]:

filtered_city_day = Imported_data_modified[Imported_data_modified['Date'] <= '2023-01-01']
filtered_city_day.set_index('Date' , inplace = True)
axes = filtered_city_day[pollutants].plot(marker = '.' , alpha = 0.5 , linestyle = 'None' , figsize = (16,30) , subplots = True)
for ax in axes:

    ax.set_xlabel('Years')
    ax.set_ylabel('ug / m3')

Points to Note¶

Pollutants like - PM2.5 , PM10 and NO2 show a clear seasonal effect, with pollution being higher in winter months as compared to the summer ones.
Increasing trend (over the 7 year period) in pollutant's concentration can be clearly seen.It paints a grim picture of pollution in india.
*Hypothesis* "Air pollution is associated with the type of climate and specificity of particular months." is True.

To dig a little deeper, i created subplots to analyze the distribution of pollutants over 5 year time period. To create the subplots I used the .subplot[] function in python.

It would help us to see the seasonal effect in more clearly.
It is imperative to test *Hypothesis* that- " Covid-19 pandemic lockdown (started in March 25 2020) had no effect on air pollution" .

In [82]:

Imported_data_modified['NH3'] = pd.to_numeric(Imported_data_modified['NH3'])

In [109]:

corr_with_AQI = Imported_data_modified.corr().PM10.sort_values(ascending = False)

In [88]:

metrices = corr_with_AQI[corr_with_AQI>0.01].index

In [62]:

Imported_data_modified['Year_Month'] = Imported_data_modified.Date.apply(lambda x : x.strftime('%Y-%m'))

In [74]:

Imported_data_modified = Imported_data_modified[Imported_data_modified['Year_Month'] <= '2023-01-01']
Imported_data_modified = Imported_data_modified[Imported_data_modified['Year_Month'] >= '2017-10-01']

In [89]:

df = Imported_data_modified.groupby(['Year_Month']).sum().reset_index()

sns.set_style('ticks')
fig, ax_ = plt.subplots(len(metrices), 1, figsize=(20,40))


fig.tight_layout(pad=4)
for i, col in enumerate(metrices,start = 0):
    
    x = df['Year_Month']
    y = df[col]
    ax_[i].plot_date(x ,y ,label=col, linestyle="--")
    ax_[i].set_xticklabels(df['Year_Month'], rotation=85);
    ax_[i].legend();

    

The trend is clearly visible , The concentraion of pollutants in summer months (i.e April to August) is lowest . While concentration is highest in the winter months (i.e November to February).

Possible reasons for this phenomenon :¶

Cold air is denser and moves slower than warm air after trapping pollutants.
Anthropogenic activities like - fireplaces and wood burners release pollutants in atmosphere, during winter season.

Also , there is a notable dip after 25 March 2020 lockdown for all pollutants. so the hypothesis that " Covid-19 pandemic lockdown (started in March 25 2020) had no effect on air pollution" . is False. The decrease in Particulate matter pollution and gaseous pollutants (CO and NO2) clearly reflects the impact of ceased industrial and vehicular activities during lockdown.

Pollution in cities :¶

To analyze the pollutantion levels in the cities. Let's start by listing down all the 26 cities present in this dataset. To identify any duplicates , due to improper naming.

In [92]:

Imported_data_modified['City'].unique()

Out[92]:

array(['Ahmedabad', 'Aizawl', 'Amaravati', 'Amritsar', 'Bengaluru',
       'Bhopal', 'Brajrajnagar', 'Chandigarh', 'Chennai', 'Coimbatore',
       'Delhi', 'Ernakulam', 'Gurugram', 'Guwahati', 'Hyderabad',
       'Jaipur', 'Jorapokhar', 'Kochi', 'Kolkata', 'Lucknow', 'Mumbai',
       'Patna', 'Shillong', 'Talcher', 'Thiruvananthapuram',
       'Visakhapatnam'], dtype=object)

In order to analyze the distribution of pollutants between the cities.

Proporation of pollutants can be plotted side by side using the treemap.

In [93]:

import plotly.express as px
from plotly.offline import init_notebook_mode, iplot 
import plotly.graph_objs as go
from plotly.offline import download_plotlyjs,init_notebook_mode,plot,iplot
!pip install chart_studio
import chart_studio.plotly as py
import cufflinks
cufflinks.go_offline()
cufflinks.set_config_file(world_readable=True, theme='pearl')

df = Imported_data_modified.drop(columns = ['Date', 'AQI_Bucket', 'AQI']).groupby('City').sum().reset_index()
fig = px.treemap(pd.melt(df, id_vars = 'City'), path=['City','variable'],
                 values=pd.melt(df, id_vars = 'City')['value'],
                 title = 'Cities and the proportion of pollutants in each')
fig.show()

North indian like cities - Delhi , Gurugram . tops this visualization.

Possible reasons for this phenomenon :¶

Northern part of india and experience very low temperature during winter season.
In northern india , Due to stubble burning (after harvest) in states like- Haryana, punjab(aka garnary of india). pollution increases significantly.
They are some of the most urbanized cities of india.

In cities like Ahmedabad and Patna, The high rate of urbanisation and industrialisation are the main reason for high pollution levels.

Let's dig down a little deeper , In order to find out the cities that reported maximum concentration of pollutants , like - PM2.5 , PM10 , SO2 etc.

In [94]:

def max_polluted_city(pollutant):
    x1 = Imported_data_modified[[pollutant , 'City']].groupby(['City']).mean().sort_values(by = pollutant , ascending = False).reset_index()
    x1[pollutant] = round(x1[pollutant],2)
    return x1[:10].style.background_gradient(cmap = 'coolwarm')

In [96]:

from IPython.display import display_html
def display_side_by_side(*args):
    html_str = ''
    for df in args:
        html_str += df.render()
    display_html(html_str.replace('table', 'table style  = "display:inline"'),raw = True)

In [100]:

pm2_5 = max_polluted_city('PM2.5')
pm10 = max_polluted_city('PM10')
no2 = max_polluted_city('NO2')
so2 = max_polluted_city('SO2')
co = max_polluted_city('CO')

display_side_by_side(pm2_5,pm10,no2,so2,co)

	City	PM2.5
0	Delhi	115.980000
1	Gurugram	114.780000
2	Patna	112.450000
3	Lucknow	96.220000
4	Mumbai	69.630000
5	Jorapokhar	64.370000
6	Guwahati	63.040000
7	Ahmedabad	60.590000
8	Jaipur	59.620000
9	Kolkata	59.360000

	City	PM10
0	Delhi	227.930000
1	Gurugram	208.870000
2	Patna	165.060000
3	Lucknow	146.860000
4	Jorapokhar	132.790000
5	Jaipur	129.820000
6	Ahmedabad	128.350000
7	Talcher	121.680000
8	Bhopal	117.490000
9	Guwahati	117.430000

	City	NO2
0	Ahmedabad	57.290000
1	Delhi	48.350000
2	Jaipur	42.220000
3	Visakhapatnam	36.470000
4	Lucknow	35.990000
5	Kolkata	34.650000
6	Patna	34.320000
7	Hyderabad	31.180000
8	Bengaluru	29.050000
9	Coimbatore	26.800000

	City	SO2
0	Ahmedabad	45.750000
1	Jorapokhar	33.450000
2	Talcher	27.240000
3	Bhopal	18.150000
4	Mumbai	17.250000
5	Guwahati	17.070000
6	Brajrajnagar	16.400000
7	Patna	15.910000
8	Visakhapatnam	14.900000
9	Delhi	14.340000

	City	CO
0	Ahmedabad	13.680000
1	Jorapokhar	9.530000
2	Lucknow	4.490000
3	Bengaluru	3.820000
4	Delhi	1.710000
5	Talcher	1.550000
6	Brajrajnagar	1.420000
7	Patna	1.360000
8	Gurugram	1.150000
9	Kochi	1.110000

As expected , the same cities has the highest mean value.

Now , According to cpcb.nic.in . The PM2.5 and PM10 are most crucial pollutant to find out the air quality.

Permissible limit for the annual average of pollutants PM2.5 is 40 ug/m3 , PM10 is 60 ug/m3. As seen in visualization above , The Indian cities PM2.5 and PM10 levels are signifcantly higher than permissible limit.

Status of indian cities :

Delhi has Avg(PM2.5) = 115.98 (3X of limit) , Avg(PM10) = 227.93 (3.5X of limit)
Gurugram has Avg(PM2.5) = 114.78 (3X of limit) , Avg(PM10) = 60 (3X of limit)
Even the cities ranked 10th in the list i.e Kolkata's Avg(PM2.5) = 59.3 (1.5X of limit) and Guwhati's Avg(PM10) = 117.4 (2X of limit)

These figures clearly shows that "air pollution is a silent crisis in india , It's an emergency".

It could be clearly the reason for the abrupt rise in diseases such as Lung cancer in these cities.

Conclusion¶

Findings¶

Air pollution is associated with the type of climate and specificity of particular months. In other words , The Pollution shows seasonal affect , as the concentraion of pollutants in summer months (i.e April to August) is lowest . And highest in the winter months (i.e November to February).
Urbanized North indian cities - Delhi , Gurugram. are seriously affected due to pollution . As during winters , low temperature coupled with stubble burning( after harvest) in the adjoining states , leads to significant rise in pollutants concentration.
There was a notable dip in pollutants concentration due to lockdown of Covid-19 pandemic .
The Average particulate matter in some polluted indian cities is over 3 times the maximum permissible limit.

Proposed solutions¶

Government should double down the effort during the winter season to control the pollution.
The notable fall in the pollutants concentration during Covid-19 pandemic , has shown that reduction in the use of polluting vehicles and industries , could lead to an immediate relief from the severe air pollution. Therefore , promotion of public transport and curbing polluting industries , could provide an impressive cut in pollution levels.
The people should be informed about the fact that pollutants level is twice to thrice times above the safe limit. And , its impact on their health.