I. DOCCUMENTARY REVIEW

PROBLEMS • What were the main sources of CO₂ emissions before the COVID-19 pandemic? • How did lockdowns and health restrictions influence the reduction of CO₂ emissions in 2020? • Which economic sectors were most impacted by the decline in emissions during the pandemic? • Did CO₂ emission levels rebound after the health crisis in 2021? • What lessons can be learned from this period for transitioning to a low-carbon economy

Objectives of the Study Objective 1 : To identify and analyze the main anthropogenic and environmental factors driving CO₂ emissions in selected countries before, during, and after the COVID-19 pandemic.

Objective 2 : To assess the environmental and socio-economic impacts of CO₂ emissions across different regions and time periods.

Objective 3 : To evaluate the influence of the COVID-19 pandemic on the evolution of CO₂ emission patterns, highlighting any significant shifts or trends.

Objective 4 : To determine the statistical correlations between key variables such as urban population, density, temperature, precipitation, pollution index, and annual CO₂ emissions using Principal Component Analysis (PCA).

Objective 5 : To understand the environmental impact of the COVID-19 pandemic, particularly in relation to emission reductions and ecosystem responses. KEYWARDS AND DEFINITION

Comparative Study: An analytical approach that examines similarities and differences between multiple subjects. In this context, it refers to the comparison of CO₂ emissions and their drivers and impacts across different countries and time periods (before, during, and after COVID-19). Causes of CO₂ Emissions: The primary factors responsible for the release of carbon dioxide into the atmosphere, including industrial activities, transportation, deforestation, urbanization, and energy consumption.[1]

Impacts of CO₂ Emissions: The consequences of elevated CO₂ levels on the environment and society, such as climate change, air pollution, public health issues, and economic disruptions. Global Scale: The study covers multiple countries across different continents, aiming to capture both universal trends and regional specificities in CO₂ emissions and their effects.

COVID-19 Pandemic: A global health crisis caused by the SARS-CoV-2 virus, which began in late 2019 and significantly disrupted human activities worldwide. It serves as a key variable in this study due to its impact on economic activities and environmental indicators.

Before, During, and After COVID-19: A temporal framework used to observe the evolution of CO₂ emissions and related factors. “Before” refers to pre-pandemic conditions (2019), “during” to the peak years of disruption (2020), and “after” to the initial recovery phase (2021).

ASSESSMENT OF GLOBAL CO₂ EMISSIONS

This report provides an updated assessment of global carbon dioxide (CO₂) emissions trends up to 2011, extending last year’s analysis. The focus is on the changes in annual emissions from 2010 to 2011. Unlike the BP reports, which concentrate on fossil fuel combustion, this study incorporates a broader range of CO₂ emission sources. These include flaring of waste gas during oil production, cement clinker production, other limestone uses, non-energy fuel uses (such as feedstocks), and several smaller sources.[2]

CHANGES IN EMISSIONS BY COUNTRY

China, now firmly within the 6 to 19 tonnes per capita range[3], comparable to major industrial powers, reflects the ongoing scale of industrialization. During the study period, carbon emissions from transportation increased in the United States, China, India, Canada, Russia, and Brazil, while they decreased in Japan. This suggests that, in most of these countries, energy conservation strategies were not effectively implemented.

FACTORS INFLUENCING EMISSION INCREASES OR REDUCTIONS

In most countries, reductions in CO₂ emissions were primarily due to the decrease in carbon intensity, meaning more efficient energy use and the adoption of cleaner technologies. In contrast, emission increases were mainly driven by the structure of electricity generation (more or less dependent on fossil fuels) and economic growth. Therefore, to achieve sustainable emission reductions, optimizing energy structures and limiting the growth of private vehicle ownership are crucial.[4]

CASE STUDY: PHOENIX, ARIZONA

Recent research has highlighted a CO₂ “dome” in the Phoenix, Arizona metropolitan area, where CO₂ concentrations peak near the city center and are 50% higher than in the surrounding rural areas. This phenomenon is attributed to anthropogenic sources and the region’s physical geography. Human-related CO₂ emission data were collected from various government and NGO sources. Additionally, soil CO₂ efflux was measured across the dominant land-use types. Over 80% of emissions came from human activities, primarily automobile traffic. Natural desert ecosystems showed minimal emissions during hot, dry periods but responded rapidly to moisture. In contrast, human maintained vegetation types (such as golf courses, lawns, and irrigated agriculture) exhibited higher CO₂ efflux, which was both temperature and soil moisture dependent.[5]

COVID-19 AND THE IMPACT ON CO₂ EMISSIONS

Before the Pandemic Before COVID-19, global CO₂ emissions increased by about 1% per year, with significant growth in the transportation sector, while renewable energy was developing alongside fossil fuels.

Reduction in Emissions During Lockdowns Due to global lockdowns, daily global CO₂ emissions dropped by 17% in April 2020, mainly due to reduced transportation and industrial activities. The surface transportation sectorcontributed to nearly half of this reduction.[6]

IMPACT ON ANNUAL EMISSIONS

The reduction in annual emissions depended on the duration of the lockdowns. If conditions had returned to normal by mid-June, the decrease would have been around 4%. However, if restrictions had continued until the end of the year, the annual reduction could have reached 7%.[7]

Rebound After the Pandemic After the lockdowns were lifted, emissions rebounded, particularly in China and the United States, where economies restarted quickly, leading to a resurgence in CO₂ emissions.

Role of Post-COVID Policies Post-COVID economic policies, such as investments in renewable energy and the energy transition, will be crucial in determining whether the emission reductions observed during the pandemic can lead to lasting changes.

LESSONS FOR THE FUTURE

The pandemic showed that it is possible to temporarily reduce emissions, but for lasting impact, global policies aimed at decarbonization and reducing reliance on fossil fuels are essential.

CONSEQUENCES OF CO₂ EMISSIONS ON TEMPERATURE AND PRECIPITATION

The accumulation of CO₂ in the atmosphere leads to a global rise in temperatures, resulting in intensified heatwaves, ice melt, disruption of ecological balances, and increased evaporation of water resources. This dynamic heightens the vulnerability of both natural and human systems, particularly in arid or densely populated areas.

At the same time, precipitation patterns become more irregular and extreme. Some regions experience prolonged droughts, while others face intense and destructive rainfall events. These disruptions compromise food security, weaken water resources, degrade soils, and increase the risk of flooding.

These climate transformations trigger a cascade of impacts, affecting agriculture, biodiversity, water resources, infrastructure, and public health. They worsen food insecurity, intensify resource-related tensions, and increase the vulnerability of populations to climate-related hazard.[8]

The COVID-19 pandemic provided a unique insight into CO₂ emissions dynamics, highlighting their main causes and impacts. While lockdowns temporarily reduced emissions, the decline was not structural. However, the crisis demonstrated that rapid changes are possible, particularly through remote work and reduced travel. To achieve a lasting transition, investments in renewable energy and low-carbon policies are essential. This period should serve as a catalyst for accelerating the necessary transformations to combat climate change.

#II.TOOLS USED

KOBO TOOLBOX Kobo Toolbox is a suite of open-source tools designed to simplify the data collection process. It enables users to create and deploy forms and to collect data using various channels, such as webforms, mobiles, and tablets. It supports offline data collection, enabling users to collect data in remote areas without internet connectivity. Kobo Collect is the mobile app component of Kobo Toolbox and allows users to collect data on Android devices. It supports multiple languages, multimedia files, and collection of geospatial data. It helps monitor data .

RSTUDIO RStudio is an integrated development environment (IDE) for the R and Python programming languages. It offers a console, a syntax-highlighting editor, tools for direct code execution, as well as features for tracing, history, debugging, and workspace management. RStudio is available in both open-source and commercial versions, running on Windows, Mac, and Linux. Zotero Zotero is a free and open-source reference management software. It allows users to collect, organize, annotate, cite, and share research. Available for Mac, Windows, Linux, and iOS, Zotero facilitates the quick creation of bibliographies and provides automatic citation features in various styles.[9]

QGIS QGIS (Quantum Geographic Information System) is a free and open-source Geographic Information System (GIS) software. It enables users to create, edit, visualize, analyze, and publish geospatial information. QGIS is valued for its flexibility and support for multiple data formats, including raster and vector formats. MICROSOFT EXCEL Microsoft Excel is a spreadsheet program used to store, organize, and manipulate data. It represents data in tables composed of rows and columns, where each cell can contain different types of data, including text and numbers. Excel allows users to perform calculations using built-in formulas and functions, facilitating financial analysis, budgeting, and expense track #III.JUSTIFICQTION OF VARIABLES URBAN POPULATION The growth of the urban population is closely linked to increased economic activity, transportation, and energy consumption. This variable helps to understand how urbanization contributes to rising CO₂ emissions, especially in large cities.

DENSITY (POPULATION DENSITY) High population density can lead to a concentration of emission sources (transportation, buildings, industries). It also influences pressure on infrastructure and resources. This variable helps assess whether more densely populated areas emit proportionally more CO₂.

CO₂ EMISSIONS (TOTAL CO₂EMISIONS) This is the study’s primary indicator. It tracks the overall evolution of CO₂ emissions in each country, comparing the periods before, during, and after the pandemic. It is essential for identifying general trends.

POLLUTION INDEX This indicator reflects air quality by taking into account several pollutants, including CO₂. It helps link emissions to environmental and health impacts, complementing purely quantitative data.

PRECIPITATION Climate change induced by greenhouse gases can disrupt rainfall patterns. This variable examines whether changes in CO₂ emissions are associated with observed variations in precipitation.

TEMPERATURE CO₂ is a greenhouse gas contributing to global warming. Studying temperature trends helps to better understand the potential climatic impacts of emissions and to identify long-term correlations.[10]

CO₂ BY TRANSPORT Transportation is one of the largest contributors to CO₂ emissions, especially in urban areas. This variable measures the impact of mobility restrictions during the pandemic on emissions from this sector.[11]

CO₂ BY AGRICULTURE Agriculture contributes to emissions through deforestation, the use of nitrogen-based fertilizers, and machinery. This variable is key to analyzing a sometimes overlooked yet significant sector, particularly in countries with strong agricultural activity.

CO₂ BY INDUSTRY Industry is a major source of CO₂ due to its intensive use of fossil fuels. This variable helps evaluate the effects of industrial slowdown during the pandemic and the subsequent economic recovery.

CO₂ BY BUILDING This variable includes emissions from residential and commercial energy use (heating, air conditioning, electricity). It is relevant to understanding the effects of lockdowns on residential energy consumption.

Conclusion:These ten variables were selected to allow a comprehensive analysis combining demographic, environmental, and sectoral factors. They enable a detailed understanding of the causes of emissions, their temporal variations, and their environmental impacts in connection with the COVID-19 pandemic

data = read.csv(file ="Variables2020.csv", header = TRUE, sep = ";", quote = "\"",
                dec = ",", row.names = 1)
data[,1:10]

##       Urb.Pop      Dens CO2.Emiss air_poll_CO     Precip  Temp.var Trans_CO2
## BFA   6573983  78.50399   5377314    967355.7  700.36096 0.3413738    760000
## CIV  14951022  90.92909  11015523   1588104.6 1150.19200 0.4674681   4270000
## EGY  46768290 109.81477 235357630   4195928.5   26.94142 0.1101767  55810000
## FRA  54739999 120.36240 280490020   2434855.0 1137.01400 1.0604340 108790000
## DEU  64410589 239.35634 648356860   2439712.2  820.91364 1.1459798 146510000
## NGA 111188136 234.96182 129497090  23092710.0 1000.73180 0.1978370  52080000
## SEN   8079308  49.92406  10820952    358039.9  688.11395 0.6394126   3270000
## ZAF  40791186  37.10731 434066530  13497122.0  521.36444 0.2706180  45430000
##     Agri_CO2 Indus_CO2 CO2.build
## BFA   650000   2190000   1560000
## CIV  6480000   3990000   2520000
## EGY 22450000  28830000  17200000
## FRA 71610000  21310000  60650000
## DEU 58050000  23740000 127250000
## NGA 79570000  28170000  42450000
## SEN  1320000     10000   4550000
## ZAF 29880000  24200000  18840000

library(corrplot)

## Warning: le package 'corrplot' a été compilé avec la version R 4.4.3

## corrplot 0.95 loaded

library(psych)

## Warning: le package 'psych' a été compilé avec la version R 4.4.3

library(Hmisc)

## Warning: le package 'Hmisc' a été compilé avec la version R 4.4.3

## 
## Attachement du package : 'Hmisc'

## L'objet suivant est masqué depuis 'package:psych':
## 
##     describe

## Les objets suivants sont masqués depuis 'package:base':
## 
##     format.pval, units

mat_cor = cor(data)
col = colorRampPalette(c("#BB4444", "#EE9988", "#FFFFFF", "#77AADD", "#4477AA"))
corrplot(mat_cor, method="color", col=col(200),  
         type="upper", order="hclust", 
         addCoef.col = "black", # Ajout du coefficient de corrélation
         tl.col="black", tl.srt=518, #Rotation des étiquettes de textes
         , sig.level = 0.1, insig = "blank", 
         # Cacher les coefficients de corrélation sur la diagonale
         diag=FALSE)

A correlation matrix is a table displaying correlation coefficients, which measure the strength and direction of relationships between pairs of variables. These coefficients range from -1 to 1 and help determine how variables are related: • Positive Correlation: When one variable increases, the other also increases, and when one decreases, the other follows. The closer the coefficient is to +1, the stronger the relationship. Notable correlations are highlighted in red when greater than 0.5. • Negative Correlation: When one variable increases, the other decreases, indicating an inverse relationship. A coefficient closer to -1 signifies a strong negative correlation. These are marked in dark blue when less than -0.5. • Zero or Weak Correlation: When changes in one variable do not significantly affect the other, the correlation is weak or nonexistent. These values, typically near zero, appear in light colors such as white or pale blue.

• By comparing these correlation matrices, we observe that the same variables remain strongly correlated over time. Specifically, there is a strong positive correlation between urban population and industry, air pollution and industry, as well as air pollution and urban population. • In general, most variables exhibit a high correlation, and there is no strong negative correlation. • Finally, in 2020, the correlation rate of transport with other variables decreased, likely due to the effects of lockdowns related to the pandemic.

# Inertia and choice of main axes
# Launching the PCA

library(FactoMineR)

## Warning: le package 'FactoMineR' a été compilé avec la version R 4.4.3

library(ggplot2)

## Warning: le package 'ggplot2' a été compilé avec la version R 4.4.3

## 
## Attachement du package : 'ggplot2'

## Les objets suivants sont masqués depuis 'package:psych':
## 
##     %+%, alpha

library(factoextra)

## Warning: le package 'factoextra' a été compilé avec la version R 4.4.3

## Welcome! Want to learn more? See two factoextra-related books at https://goo.gl/ve3WBa

pca_1 = PCA(X = data, scale.unit = TRUE, ncp = 10, ind.sup = NULL, 
            quanti.sup = NULL, quali.sup = NULL, row.w = NULL, 
            col.w = NULL, graph = FALSE, axes = c(1,2))

fviz_eig(pca_1, addlabels=TRUE, hjust = -0.3) +
  ylim(0, 65)

Comparative Analysis of the First Three Principal Components

A Principal Component Analysis (PCA) was conducted over three consecutive years. The cumulative percentages of variance explained by the first three principal components are as follows:

Year 2019: 86.2%

Year 2020: 90.0%

Year 2021: 90.6%

These results indicate that a significant portion of the information is concentrated within the first three dimensions, with a gradual improvement in representativeness from one year to the next. In 2019, the variance is more dispersed, reflecting a more complex structure. In contrast, the years 2020 and 2021 show a stronger concentration, suggesting a more stable and coherent structure.This trend, although counterintuitive for 2020 a year marked by the COVID-19 pandemic may be explained by the global homogenization of dynamics (lockdowns, reduced activities), which likely reduced variability across countries and strengthened the overall coherence of the data.

CERCLE DE CORRELATION

fviz_pca_var(pca_1, col.var = "cos2" , gradient.col = c("blue" , "green" , "red"), repel = TRUE )

In 2019, the variables were generally correlated. Population density contributed to the formation of the second axis, indicating its influence on the spatial variability of the data.

In 2020, the Covid effect disrupted this structure: the transport variable was no longer correlated, likely due to lockdowns. Density no longer contributed to the second axis, reflecting a reduced explanatory role due to the widespread decline in activity.

In 2021, there was a partial return to the pre-Covid structure: density once again contributed to axis 2. However, a new pattern emerged—temperature became correlated with precipitation, which was not the case in 2019. This may be explained by climate change or specific weather conditions that year.

library(FactoMineR)
pca_1 = PCA(X = data, scale.unit = TRUE, ncp = 10, ind.sup = NULL, 
            quanti.sup = NULL, quali.sup = NULL, row.w = NULL, 
            col.w = NULL, graph = FALSE, axes = c(1,2))

fviz_pca_ind(pca_1, col.ind = "contrib", gradient.cols = c("blue" , "green" , "red"), repel = TRUE)

#III. Detailed Analysis of the 2019 Figure (Before COVID-19)

Country Distribution

CHN (China) is positioned far to the right on Dim1, with a very strong contribution (red color), which means it stands out significantly from other countries, likely due to high values in variables such as CO₂ emissions, urban density, or urban population. USA is also distant, though not as extreme as China.

African countries (CIV, NGA, SEN, ZAF, EGY, BF…) are clustered on the left and near the center, suggesting they have relatively low and similar values for the studied variables.

France (FRA) and Germany (DEU) are higher on Dim2, which may reflect a different urban or environmental profile.

Interpretation of the Dimensions

Dim1 (62.7%) likely captures the overall level of emissions or industrialization (countries further to the right are likely higher emitters).

Dim2 (17%) captures finer distinctions, potentially related to urban management or infrastructure. II. Detailed Analysis of the 2020 Figure (Year of COVID-19)

Observed Changes

CHN remains far to the right on Dim1 with a strong contribution (in red), meaning it continues to stand out in terms of emissions or industrial activity.

USA has shifted further to the right on Dim1 and significantly upward on Dim2, indicating a more unique profile in 2020 compared to 2019.

African countries remain grouped, though some (like EGY, NGA) moved slightly downward.

France and Germany show slight dispersion but are still close together.

Possible Impacts of COVID-19 Here’s how the pandemic may have influenced the PCA distribution:

Sharp Drop in CO₂ Emissions in Industrialized Countries In 2020, lockdowns led to a significant drop in economic activities, transportation, and industry.This likely caused lower CO₂ emissions in countries like the USA and Europe, altering their PCA positions. China’s Quick Recovery

Although the outbreak started in China, the country resumed industrial activity quickly, by mid-2020.

This fast rebound may explain why China remained strongly positioned on Dim1, even more than in 2019.

African Countries: Minimal Change

Many African countries have lower levels of industrialization and thus less variation in CO₂ emissions.

This explains why they remain clustered near the center, showing little shift during the pandemic year.

USA: A Combination of Crises The US experienced a severe health crisis, economic slowdowns, and lifestyle shifts (remote work, reduced mobility).

This could explain the sharp rise on Dim2, possibly related to internal environmental or social changes.

Interpretive Conclusion

The year 2020 represents a break in emissions and urban dynamics due to the COVID-19 pandemic. China stood out by restarting its industrial activities quickly, maintaining its environmental impact, while the USA and some European countries experienced slowdowns and lifestyle changes that altered their profiles. African countries, with limited industrialization, remained relatively stable throughout this period.

The graph of the contributions of variables and individuals to dimensions 1 and 2 shows that the contributions are different for each variable and individual. To interpret this, we can analyze which variables contribute the most to each dimension. High contributions indicate that these variables are important in defining the corresponding axis. Similarly, for individuals, those with higher contributions are more influential in shaping the principal components. By examining the distribution, we can identify patterns, such as clusters of similar individuals or key variables driving the differentiation along the dimensions.

CONTRIBUTIONS OF INDIVIDUALS AND VARIABLES

fviz_contrib(pca_1, choice = "ind", axes = 1)

fviz_contrib(pca_1, choice = "var", axes = 1)

fviz_contrib(pca_1, choice = "var", axes = 2)

DENDROGRAM

# Classification of individuals

res.PCA<-PCA(data,graph=FALSE)
res.HCPC<-HCPC(res.PCA,nb.clust=3,consol=FALSE,graph=FALSE)
plot.HCPC(res.HCPC,choice='tree',title='Arbre hiérarchique')

Hierarchical Tree (Classification) • The tree groups countries into clusters. • There appear to be well-defined groups, likely: o Industrialized countries with high CO₂ emissions (USA, CHN, DEU). o Developing countries with lower emissions (BFA, SEN). o An intermediate group (EGY, ZAF, CIV).

Conclusion:

• Your PCA highlights a strong structuring of countries based on their level of pollution and industrialization. • Dim 1 primarily separates high-emission countries from low-emission ones. • Dim 2 could be related to climatic and environmental factors

plot.HCPC(res.HCPC,choice='map',draw.tree=FALSE,title='Plan factoriel')

plot.HCPC(res.HCPC,choice='3D.map',ind.names=FALSE,centers.plot=FALSE,angle=60,title='Arbre hiérarchique sur le plan factoriel')

REMARK The graph of individuals provides an overview of the similarities and differences between individuals. Then we can see the positions of the following countries: USA and CHINA located at the extreme right of the Dim 1,2 axis, it has very different characteristics from the other countries for this dimension; This graph helps identify groups of countries with marked similarities or differences along the main dimensions of the ACP. We find that USA, CHINA behave as atypical elements. For a better result, we removed them. After removing these countries from our data, we obtain the graph of individuals After a comparative study, we observed the impact of COVID-19 on our data. To achieve a better Principal Component Analysis (PCA), we decided to exclude China and the United States to prevent a single variable from dominating a class. We will resume the study to ensure a proper grouping and a relevant cartographic representation. For this part, we will focus on the year 2020.

res.pca <- PCA(data, scale.unit = TRUE, ncp = 5, graph = TRUE)

valeur.propre <- get_eigenvalue(res.pca)
valeur.propre

##        eigenvalue variance.percent cumulative.variance.percent
## Dim.1 5.545589997      55.45589997                    55.45590
## Dim.2 2.301704045      23.01704045                    78.47294
## Dim.3 1.441519205      14.41519205                    92.88813
## Dim.4 0.408683795       4.08683795                    96.97497
## Dim.5 0.222975865       2.22975865                    99.20473
## Dim.6 0.071898386       0.71898386                    99.92371
## Dim.7 0.007628708       0.07628708                   100.00000

fviz_eig(res.pca, 
         addlabels = TRUE,         # Ajouter les pourcentages
         ylim = c(0, 75),          # Échelle verticale
         barfill = "skyblue",      # Couleur des barres
         barcolor = "black",       # Contour
         linecolor = "red")        # Ligne du coude (critère de Kaiser)

• The first principal component (Dim1) explains 55.5% of the data variability. • The second component (Dim2) explains 23% of the variability. • Together, the first two dimensions explain 78.5% of the total information, which is very high. • This means that the data can be effectively summarized using only two axes instead of the initial 10 variables. • CO2.Emiss and Trans_CO2 have a very strong correlation (0.86), indicating they tend to move together. • Temp.var shows a negative correlation with some variables like Precip (-0.47), meaning that higher temperature variability is associated with lower precipitation. • Urb.Pop and Indus_CO2 are strongly correlated (0.91), suggesting that urban growth is linked to industrial CO2 emissions.

mat_cor = cor(data)
col = colorRampPalette(c("#BB4444", "#EE9988", "#FFFFFF", "#77AADD", "#4477AA"))
corrplot(mat_cor, method="color", col=col(200),  
         type="upper", order="hclust", 
         addCoef.col = "black", # Ajout du coefficient de corrélation
         tl.col="black", tl.srt=518, #Rotation des étiquettes de textes
         , sig.level = 0.1, insig = "blank", 
         # Cacher les coefficients de corrélation sur la diagonale
         diag=FALSE)

library(corrplot)
library(psych)
library(Hmisc)

Colors: • Dark blue → Strong positive correlation (close to 1). • Red → Negative correlation (close to -1). • White → Weak correlation (close to 0). • Examples: • CO2.Emiss and Trans_CO2 have a very strong correlation (0.86), indicating they tend to move together. • Temp.var shows a negative correlation with some variables like Precip (-0.47), meaning that higher temperature variability is associated with lower precipitation. • Urb.Pop and Indus_CO2 are strongly correlated (0.91), suggesting that urban growth is linked to industrial CO2 emissions.

fviz_pca_var(pca_1, col.var = "cos2" , gradient.col = c("blue" , "green" , "red"), repel = TRUE )

fviz_pca_var(pca_1, col.var = "contrib" , gradient.col = c("blue" , "green" , "red"), repel = TRUE )

The two graphs show a Principal Component Analysis (PCA). The first displays the quality of representation of the variables (cos²), and the second shows their contribution to the principal axes. Dim1 (55.5%) is related to urbanization, pollution, and CO2 emissions (transport, buildings, industry). Dim2 (23.5%) appears to reflect climatic factors such as precipitation and temperature variation. Variables like CO2.build, Urb.Pop, air.pol, and ind.CO2 are correlated and strongly associated with Dim1. Precip and Temp var are opposed to pollution-related variables, suggesting an inverse relationship between climate évolution .

DENDROGRAM The dendrogram below shows us that 3 main classes are created according to the similarities of individuals Dendrogram The dendrogram below shows us that 3 main classes are created according to the similarities of individuals

plot.HCPC(res.HCPC,choice='tree',title='Arbre hiérarchique')

The three (3) classes thus created are: Class 1: CIV, BFA, SEN Class 2: FRA, DEU Class 3: NGA, EGY, ZAF THE FACTORIAL PLAN The factorial design projects countries into the space of the first two principal dimensions (Dim1 and Dim2)

plot.HCPC(res.HCPC,choice='map',draw.tree=FALSE,title='Plan factoriel')

3D REPRESENTATION OF HIERARCHICAL CLASSIFICATION

plot.HCPC(res.HCPC,choice='3D.map',ind.names=FALSE,centers.plot=FALSE,angle=60,title='Arbre hiérarchique sur le plan factoriel')

#IV.LINEAR REGRESSION DEFINITION Linear regression is a statistical method that models the relationship between a dependent variable (or response) and one or more independent variables (or predictors). It is based on the assumption that this relationship is linear, that is, that a change in the independent variable results in a proportional change in the dependent variable. SELECTION OF VARIABLES We will use the following variables: Urban Pop Transport Density Industry

# REGRESSIONS LINEAIRES
regressionL <- lm(Trans_CO2 ~ Urb.Pop, data = data)
summary(regressionL)

## 
## Call:
## lm(formula = Trans_CO2 ~ Urb.Pop, data = data)
## 
## Residuals:
##       Min        1Q    Median        3Q       Max 
## -60176983 -19612540 -10896415  12214413  75777461 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.356e+07  2.702e+07   0.502    0.634
## Urb.Pop     8.877e-01  4.966e-01   1.788    0.124
## 
## Residual standard error: 46040000 on 6 degrees of freedom
## Multiple R-squared:  0.3475, Adjusted R-squared:  0.2388 
## F-statistic: 3.196 on 1 and 6 DF,  p-value: 0.1241

library(car)  # regression linéaire

## Warning: le package 'car' a été compilé avec la version R 4.4.3

## Le chargement a nécessité le package : carData

## Warning: le package 'carData' a été compilé avec la version R 4.4.3

## 
## Attachement du package : 'car'

## L'objet suivant est masqué depuis 'package:psych':
## 
##     logit

library(carData)  # regression linéaire
library("clusterSim")

## Warning: le package 'clusterSim' a été compilé avec la version R 4.4.3

## Le chargement a nécessité le package : cluster

## Le chargement a nécessité le package : MASS

library(DataExplorer)

## Warning: le package 'DataExplorer' a été compilé avec la version R 4.4.3

library(FactoInvestigate)

## Warning: le package 'FactoInvestigate' a été compilé avec la version R 4.4.3

attach(data)
library(corrplot)
library(psych)
library(Hmisc)
library(ggplot2) # graphique
library(factoextra)  # graphique
library(FactoMineR)  # lancement ACP
# Graphe de la régression linéaire simple
ggplot(data, aes(x = Urb.Pop, y = Trans_CO2)) +
  geom_point(color = "blue") +
  geom_smooth(method = "lm", se = TRUE, color = "red") +
  ggtitle("Régression linéaire : Effet de Urb.Pop sur Trans_CO2 ") +
  xlab("Urb.Pop") + ylab("Trans_CO2") +
  theme_minimal()

## `geom_smooth()` using formula = 'y ~ x'

regressionM <- lm(Dens ~ Urb.Pop + Indus_CO2 + Trans_CO2, data = data)
summary(regressionM)

## 
## Call:
## lm(formula = Dens ~ Urb.Pop + Indus_CO2 + Trans_CO2, data = data)
## 
## Residuals:
##     BFA     CIV     EGY     FRA     DEU     NGA     SEN     ZAF 
##  20.890  18.432  33.394 -46.531  32.853   1.685 -22.804 -37.918 
## 
## Coefficients:
##               Estimate Std. Error t value Pr(>|t|)  
## (Intercept)  5.080e+01  2.640e+01   1.925   0.1266  
## Urb.Pop      2.431e-06  8.047e-07   3.021   0.0391 *
## Indus_CO2   -4.434e-06  2.531e-06  -1.752   0.1547  
## Trans_CO2    7.126e-07  4.128e-07   1.726   0.1594  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 42.12 on 4 degrees of freedom
## Multiple R-squared:  0.8307, Adjusted R-squared:  0.7037 
## F-statistic: 6.542 on 3 and 4 DF,  p-value: 0.05061

vif(regressionM)  # Multicolinéarité

##   Urb.Pop Indus_CO2 Trans_CO2 
##  3.137864  3.814946  1.872009

INTERPRETATION The image shows a linear regression graph illustrating the relationship between Urb.Pop (urban population) and Trans_CO2 (CO2 emissions from transportation). Graph Analysis: 1. Blue points: These represent observed data. Each point corresponds to a (Urb.Pop, Trans_CO2) pair. 2. Red line: This is the linear regression line, which represents the general trend between urban population and CO2 emissions. 3. Gray area: This is the confidence interval, showing the uncertainty around the regression.

Interpretation:

• The red line shows an increasing trend: as the urban population grows, CO2 emissions also increase. • The gray area is wider at the extremes, indicating more uncertainty in those regions. • The blue points do not perfectly align with the red line, meaning

there is some data dispersion (suggesting that other factors influence CO2 emissions besides urban population). CONCLUSION: There is a positive relationship between urban population and CO2 emissions from transportation: the larger the urban population, the higher the emissions. However, the data dispersion indicates that the relationship is not perfect and that other variables may also play a role.

4. Variable maps

FIGURE 1 : AGRI CO2

FIGURE 2 : Temperature balance

FIGURE 3 : CARTE SIG

FIGURE 4 : Air pollution CO

$FIGURE 5 : HIERACHICAL$

FIGURE 5 : HIERACHICAL

FIGURE 6 : Urban Population

Bibliography

[1] « Climate impacts of air pollution ». Consulté le: 9 avril 2025. [En ligne]. Disponible sur: https://www.who.int/teams/environment-climate-change-and-health/air-quality-energy-and-health/health-impacts/climate-impacts-of-air-pollution [2] N. Bahman et al., « A critical review on global CO2 emission: where do industries stand? », Reviews on Environmental Health, vol. 38, no 4, p. 681‑696, déc. 2023, doi: 10.1515/reveh-2022-0105. [3] J. G. J. Olivier, J. a. H. W. Peters, et G. Janssens-Maenhout, « Trends in global CO2 emissions. 2012 Report », juill. 2012, doi: 10.2788/33777. [4] S. Solaymani, « CO2 emissions patterns in 7 top carbon emitter economies: The case of transport sector », Energy, vol. 168, p. 989‑1001, 2019. [5] B. Koerner et J. Klopatek, « Anthropogenic and natural CO2 emission sources in an arid urban environment », Environmental Pollution, vol. 116, p. S45‑S51, 2002. [6] C. Le Quéré et al., « Temporary reduction in daily global CO2 emissions during the COVID-19 forced confinement », Nat. Clim. Chang., vol. 10, no 7, p. 647‑653, juill. 2020, doi: 10.1038/s41558-020-0797-x. [7] U. Nations, « No, the pandemic did not help climate action », United Nations. Consulté le: 9 avril 2025. [En ligne]. Disponible sur: https://www.un.org/en/desa/no-pandemic-did-not-help-climate-action?utm_source=chatgpt.com [8] X. Zhou, E. Weng, et Y. Luo, « Modeling patterns of nonlinearity in ecosystem reponses to temperature, CO2, and precipitation changes », Ecological Applications, vol. 18, no 2, p. 453‑466, 2008. [9] « Zotero | Storage ». Consulté le: 9 avril 2025. [En ligne]. Disponible sur: https://www.zotero.org/storage [10] « Liens entre CO2 et température, par Frédéric Parrenin – Le Climat en Questions ». Consulté le: 9 avril 2025. [En ligne]. Disponible sur: https://www.climat-en-questions.fr/reponse/liens-entre-co2-et-temperature-par-frederic-parrenin/ [11] C. Linton, Grant-Muller ,Susan, et W. F. and Gale, « Approaches and Techniques for Modelling CO2 Emissions from Road Transport », Transport Reviews, vol. 35, no 4, p. 533‑553, juill. 2015, doi: 10.1080/01441647.2015.1030004. [12] H. Ritchie, P. Rosado, et M. Roser, « CO₂ and Greenhouse Gas Emissions », Our World in Data, déc. 2023, Consulté le: 9 avril 2025. [En ligne]. Disponible sur: https://ourworldindata.org/co2-and-greenhouse-gas-emissions [13] H. Ritchie, « CO₂ emissions dataset: our sources and methods », Our World in Data, févr. 2022, Consulté le: 9 avril 2025. [En ligne]. Disponible sur: https://ourworldindata.org/co2-dataset-sources [14] « Annual precipitation », Our World in Data. Consulté le: 9 avril 2025. [En ligne]. Disponible sur: https://ourworldindata.org/grapher/average-precipitation-per-year [15] « Population density », Our World in Data. Consulté le: 9 avril 2025. [En ligne]. Disponible sur: https://ourworldindata.org/grapher/population-density

Comparative Study of the Causes and Impacts of CO₂ Emissions Worldwide Before, During, and After COVID-19 (2019-2021)

KONAN Angoua.U. Emmanuel, LOMPO Palamanga, NANA Tinwendé Jedida, RABDO Guélilou

2025-04-09

INTRODUCTION

I. DOCCUMENTARY REVIEW

4. Variable maps