Indian Crop Production Anlysis and Relevent Climate Effects¶

Presenters:¶

  • Alan Wang

Table of Contents

  • Indian Crop Production Anlysis and Relevent Climate Effects
      • Presenters:
      • Introduction
        • India economy sector distribution
        • Population dependency
      • Indian agriculture is particularly vulnerable due to:
        • Long coatline and relatively low altitude:
      • Indian top ten crop production since 2010
      • We can see the connection between temperature and CO2 emission
      • What happened to the cotton production in Gujarat and Maharashtra?
        • We can see that the production in most districts in Gujarat dropped a lot in two year
      • Blue: Decrease in production
      • Yellow or light blue: not increase or decrease that much
      • Red: Increase in production
      • The production in eastern part tends to grow by a little bit while the production in other area drops a lot
      • Blue: Decrease in Area
      • Yellow or light blue: not increase or decrease that much
      • Red: Increase in Area
      • The declination resulted from multiple reasons:
      • This is the prices and rate of change data of Soybean and Cotton since 2007
      • We can see that around the same period from 2011 to 2012, the price of cotton slumped by more than 40%. That's partly why the production also decreased
      • Some suggestions
        • If you are interested
        • all reported huge decline of cotton production in India
    • References
In [1]:
%matplotlib inline
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
#import seaborn as sns
#from scipy.stats import ks_2samp
#import glob
#import time
import requests
import bs4
#import lxml
import altair as alt
import folium
import geopandas as gdp
import math
import json
#from mpl_toolkits import mplot3d
import plotly
import plotly.express as px
import plotly.graph_objects as go
from ipywidgets import interact
In [2]:
#read the data
co2 = pd.read_csv("co2.csv")
temp = pd.read_csv("temperatures.csv")
crop = pd.read_csv("crops.csv")
pop = pd.read_excel('Population Estimates for Districts and PCs in India, 2020.xlsx')
geoinfo = pd.read_csv('district wise centroids.csv')
cs = pd.read_excel('cotton-180.xlsx',skiprows = 1)

Introduction¶

India is highly vulnerable to climate change due to its considerable and growing population and growing economy, both closely tied to a limited natural resource base. More than half of the 1.3 billion population directly depend on climate sensitive sectors (agriculture, forestry, fisheries) and natural resources (water, biodiversity and coastlands) for subsistence and livelihood.[2] As India strives to develop its economy, rising industrialization, urbanization, agricultural commercialization, and expansion of power generation capacity based predominately on non-renewable fuels will rapidly increase green-house gas (GHG) emissions.

India economy sector distribution¶

In [3]:
#scraping resource from[3]
sector = requests.get('https://statisticstimes.com/economy/country/india-gdp-sectorwise.php#:~:text=Sector%2Dwise%20GDP%20in%20India,Financial%2C%20real%20estate%20%26%20prof%20servs').text
sector = bs4.BeautifulSoup(sector)
tab = sector.find('table', attrs = {'id': 'table_id4'})
headers = []
for i in tab.find_all('th'):
    headers.append(i.text)
head =[]
for h in headers[:7]:
    head.append(h)
    head.append(h+'1')
head = head[1:]
mydata = pd.DataFrame(columns = head)
for j in tab.find_all('tr')[2:]:
    row_data = j.find_all('td')
    row = [k.text for k in row_data]
    mydata.loc[len(mydata)] = row
#slicing
mydata = mydata[['Year1','Agriculture & Allied', 'Agriculture', 
                 'Industry', 'Mining & Quarrying', 'Manufacturing', 
                 'Services']]
mydata['Year1'] = mydata['Year1'].str[:4].astype('int')
#Dataframe slicing,cleaning
mydata = mydata.sort_values(by = 'Year1', ascending = False).set_index('Year1')
value = [18.2, 15.42+ 19.53, 22.05, 2.7+ 5.16,12.89]
value.append(100-np.array(value).sum())
value.sort(reverse = False)
mydata.iloc[0]
Out[3]:
Agriculture & Allied    18.20
Agriculture             15.79
Industry                24.77
Mining & Quarrying       2.13
Manufacturing           12.89
Services                57.03
Name: 2013, dtype: object
In [4]:
# Weighted Pie
source = pd.DataFrame({"category": 
                       ['Other','1.Construction', '2.Manufacturing', '3.Agriculture & Allied', '4.Financial, real-estate, etc', '5.Other servs'],"values": value})
base = alt.Chart(source).encode(
    theta=alt.Theta("values:Q", stack=True),
    radius=alt.Radius("values", scale=alt.Scale(type="sqrt", zero=True, rangeMin=20)),
    color="category:N",
)

c1 = base.mark_arc(innerRadius=20, stroke="#fff")

c2 = base.mark_text(radiusOffset=10).encode(text="values:Q")

c1 + c2
Out[4]:

Population dependency¶

About 60 to 88% of the human population depends on agriculture for their livelihoods, of which 12 to 93% of the people live in rainfed areas and 26 to 84% of the arable land.[1]

In [5]:
#Population dependency pie chart
source = pd.DataFrame(
    {"category": ["Agriculture", "Industry", "Service"], "value": [70, 12 ,18]}
)

base = alt.Chart(source).encode(
    theta=alt.Theta("value:Q", stack=True), color=alt.Color("category:N", legend=None)
)

pie = base.mark_arc(outerRadius=120)
text = base.mark_text(radius=150, size=12).encode(text="category:N")

pie + text
Out[5]:

Indian agriculture is particularly vulnerable due to:¶

  1. Predominantly unirrigated (rain-fed) agriculture[1] largely contingent on monsoon rains,
  2. Himalayan glacier-fed, river based irrigation in the north and east,
  3. Massive livestock population, specifically cattle and water buffalo,
  4. Densely populated, low lying 7,500 kilometer coastline,
  5. Large number of resource-poor farmers and other producers with limited ability to adopt sustainable resource technologies.
In [6]:
pop['District Name, State Letter Code'] = pop['District Name, State Letter Code'].str.split(',').apply(lambda x:x[0])
toge = pop.merge(geoinfo, left_on = 'District Name, State Letter Code', right_on = 'District', how = 'inner')
s_pop = pop.groupby('State Name').sum().reset_index()
s_pop.sort_values(by = 'District Population').head()
geodf = gdp.read_file('https://raw.githubusercontent.com/Subhash9325/GeoJson-Data-of-Indian-States/master/Indian_States')
s_pop['State Name'] = (s_pop['State Name'].replace('Andaman & Nicobar Islands','Andaman and Nicobar')
                      .replace('Jammu & Kashmir','Jammu and Kashmir')
                      .replace('NCT of Delhi','Delhi' )
                      .replace('Odisha', 'Orissa')
                      .replace('Uttarakhand', 'Uttaranchal')
                      .replace('Dadra and Nagar Haveli and Daman and Diu','Dadra and Nagar Haveli' ))
s_pop.head()
Out[6]:
State Name District Population
0 Andaman and Nicobar 4.309542e+05
1 Andhra Pradesh 5.649770e+07
2 Arunachal Pradesh 1.592486e+06
3 Assam 3.561591e+07
4 Bihar 1.184479e+08
In [7]:
s_pop2 = s_pop.set_index('State Name')['District Population']
center_lat = toge['Latitude'].mean()
center_long = toge['Longitude'].mean()

m = folium.Map(location=[center_lat, center_long],tiles='cartodbpositron')#'CartoDB positron'

folium.FitBounds([(center_lat-15,center_long-19), (center_lat+12,center_long+19)]).add_to(m)
url = (
    "https://raw.githubusercontent.com/Subhash9325/GeoJson-Data-of-Indian-States/master/Indian_States"
)
import branca
steps = s_pop2.to_numpy()
colorscale = branca.colormap.linear.YlOrBr_05.to_step(steps, np.arange(0,1,0.001)).scale(1e+04, 1.2e+08)
def style_function(feature):
    popu = s_pop2.get((feature['properties']['NAME_1']), None)
    return {
        "fillOpacity": 0.6,
        "weight": 0,
        "fillColor": "#black" if popu is None else colorscale(popu),
    }

folium.GeoJson(
    json.loads(requests.get(url).text),
    style_function=style_function,
).add_to(m)

m
#folium.TileLayer(tiles = 'MapQuest Open Aerial', attr = 'https://www.mapquest.com/').add_to(m)
Out[7]:
Make this Notebook Trusted to load map: File -> Trust Notebook

Long coatline and relatively low altitude:¶

In [8]:
from IPython.display import Image
Image(filename='altitude.png', width = 600, height = 800)
Out[8]:

Indian top ten crop production since 2010¶

In [9]:
top_ten = crop[(crop['Crop_Year'] > 2009) & (crop['Crop_Year'] < 2015)].groupby('Crop').mean()['Production'].nlargest(11).drop('Total foodgrain').index
recent_top_10 = crop.set_index("Crop").loc[top_ten].reset_index()
recent_five_year = recent_top_10[(recent_top_10['Crop_Year'] > 2009) & (recent_top_10['Crop_Year'] < 2015)].dropna()[['Crop_Year','Crop', 'Production']].groupby(['Crop_Year', 'Crop']).mean().reset_index()
recent_five_year['Crop_Year'] = recent_five_year['Crop_Year'].astype('string')
recent_five_year['Production'] = recent_five_year['Production'].apply(lambda x: x**(1/3))#math.log(x,1.001)

stack_bar = alt.Chart(recent_five_year).mark_bar().encode(
    x=alt.X('Production',title = 'Cube Root Production'),
    y='Crop_Year',
    color='Crop'
).properties(title = 'stack bar chart')
stack_bar
Out[9]:
In [ ]:

A recent report[4] by the Government of India Ministry of Earth Sciences notes that since the middle of the twentieth century, India has experienced a rise in average temperatures, a decrease in monsoon precipitation, a rise in extreme temperature, rainfall and drought events, an increase in intensity of severe cyclones and other changes in the monsoon system. And we expect these climate change distresses are expected to continue during the twenty-first century.

In [10]:
tem = temp[['YEAR','ANNUAL','JAN-FEB', 'MAR-MAY','JUN-SEP','OCT-DEC']].melt('YEAR', var_name = 'category', value_name = 'temperature')
# Create a selection that chooses the nearest point & selects based on x-value
nearest = alt.selection(type='single', nearest=True, on='mouseover',
                        fields=['YEAR'], empty='none')
# The basic line
line = alt.Chart(tem).mark_line(interpolate='basis').encode(
    alt.X('YEAR:Q'),
    alt.Y('temperature:Q',scale = alt.Scale(zero = False)),
    color='category:N'
)
# Transparent selectors across the chart. This is what tells us
# the x-value of the cursor
selectors = alt.Chart(tem).mark_point().encode(
    x='YEAR:Q',
    opacity=alt.value(0),
).add_selection(
    nearest
)
# Draw points on the line, and highlight based on selection
points = line.mark_point().encode(
    opacity=alt.condition(nearest, alt.value(1), alt.value(0))
)
# Draw text labels near the points, and highlight based on selection
text = line.mark_text(align='left', dx=5, dy=-5).encode(
    text=alt.condition(nearest, 'temperature:Q', alt.value(' '))
)
# Draw a rule at the location of the selection
rules = alt.Chart(tem).mark_rule(color='gray').encode(
    x='YEAR:Q',
).transform_filter(
    nearest
)

# Put the five layers into a chart and bind the data
alt.layer(
    line, selectors, points, rules, text
).properties(
    width=600, height=300)
Out[10]:

We can see the connection between temperature and CO2 emission¶

In [11]:
co2_temp = co2.merge(temp, left_on = 'Year', right_on = 'YEAR')[['YEAR','C02_Emission_Kt', 'ANNUAL']]

fig = go.Figure(data=[go.Scatter3d(
    x=co2_temp['YEAR'],
    y=co2_temp['C02_Emission_Kt'],
    z=co2_temp['ANNUAL'],
    mode='markers',
    marker=dict(
        size=2,
        color=co2_temp['ANNUAL'],                # set color to an array/list of desired values
        colorscale='amp',   # choose a colorscale
        opacity=0.8
    )
)])

#fig.update_layout(margin=dict(l=0, r=0, b=0, t=0))
fig.show()

Factors including excessive temperatures, acute monsoon system transformations, fragile glacial fed river systems, extreme weather events, and inundation of coastlines due to rising sea levels all will impact India’s existing water-based ecosystems. Agriculture is the most vulnerable sector to climate change, owing to its sensitivity to precise weather parameters and likely economic impacts. The changes in climatic events such as temperature, rainfall and atmospheric carbon dioxide (CO2) can significantly affect crop yields.

In [12]:
states_top_ten=dict()
crop = crop.dropna()
#vistual
for i in top_ten:
    wheat = crop[(crop['Crop'] == i)][['State_Name','Crop_Year','Production']]
    wheat = wheat.groupby(['State_Name','Crop_Year']).sum().reset_index()
#wheat = wheat.melt('Crop_Year', var_name = 'State_Name', value_name = 'Production')
    get_rid = wheat.groupby('State_Name').max()
    get_rid = get_rid[get_rid['Production']> np.mean((wheat.groupby('State_Name')['Production'].mean()))/2].index
    wheat = wheat.set_index('State_Name').loc[get_rid].reset_index()
    highlight = alt.selection(type='single', on='mouseover',
                          fields=['State_Name'], nearest=True)
    
    base = alt.Chart(wheat).encode(
    alt.X('Crop_Year:Q',scale=alt.Scale(zero = False)),
    alt.Y('Production:Q'),
    color='State_Name:N'
    )

    points = base.mark_circle().encode(
    opacity=alt.value(0)
    ).add_selection(
    highlight
    ).properties(
    width=600
    )

    lines = base.mark_line().encode(
    size=alt.condition(~highlight, alt.value(1), alt.value(3))
    )
    states_top_ten[i] = points + lines
#interact
def inter(crop = 'Coconut '):
    return states_top_ten[crop]
top_ten
Out[12]:
Index(['Coconut ', 'Sugarcane', 'Wheat', 'Jute', 'Rice', 'Cotton(lint)',
       'Mango', 'Potato', 'Paddy', 'Soyabean'],
      dtype='object', name='Crop')
In [13]:
interact(inter, x = top_ten);

interactive(children=(Text(value='Coconut ', description='crop'), Output()), _dom_classes=('widget-interact',)…

What happened to the cotton production in Gujarat and Maharashtra?¶

We notice that the cotton production in Gujarat frop tremendously from 2011 to 2012 and even disappeared after 2012. (Although this missingness is probably due to the incompleteness of given data)

In [14]:
Gujarat_2011 = crop[(crop['State_Name'] == 'Gujarat') &(crop['Crop'] == 'Cotton(lint)') & (crop['Crop_Year'] == 2011)]
Gujarat_2012 = crop[(crop['State_Name'] == 'Gujarat') &(crop['Crop'] == 'Cotton(lint)') & (crop['Crop_Year'] == 2012)]
Gujarat_2012.head()
Out[14]:
State_Name District_Name Crop_Year Season Crop Area Production
57515 Gujarat AHMADABAD 2012 Kharif Cotton(lint) 176800.0 302100.0
57864 Gujarat AMRELI 2012 Kharif Cotton(lint) 289300.0 242700.0
58103 Gujarat ANAND 2012 Kharif Cotton(lint) 7600.0 20800.0
58500 Gujarat BANAS KANTHA 2012 Kharif Cotton(lint) 45600.0 160700.0
58907 Gujarat BHARUCH 2012 Kharif Cotton(lint) 118900.0 271700.0

We can see that the production in most districts in Gujarat dropped a lot in two year¶

In [15]:
two_year = crop[(crop['State_Name'] == 'Gujarat') &(crop['Crop'] == 'Cotton(lint)') & ((crop['Crop_Year'] == 2011) | (crop['Crop_Year'] == 2012))]
X = list(map(lambda x:x[:6],(two_year['District_Name'].unique())))
Y = Gujarat_2011['Production']
Z = Gujarat_2012['Production']
  
X_axis = np.arange(len(X))
plt.figure(figsize=(20, 8))
plt.bar(X_axis - 0.2, Y, 0.4, label = '2011')
plt.bar(X_axis + 0.2, Z, 0.4, label = '2012')

plt.xticks(X_axis, X)
plt.xlabel("Districts")
plt.ylabel("Production")
plt.title("Production distribution")
plt.legend()

plt.show()
In [16]:
# Data process
effect = Gujarat_2011.merge(Gujarat_2012, on = 'District_Name')
effect['dif_prod'] = effect['Production_x'] - effect['Production_y']
effect['dif_area'] = effect['Area_x'] - effect['Area_y']
effect = effect[['District_Name', 'dif_area', 'dif_prod']]
effect['District_Name'] = effect['District_Name'].apply(lambda x: x[0]+x[1:].lower())
effect.loc[6,'District_Name'] = 'Dahod'
effect.loc[3,'District_Name'] = 'Banas Kantha'
effect.loc[14,'District_Name'] = 'Panch Mahals'
effect.loc[18,'District_Name'] = 'Sabar Kantha'
In [17]:
effect1 = effect.set_index('District_Name')['dif_prod']
m = folium.Map(location=[22.2587, 71.1924],tiles='cartodbpositron')#'CartoDB positron'
inp = effect1.to_numpy()
effect1 = pd.Series(inp, index = effect1.index)
folium.FitBounds([(22.2587-2,71.1924-2), (22.2587+2,71.1924+2)]).add_to(m)

import branca

colorscale = branca.colormap.linear.RdYlBu_08.to_step(data=inp, quantiles=np.arange(0,1,0.001)).scale(-150000, 129400.0)
def style_function(feature):
    popu = effect1.get((feature['properties']['NAME_2']), None)
    #print(feature['properties']['NAME_2'])
    #print(popu)
    return {
        "fillOpacity": 0.6,
        "weight": 0,
        "fillColor": "#black" if popu is None else colorscale(popu),
    }

with open(os.path.join('gujarat.geojson')) as f:
    location = json.load(f)
folium.GeoJson(
    location,
    style_function=style_function,
).add_to(m)
Out[17]:
<folium.features.GeoJson at 0x7fcbdfbe2040>

Blue: Decrease in production¶

Yellow or light blue: not increase or decrease that much¶

Red: Increase in production¶

In [18]:
m
Out[18]:
Make this Notebook Trusted to load map: File -> Trust Notebook

The production in eastern part tends to grow by a little bit while the production in other area drops a lot¶

In [19]:
effect2 = effect.set_index('District_Name')['dif_area']
m = folium.Map(location=[22.2587, 71.1924],tiles='cartodbpositron')#'CartoDB positron'
inp = (effect2.to_numpy())
effect1 = pd.Series(inp, index = effect1.index)
folium.FitBounds([(22.2587-2,71.1924-2), (22.2587+2,71.1924+2)]).add_to(m)
import branca

colorscale = branca.colormap.linear.RdYlBu_08.to_step(data=inp, quantiles=np.arange(0,1,0.001)).scale(-110200,119400)
def style_function(feature):
    popu = effect1.get((feature['properties']['NAME_2']), None)
    #print(feature['properties']['NAME_2'])
    #print(popu)
    return {
        "fillOpacity": 0.6,
        "weight": 0,
        "fillColor": "#black" if popu is None else colorscale(popu),
    }

with open(os.path.join('gujarat.geojson')) as f:
    location = json.load(f)
folium.GeoJson(
    location,
    style_function=style_function,
).add_to(m)
Out[19]:
<folium.features.GeoJson at 0x7fcbdf821d90>

Blue: Decrease in Area¶

Yellow or light blue: not increase or decrease that much¶

Red: Increase in Area¶

In [20]:
m
Out[20]:
Make this Notebook Trusted to load map: File -> Trust Notebook

The declination resulted from multiple reasons:¶

  • climate: delayed monsoons,
  • high cost of cultivation,
  • market: low market prices
  • policy: inconsistent export policy

This is the prices and rate of change data of Soybean and Cotton since 2007¶

In [21]:
cs
Out[21]:
Month Cotton Price (Indian Rupee per Kilogram) Soybeans Price (Indian Rupee per Metric Ton) Cotton ROC Soybeans ROC Cotton / Soybeans Price Ratio
0 2007-04-01 53.11 13477.80 - - 0.0039
1 2007-05-01 49.77 13624.53 -0.063 0.0109 0.0037
2 2007-06-01 54.62 14715.08 0.0976 0.08 0.0037
3 2007-07-01 60.62 15176.34 0.1099 0.0313 0.0040
4 2007-08-01 60.00 15714.78 -0.0103 0.0355 0.0038
... ... ... ... ... ... ...
176 2021-12-01 200.18 41859.34 -0.0367 0.0199 0.0048
177 2022-01-01 216.64 45132.08 0.0823 0.0782 0.0048
178 2022-02-01 228.85 49643.56 0.0563 0.1 0.0046
179 2022-03-01 237.12 54941.77 0.0361 0.1067 0.0043
180 2022-04-01 260.54 54911.05 0.0988 -0.0006 0.0047

181 rows × 6 columns

In [22]:
cs1 = cs.iloc[1:][['Month','Cotton ROC', 'Soybeans ROC']].melt('Month', var_name = 'Crops', value_name = 'ROC')
cs1
Out[22]:
Month Crops ROC
0 2007-05-01 Cotton ROC -0.063
1 2007-06-01 Cotton ROC 0.0976
2 2007-07-01 Cotton ROC 0.1099
3 2007-08-01 Cotton ROC -0.0103
4 2007-09-01 Cotton ROC 0.0085
... ... ... ...
355 2021-12-01 Soybeans ROC 0.0199
356 2022-01-01 Soybeans ROC 0.0782
357 2022-02-01 Soybeans ROC 0.1
358 2022-03-01 Soybeans ROC 0.1067
359 2022-04-01 Soybeans ROC -0.0006

360 rows × 3 columns

In [23]:
# Create a selection that chooses the nearest point & selects based on x-value
nearest = alt.selection(type='single', nearest=True, on='mouseover',
                        fields=['Month'], empty='none')
# The basic line
#highlight = alt.selection(type='multi', on='mouseover',
#                          fields=['Month'], nearest=True)

line = alt.Chart(cs1).mark_line().encode(
    alt.X('Month:T'),
    alt.Y('ROC:Q',scale = alt.Scale(zero = False)),
    color='Crops:N',
    size=alt.condition(~nearest, alt.value(1.6), alt.value(12))
)
# Transparent selectors across the chart. This is what tells us
# the x-value of the cursor
selectors = alt.Chart(cs1).mark_point().encode(
    x='Month:T',
    opacity=alt.value(0),
).add_selection(
    nearest
)
# Draw points on the line, and highlight based on selection
points = line.mark_point().encode(
    opacity=alt.condition(nearest, alt.value(1), alt.value(0))
)
# Draw text labels near the points, and highlight based on selection
text = line.mark_text(align='left', dx=5, dy=-5).encode(
    text=alt.condition(nearest, 'ROC:Q', alt.value(' '))
)
# Draw a rule at the location of the selection
rules = alt.Chart(cs1).mark_rule(color='gray').encode(
    x='Month:T',
).transform_filter(
    nearest
)

# Put the five layers into a chart and bind the data
alt.layer(
    line, selectors, points, rules, text
).properties(
    width=630, height=300)
Out[23]:

We can see that around the same period from 2011 to 2012, the price of cotton slumped by more than 40%. That's partly why the production also decreased¶

Of course there are many other reasons, but this is definitely the most significant one.

Some suggestions¶

  1. Strengthening the policy-relevant data base: Policy makers should have access to the most relevant, reliable, and timely information about policy options and expected impact of each option on the various stakeholder groups. An improved database, timely market information, including reliable forecasts in developing countries, would greatly facilitate policy design and implementation.
  2. Appropriate use of trade policy and careful interference in price signals: World market should be able to dampen price volatility in the domestic market.
  3. Assure that appropriate price information are sent to domestic producers and consumers. For example, some knowledge of market saturation should be given to the producers so that farmers can act early. It's clear the price volatility was skyrocketing around and before 2011.
  4. Facilitating effective risk management tools: More effective risk management tools are needed by farmers, traders and consumers to cope with extreme weather events, global warming and market fluctuations. Examples are better market forecasts, timely dissemination of weather-related information, and research to develop risk-reducing technology—such as drought- and flood-tolerant and pest-resistant crops.
  5. When such event happens, targeted compensatory measures could be used to protect specific groups, such as poor consumers or producers.

If you are interested¶

https://www.downtoearth.org.in/news/farmers-shun-cotton-38429 https://economictimes.indiatimes.com/news/economy/agriculture/cotton-productivity-on-decline-across-india/articleshow/16667755.cms?from=mdr

all reported huge decline of cotton production in India¶

References¶

  • [1]Devendra, C. “Rainfed areas and animal agriculture in Asia: the wanting agenda for transforming productivity growth and rural poverty.” Asian-Australasian journal of animal sciences vol. 25,1 (2012): 122-42. doi:10.5713/ajas.2011.r.09
  • [2]Majra, J P, and A Gur. “Climate change and health: Why should India be concerned?.” Indian journal of occupational and environmental medicine vol. 13,1 (2009): 11-6. doi:10.4103/0019-5278.50717
  • [3]https://mospi.gov.in/documents/213904/416359//Press%20Note_31-05-2021_m1622547951213.pdf/7140019f-69b7-974b-2d2d-7630c3b0768d
  • [4]See: Assessment of Climate Change Over the Indian Region, June 2020.
In [ ]: