Dataset: CO2 Emissions Around the World

● Data Source
Kaggle:
Dataset: CO2 Emissions Around the World
Data Source:
https://www.kaggle.com/datasets/koustavghosh149/co2-emission-around-theworld?resource=download
Dataset: World Population by Countries Dataset (1960-2021)
Data Source:
https://www.kaggle.com/datasets/kaggleashwin/population-dataset
Dataset: World GDP (GDP, GDP per capita, and annual growths)
Data Source:
https://www.kaggle.com/datasets/zgrcemta/world-gdpgdp-gdp-per-capita-and-annual-growths
Kaggle is the largest data scientist community in the world and was founded in 2010. Its primary
function is to act as a platform for programmers and data scientists to host machine learning
challenges, data visualization, etc. Additionally, Kaggle offers a public data platform where users can
access its datasets for data exploration and model creation.
● Data Description
The CO2 emissions database is the main dataset for this project, includes per capita co2 emissions
for 216 countries from 1990 to 2019.The main variables include country name, country code, region,
Indicator Name, and years(1990-2019).The team will use the above variables for subsequent data
processing and visualization analysis of co2 emissions per capita for each country and region.
The population dataset and the GDP dataset will be used as secondary datasets to assist the team in
their exploration of the causes of changes in per capita co2 emissions. Both datasets include the
same country information as the above dataset, one includes the size of each country’s population,
and the other includes the GDP of each country. The time interval is from 1990 to 2019.Variables in
both dataset include country name, country code and year(1990-2019).
● Background and Visualization Purpose
Background
CO2 is an important component of air and has an important impact on many factors in life. In recent
years CO2 emissions have become one of the key topics of concern in many countries, as excessive
CO2 emissions affect not only environmental change but also human life, so it is increasingly
important to reduce CO2 emissions while reducing the impact on other factors.
Purpose
In order to better understand how CO2 emissions have changed in recent years in different countries
and regions, and to understand the impact of measures implemented by countries in the face of
environmental pressures, we have downloaded relevant data sets from Kaggle and used data
visualisation tools to explore and analyse the impact of changes in population and GDP on CO2
emissions, and to present our understanding and proposals.
● Data Visualization Tools
Tableau
In Tableau, first of all, we use the data of different countries of CO2 emissions and GDP to generate
a map to see an overall condition as the year past from 1981 to 2019, then we will then use
histograms to show the ranking of CO2 emissions for different countries, after which we will create a
graph of the growth rate of GDP per capita, a comparison of linear growth rates for different countries
and regions to see how CO2 emissions are affected by different regions, and finally, we will
summarise them and form a story to show our conclusions.
R
In R, our initial intention is to use scatter plots and the linear regression to look at the distribution of
CO2 emissions, GDP and population, to fit smoothed trend lines and to compare data across
countries to provide data to support further analysis, and then we may select some representative
data for more detailed comparisons and create graphs such as pie charts for detailed analysis. Our
aim is to identify the most relevant variables to predict the global carbon emissions.
● Analysis steps
1.Determine the direction of the problem
The problem-side team has determined to explore the factors that influence per capita co2 emissions
in each country, and how that data changes over time. For the impact factor, due to the size of the
population has a strong influence on co2 emissions and the value of a country’s GDP is an important
indicator for assessing the level of the economy, so changes in country’s population and GDP are
also important points of analysis.
2.Data preparation
After deciding on the direction, the team found three public datasets from Kaggle corresponding to
co2 emissions per capita population and GPD respectively. All datasets were uploaded to the Kaggle
platform within 1 year, and these three datasets will provide data support for the subsequent analysis
steps.
3.Data pre-processing
Select the data we need such as CO2 emissions statistics for different countries and regions from
1981-2019, population and GDP statistics for the same years, remove redundant data and unwanted
data, clean the data and do preliminary filtering
4.Data visualization
Use Tableau, R to visualise data, create scatter charts to analyse the relationship between two
variables, create line charts to analyse growth rates, create histogram charts to compare data
between countries and consolidate individual charts
5. Summary of findings and analysis
Identify representative countries, find information to learn more about the country in that year, analyze
the possible economic and political measures behind the data, and get a preliminary analysis of the
country’s CO2 emissions and the group’s recommendations.
● Initial findings
Tableau
Through Tableau, we conducted an initial exploration of the data, using the median as a
measure, and had the following findings:
1) The US, China and Western Europe are the three largest economies. The US has the
largest CO2 emission per capita, with China being the second largest, and Western Europe
being the smallest.
2) The Top3 countries of CO2 emission per capita are Qatar, United Arab Emirates and
Bahrain; the Bottom3 countries of CO2 emission per capita are Burundi, Congo, Dem. Rep.
and Central African Republic.
3) In terms of total CO2 emission, China is in first place, the US is second, and the Russian
Federation is third.
R
Through R, we obtained that the 99th percentile of each country’s 30-year average per capita
CO2 value was 22.94559007, and the 1st percentile was 0.04973599.
We screened out the countries whose average per capita CO2 values are more significant
than 99th percentile and less than 1st percentile and plotted the development trend of their
per capita CO2 emissions over time.
According to the picture, it can be concluded as follows:
1) Qatar and the United Arab Emirates are the two countries with the highest annual average
CO2 emissions per capita over 30 years, while Burundi and Congo, Dem. Rep are the two
countries with the lowest average yearly CO2 emissions per capita.
2) As the country with the highest per capita CO2 emissions, Qatar’s per capita CO2
emissions from 1990 to 2019 showed a trend of rising first and then declining, reaching a
peak in 2000-2005.
3) The two countries with higher per capita CO2 emissions showed a downward trend in
recent years.
4) The two countries with lower CO2 emissions per capita have fluctuating trends, reaching a
small peak between 2010-2015 and increasing in recent years.