JURY MEMBERS :

Dr. Maïmouna Bologo/Traoré, Dr. Malicki Zorom, Dr. Yohan Richardson and Mr. Sina Thiam

ABSTRACT

  Faced with the growing effects of climate change, Sahelian countries are confronted with major challenges in terms of environmental and socio-economic vulnerability. This study aims to assess the climate vulnerability of these countries in order to guide more targeted emergency and adaptation policies. To this end, we analyzed the influence of various climatic factors (temperature anomalies and extremes, precipitation), environmental factors (water stress, forest cover changes), and socio-economic factors (population growth, access to essential services) on the vulnerability of infrastructures and populations. Through Principal Component Analysis (PCA) complemented by Hierarchical Cluster Analysis and GIS mapping, three vulnerability profiles emerge: highly exposed countries (e.g., Niger, Chad, Mauritania), countries in transition (e.g., Mali, Burkina Faso), and relatively resilient countries (e.g., Cameroon, Nigeria). These findings call for targeted public policies, by optimizing water resource management, strengthening health infrastructures, and developing climate-adapted agricultural practice.

Introduction

  The impact of climate change on infrastructure and human societies is now a major global issue. Sahelian countries, in particular, are facing extreme climatic conditions that compromise their natural resources and accentuate their socio-economic vulnerability[1]. Rising temperatures, erratic rainfall and the increased frequency of extreme weather events are exacerbating water stress, threatening food security and amplifying pressures on infrastructure and populations[2]. Despite the urgency of the situation, these countries are struggling to put in place effective adaptation strategies due to economic, institutional and environmental challenges. In addition, water resources management and infrastructure adaptation are hampered by the lack of accurate data and policies adapted to local realities. It is therefore imperative to assess the climate vulnerability of Sahelian countries through an approach that integrates climatic, environmental and socio-economic dimensions[3]. This study is part of an analytical approach aimed at understanding the effects of climate change on infrastructure and natural resources in the Sahel region. The main objective is to identify the determinants of climate vulnerability, assess disparities between countries and propose recommendations to build resilience. More specifically, This analysis seeks to answer the following questions:

To address these questions, we adopted a quantitative approach combining Principal Component Analysis (PCA) to identify relationships between variables, Hierarchical Cluster Analysis (HCA) to group countries according to their vulnerability profiles, and GIS mapping (QGIS) to visualize territorial disparities. This study thus aims to provide practical recommendations to policymakers and development actors in order to optimize resource management and improve infrastructure adaptation in the face of growing climate challenges.

I. Literature Review

1. General Framework of the Study

  Climate change is one of the major challenges of the 21st century, affecting various aspects of the environment, particularly in Sahelian countries in Africa. Numerous studies highlight the harmful effects of rising temperatures, changing precipitation patterns, and increasing extreme climate events on ecosystems, biodiversity, and natural resources in this vulnerable region.

2. Impact of Climate Change on Ecosystems

  Climate change is profoundly altering Sahelian ecosystems. Rising temperatures and decreasing rainfall are causing increased desertification and a scarcity of natural resources. According to the Intergovernmental Panel on Climate Change (IPCC), these changes disrupt fauna and flora, particularly species adapted to semi-arid conditions. Aquatic ecosystems, such as Lake Chad, are also undergoing major transformations, with a drastic reduction in surface area, threatening the livelihoods of local populations【2】

3. Deforestation and Loss of Biodiversity

  Climate change exacerbates deforestation in Sahelian countries, notably in Niger, Mali, and Burkina Faso, where human pressure and extreme weather conditions weaken forests. The World Bank notes that desert expansion, combined with intensive land use, leads to increased soil degradation and reduced biodiversity. The gradual disappearance of forests also lowers carbon sequestration capacity, further intensifying global warming【4】

4. Effects on Water Resources and Agriculture

  Water resources are particularly threatened in the Sahel. Changing rainfall patterns result in prolonged droughts and more frequent floods, affecting access to drinking water and agricultural production. The African Development Bank indicates that farmers in Senegal and Chad suffer significant yield losses due to erratic rainfall and soil degradation. These conditions endanger food security and fuel conflicts between farmers and herders over access to natural resources【5】

5. Challenges in Financing and Adaptation Strategies

  Adapting to climate change is a major challenge for Sahelian countries, which often lack the financial resources to implement resilience strategies. A report by the United Nations Development Programme (UNDP) highlights that initiatives like the Great Green Wall, aimed at restoring degraded lands and halting desertification, can mitigate some effects of climate change. However, such projects require substantial funding, and Sahelian countries struggle to secure sufficient support from international institutions【6】.

6. Case Studies of Certain Countries

  In the Sahel, several countries are developing strategies to combat the effects of climate change. In Niger, sustainable land management projects, such as assisted natural regeneration programs, have improved agricultural productivity and restored degraded lands. In Mali, the promotion of appropriate agricultural techniques like drip irrigation contributes to better water resource management. In Burkina Faso, agroecological practices, such as stone bunds to fight soil erosion, have shown positive results in protecting the environment and enhancing farmers’ resilience【7】.

II. Methodology

  To analyze the dynamics at play in the impact of climate change, this study adopts a rigorous quantitative approach. The chosen methodology combines data from well-established sources with advanced statistical processing to reveal significant trends. It is structured around three main stages: data collection and preparation, statistical analysis using Principal Component Analysis (PCA), and interpretation of results for a better understanding of the phenomena studied. Tools Used in This Study:

1. Data Collection

1.1. Data Collection Tools

The data used in this study come from sources known for their reliability, methodological rigor, and comprehensiveness:

  • World Bank: A leading institution for economic and social data, providing validated statistics used by decision-makers and researchers worldwide. Its broad data coverage allows in-depth analysis of economic and environmental dynamics【8】.

  • ND-GAIN: The Notre Dame Global Adaptation Initiative measures a country’s exposure, sensitivity, and ability to adapt to the impacts of climate change. This platform evaluates overall vulnerability across six vital sectors food, water, health, ecosystem services, human habitat, and infrastructure【9】.

  • Our World in Data: This platform offers data from academic research and recognized institutions, covering a wide range of environmental, economic, and social indicators. Its open access and frequent updates make it a valuable source for analyzing global trends【10】.

1.2. Target Population

The target population for our study includes:

  • Local communities (farmers, herders, fishermen, rural and urban residents)

  • Local and regional authorities (government officials, policymakers)

  • Local and international NGOS

  • Experts in climatology and ecology (researchers and scientists)

  • Local economic actors (agricultural businesses, water and energy sectors)

  This diversity ensures the collection of both quantitative and qualitative field data. The questionnaire was administered to households to understand their perceptions and vulnerabilities related to climate change. Meanwhile, the interview guides were addressed to experts, NGOs, policymakers, and other stakeholders to gather specialized information on causes, effects, and adaptation policies. Appendix 1: Questionnaire and interview guide

1.3. Analyzed Variables

  As part of this study on the impact of climate change, several variables were used. The table below presents the adopted abbreviations along with their definitions:

MAP REPRESENTING THE STUDY AREA

Table of variables Data table

RTI <- read.csv(file = "RR.csv", header = TRUE, row.names = 1, sep = ";", dec = ",", fileEncoding = "ISO-8859-1")
RTI
##                PRECIP TEMP_EX TEMP_ANOM   H2O_STR     H2O_USE     FOREST
## Burkina     700.36096     1.3 0.3413738 0.4281452   6.5440000 22.9035088
## Cameroon   1623.25790     0.1 0.6139796 0.3324320   0.3986813 43.1479766
## Chad        203.81463     2.7 0.4065404 0.5122961   5.8640000  3.5117535
## Gambia      944.07180     0.2 0.5789345 0.3630312   3.3866667 24.5454545
## Guinea     1789.59750     0.2 0.6665189 0.3271297   0.3938053 25.3499919
## Mali        276.58563     0.7 0.5571928 0.4742314   8.6433333 10.8966636
## Mauritania   41.42700     3.1 0.9427039 0.5917325 337.0500000  0.3088192
## Niger        29.05078     7.5 0.4446030 0.8215759  73.8087943  0.8719823
## Nigeria     820.79810     0.1 0.1978370 0.3818094   5.6434389 24.1043842
## Senegal     488.93503     0.6 0.6394126 0.5080147  11.8666667 42.1137485
##              VUL_HAB RL_INDEX  VUL_FOOD VUL_HEALTH POP_GROW       GES
## Burkina    0.5917942     0.99 0.6451618  0.6430062    2.415  54928136
## Cameroon   0.6687529     0.84 0.5693268  0.5851721    2.696  65614304
## Chad       0.6372580     0.91 0.6991467  0.7965270    3.355 109226270
## Gambia     0.6191002     0.96 0.6182748  0.5195413    2.396   3613415
## Guinea     0.6400532     0.89 0.5302772  0.5257007    2.527  46723690
## Mali       0.5996820     0.98 0.6183859  0.6360676    3.077  64378856
## Mauritania 0.5333328     0.97 0.5836399  0.6207052    2.900  17282982
## Niger      0.6238498     0.93 0.7846569  0.7528860    3.289  55550400
## Nigeria    0.6249186     0.85 0.5833909  0.5611565    2.108 414200540
## Senegal    0.5598835     0.94 0.6434956  0.6626767    2.590   4455574

2. Data Analysis

  The data analysis was conducted using Principal Component Analysis (PCA), a statistical technique used to explore multidimensional quantitative data. The purpose of PCA is to reduce the dimensionality of the dataset while preserving the essential information, which facilitates interpretation of relationships among variables. PCA works by transforming the initial variables into a new set of uncorrelated variables, called principal components. These components are ordered so that the first captures the greatest amount of total data variance, followed by others that progressively account for less. The analysis was carried out using RStudio, following several methodological steps:

3. Résults and Interprétation

This section presents a synthesis of the statistical analyses carried out to identify the interactions between climatic, environmental, and socio-economic variables, as well as their implications for the vulnerability of Sahelian countries.

library(readr)
library(FactoMineR)
library(factoextra)
## Le chargement a nécessité le package : ggplot2
## Welcome! Want to learn more? See two factoextra-related books at https://goo.gl/ve3WBa
library(corrplot)
## corrplot 0.95 loaded
library(ggrepel)
library(viridis)
## Le chargement a nécessité le package : viridisLite
library(Factoshiny)
## Le chargement a nécessité le package : shiny
## Le chargement a nécessité le package : FactoInvestigate
library(cluster)
library(RColorBrewer)
library(factoextra)

Loading and Preparing the Data

guess_encoding("RR.csv")
## # A tibble: 1 × 2
##   encoding confidence
##   <chr>         <dbl>
## 1 ASCII             1
RTI <- read.csv(file = "RR.csv", header = TRUE, row.names = 1, sep = ";", dec = ",", fileEncoding = "ISO-8859-1")
RTI_scaled <- scale(RTI)
res.pca <- PCA(RTI_scaled, graph = FALSE)

3.1. Determining the Number of Components to Retain

# Extraction and Display of Eigenvalues
eig_values <- res.pca$eig
print(eig_values)
##        eigenvalue percentage of variance cumulative percentage of variance
## comp 1 5.77928066             48.1606722                          48.16067
## comp 2 2.70749476             22.5624563                          70.72313
## comp 3 1.28780666             10.7317222                          81.45485
## comp 4 1.07790593              8.9825495                          90.43740
## comp 5 0.52902336              4.4085280                          94.84593
## comp 6 0.43346410              3.6122008                          98.45813
## comp 7 0.10186281              0.8488567                          99.30699
## comp 8 0.05826520              0.4855433                          99.79253
## comp 9 0.02489652              0.2074710                         100.00000
# Variance Explained by Each Component 
fviz_eig(res.pca, addlabels = TRUE, barfill = "steelblue", barcolor = "black", 
         main = "Explained Variance")

  Principal Component Analysis (PCA) is used to reduce the dimensionality of the data while preserving most of the information. The scree plot below shows the proportion of variance explained by each component.

  In our case, the first two principal components together explain 70.8% of the total variance:

Axis 1: 48.2%

Axis 2: 22.6%

  According to Kaiser’s criterion (which recommends retaining components with eigenvalues greater than 1), as well as the presence of a clear elbow in the scree plot, it is appropriate to retain the first two components for interpretation.

3.2. Countries Contributions to Principal Components 1 and 2 and Factorial plane

# Extract the contributions of individuals (countries) on axes 1 and 2
contrib_ind <- res.pca$ind$contrib[, 1:2]
# Convert to a data frame for easier manipulation
contrib_df <- as.data.frame(contrib_ind)
# Add a column with the total contribution on both axes
contrib_df$Total <- rowSums(contrib_df)
# Add a column for country names
contrib_df$Pays <- rownames(contrib_df)
# Create a horizontal bar chart of total contributions
ggplot(contrib_df, aes(x = reorder(Pays, -Total), y = Total)) +
  geom_bar(stat = "identity", fill = ifelse(contrib_df$Pays == "Mauritania", "red", "steelblue")) +  # Highlight Mauritania in red
  labs(title = "Contribution of Countries to Axes 1 and 2",
       x = "Country", y = "Total Contribution (%)") +
  theme_minimal() +
  coord_flip()

fviz_pca_ind(res.pca, repel = TRUE, col.ind = "red", title = "PCA Factorial Plane")

  The graph below represents the total contribution of countries to the first two axes of the principal component analysis.

  It shows that Mauritania has a significantly higher contribution compared to the other countries, indicating that it strongly influences the orientation of the factorial axes.

  This structuring role, combined with its possible outlier position in the factorial plane, makes it an atypical individual within the studied group.

3.3. Analysis without Mauritania

loading and Preparing the Data

RTI <- read.csv(file = "RR.csv", header = TRUE, row.names = 1, sep = ";", dec = ",", fileEncoding = "ISO-8859-1")
RTI <- RTI[rownames(RTI) != "Mauritania", ]  # Exclusion of Mauritania
RTI_scaled <- scale(RTI)
res.pca <- PCA(RTI_scaled, graph = FALSE)

3.3.1. Determining the Number of Components to Retain

# Extraction and Display of Eigenvalues
eig_values <- res.pca$eig
print(eig_values)
##        eigenvalue percentage of variance cumulative percentage of variance
## comp 1 6.34643455             52.8869546                          52.88695
## comp 2 2.21516772             18.4597310                          71.34669
## comp 3 1.68947403             14.0789503                          85.42564
## comp 4 0.82041528              6.8367940                          92.26243
## comp 5 0.56605286              4.7171072                          96.97954
## comp 6 0.24152325              2.0126937                          98.99223
## comp 7 0.08376709              0.6980590                          99.69029
## comp 8 0.03716521              0.3097101                         100.00000
# Variance Explained by Each Component 
fviz_eig(res.pca, addlabels = TRUE, barfill = "steelblue", barcolor = "black", 
         main = "Explained Variance")

  Principal Component Analysis (PCA) was conducted again after excluding Mauritania from the dataset. This adjustment aimed to assess the influence of atypical countries on the overall structure of the data.

  Following this exclusion, the first two principal components explain a total of 71.4% of the variance:

Axis 1: 52.9%

Axis 2: 18.5%

  This slight increase in the variance explained by the first axis suggests that Mauritania had a strong structuring effect on the original PCA results. The current analysis, based on Kaiser’s criterion (eigenvalues > 1) and the scree plot’s clear elbow, confirms that the first two components remain sufficient for interpretation.

3.3.2. Countries Contributions to Principal Components 1 and 2 and Factorial plane

# Extract the contributions of individuals (countries) on axes 1 and 2
contrib_ind <- res.pca$ind$contrib[, 1:2]
# Convert to a data frame for easier manipulation
contrib_df <- as.data.frame(contrib_ind)
# Add a column with the total contribution on both axes
contrib_df$Total <- rowSums(contrib_df)
# Add a column for country names
contrib_df$Pays <- rownames(contrib_df)
# Create a horizontal bar chart of total contributions
ggplot(contrib_df, aes(x = reorder(Pays, -Total), y = Total)) +
  geom_bar(stat = "identity", fill = ifelse(contrib_df$Pays == "Nigeria", "red", "steelblue")) +  # Highlight Mauritania in red
  labs(title = "Contribution of Countries to Axes 1 and 2",
       x = "Country", y = "Total Contribution (%)") +
  theme_minimal() +
  coord_flip()

fviz_pca_ind(res.pca, repel = TRUE, col.ind = "red", title = "PCA Factorial Plane")

  After removing Mauritania and rerunning the PCA, Nigeria emerged as a new outlier, showing a noticeably high contribution to the principal axes and a clear separation in the factorial plane. These observations indicate that Nigeria also plays a structuring role in the distribution of individuals within the principal component space.

  This influence may be due to Nigeria’s extreme socio-economic or environmental characteristics compared to other countries in the study. Therefore, Nigeria should also be considered an atypical individual, as its weight could potentially overshadow other underlying structures if not properly accounted for in the interpretation

3.4. Analysis without Mauritania and Nigeria

RTI <- read.csv(file = "RR.csv", header = TRUE, row.names = 1, sep = ";", dec = ",", fileEncoding = "ISO-8859-1")
RTI <- RTI[!(rownames(RTI) %in% c("Mauritania", "Nigeria")), ]  # Exclusion of Mauritania and Nigeria
RTI_scaled <- scale(RTI)
res.pca <- PCA(RTI_scaled, graph = FALSE)

3.4.1. Determining the Number of Components to Retain

# Extraction and Display of Eigenvalues
eig_values <- res.pca$eig
print(eig_values)
##        eigenvalue percentage of variance cumulative percentage of variance
## comp 1 6.81524502             56.7937085                          56.79371
## comp 2 2.44358723             20.3632269                          77.15694
## comp 3 1.38246011             11.5205009                          88.67744
## comp 4 0.66591414              5.5492845                          94.22672
## comp 5 0.50215763              4.1846469                          98.41137
## comp 6 0.13173854              1.0978212                          99.50919
## comp 7 0.05889734              0.4908112                         100.00000
# Variance Explained by Each Component 
fviz_eig(res.pca, addlabels = TRUE, barfill = "steelblue", barcolor = "black", 
         main = "Explained Variance")

  A further refinement of the Principal Component Analysis (PCA) was performed by excluding both Mauritania and Nigeria, two countries that exhibited atypical behaviors in the previous analyses.

  After this adjustment, the first two principal components explain a total of 77.2% of the variance:

Axis 1: 56.8%

Axis 2: 20.4%

  The increase in the variance explained, particularly by the first axis, indicates that both Mauritania and Nigeria had a significant influence on the orientation of the factorial axes. Their exclusion results in a more balanced representation of the remaining countries, allowing for a clearer interpretation of the underlying patterns in the data. The scree plot and Kaiser’s criterion continue to support the selection of the first two principal components for meaningful analysis.

3.4.2. Matrice de corrélation

corrplot(cor(RTI),
         method = "color",
         type = "upper",
         order = "hclust",
         diag = FALSE,
         col = viridis(100),
         tl.col = "black",
         tl.srt = 45,
         tl.cex = 0.8,
         number.cex = 0.7,
         number.digits = 2,
         addCoef.col = "black",
         addgrid.col = "gray90",
         mar = c(1, 1, 2, 1),
         title = "Matrice de Corrélations",
         cex.main = 1.2)

  The correlation matrix analysis revealed significant relationships between the variables studied:

  • Strong Correlations (|r| ≥ 0.7) :
    • Water Stress (H2O_STR) and Extreme Temperatures (TEMP_EX) :r = 0.94. Heatwaves exacerbate water stress by increasing evaporation and water demand. This underscores the urgency of water management policies in arid regions (e.g., Niger, Chad).
    • Water Stress (H2O_STR) and Population Growth (POP_GROW) :r = 0.76. Population growth intensifies water demand for agriculture and domestic use, worsening shortages.
    • Food Vulnerability (VUL_FOOD) and Health Vulnerability (VUL_HEALTH) :r = 0.83. Food insecurity and health issues (e.g., malnutrition, waterborne diseases) are interconnected, requiring integrated interventions.
    • Extreme Temperatures (TEMP_EX) and Water Use (H2O_USE)* :r = 0.95. High temperatures drive increased water withdrawals for irrigation and industry, threatening local resources.
  • Significant Negative Correlations :
    • Precipitation (PRECIP) and Extreme Temperatures (TEMP_EX) : r = -0.64. Regions with extreme temperatures tend to receive less rainfall, accelerating desertification.
    • Forest Cover (FOREST) and Greenhouse Gas Emissions (GES) :r = -0.38. Deforestation (for agriculture or urbanization) releases GHGs, reducing ecosystems’ carbon sequestration capacity.
  • Absence of Correlation (|r| ≤ 0.3) :
    • Red List Index (RL_INDEX) and Precipitation (PRECIP) :r = 0.01. Biodiversity threats are not directly linked to rainfall, suggesting other factors (e.g., poaching, habitat fragmentation).
    • Temperature Anomalies (TEMP_ANOM) and Health Vulnerability (VUL_HEALTH) :r = -0.20. Annual temperature variations do not directly affect health, but indirect impacts (e.g., droughts) may be critical.

3.4.3. Principal Component Analysis (PCA)

3.4.3.1. Variable Correlation Plot

fviz_pca_var(res.pca, col.var = "cos2", repel = TRUE, title = "Corrélations des variables")

  • Variables such as VUL_FOOD, H2O_STR, H2O_USE, TEMP_EX, VUL_HEALTH, and POP_GROW are highly positively correlated and well represented (cos² > 0.8) on Dim1. These represent countries facing high exposure to water stress, extreme temperatures, demographic pressure, and multidimensional vulnerability (health and food security).

  • VUL_HAB and GES (Greenhouse Gas Emissions) are also well represented but more aligned with Dim2, indicating a different pattern of vulnerability related to housing and emissions.

  • FOREST, TEMP_ANOM, and PRECIP are located on the opposite side of Dim1, indicating that countries with higher forest coverage, more rainfall, and lower anomalies in temperature tend to be less vulnerable in the same ways.

  • The RL_INDEX (Resilience Index) is strongly negatively projected on Dim2, indicating that countries with a high resilience index have an adaptive advantage that sets them apart from the vulnerable ones.

3.4.3.2. PCA Biplot (Countries + Variables)

fviz_pca_biplot(res.pca, repel = TRUE, col.var = "blue", col.ind = "red", title = "Plan Factoriel ACP")

This graph allows for simultaneous interpretation of country positions and the influence of variables.

  • Countries on the right (Dim1 positive): Niger and Chad stand out as highly exposed to extreme temperatures, water scarcity, and high population growth, combined with major vulnerabilities in food, health, and water access.

  • Countries on the left (Dim1 negative): Cameroon, Guinea, and The Gambia are characterized by better climate conditions: more forest coverage, higher precipitation, and reduced climate stress. These countries appear to have relatively better environmental profiles.

  • Countries at the bottom (Dim2 negative): Burkina Faso, Senegal: These countries exhibit low resilience indexes, indicating reduced adaptive capacity despite moderate climate exposure.

  • Countries at the top (Dim2 positive): Guinea and Cameroon also score higher on Dim2, suggesting a relatively higher capacity to adapt due to their environmental context.

3.4.4. Contribution of Variables to Principal Component Analysis (PCA)

fviz_pca_contrib(res.pca, choice = "var", axes = 1:2, title = "Contribution des variables aux axes")
## Warning in fviz_pca_contrib(res.pca, choice = "var", axes = 1:2, title =
## "Contribution des variables aux axes"): The function fviz_pca_contrib() is
## deprecated. Please use the function fviz_contrib() which can handle outputs of
## PCA, CA and MCA functions.

fviz_contrib(res.pca, choice = "var", axes = 1, top = 12, title = "Contribution des variables - Axe 1")

fviz_contrib(res.pca, choice = "var", axes = 2, top = 12, title = "Contribution des variables - Axe 2")

PCA identifies the variables that most significantly influence the dataset’s structure, revealing the fundamental dynamics shaping climate vulnerability patterns across Sahelian nations.

RTI$TEMP_POP <- RTI$TEMP_EX * RTI$POP_GROW

ggplot(RTI, aes(x = TEMP_POP, y = H2O_STR)) +
  geom_smooth(method = "lm", formula = y ~ poly(x, 2), color = "red") +
  geom_point(aes(size = FOREST, color = PRECIP)) +
  geom_text_repel(aes(label = rownames(RTI)), size = 3) +
  scale_color_viridis(name = "Précipitations") +
  labs(x = "TEMP_EX × POP_GROW",
       y = "Stress hydrique",
       title = "Effet combiné sur les ressources en eau") +
  theme_minimal()

  This graph illustrates how climatic and demographic dynamics together intensify water stress in several Sahelian countries.

#3.4 Dendogramme

dist_matrix <- dist(RTI_scaled)
hc <- hclust(dist_matrix, method = "ward.D2")
nb_clusters <- 3
colors <- brewer.pal(nb_clusters, "Dark2")

# Affichage du dendrogramme
fviz_dend(hc, k = nb_clusters, cex = 0.8,
          k_colors = colors, rect = TRUE, rect_fill = TRUE,
          rect_border = "black", lwd = 1.2,
          main = "Dendrogramme des pays - Classification hiérarchique")
## Warning: The `<scale>` argument of `guides()` cannot be `FALSE`. Use "none" instead as
## of ggplot2 3.3.4.
## ℹ The deprecated feature was likely used in the factoextra package.
##   Please report the issue at <https://github.com/kassambara/factoextra/issues>.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.

Structure of the Dendrogram

  The vertical axis (Height) represents the level of dissimilarity between clusters: the higher the merging occurs, the more dissimilar the groups are three main homogeneous clusters emerge (highlighted in grey):

DENDROGRAM MAP

III. Discussion

  The overall results highlight complex interactions between climatic, environmental, and socio-economic factors.*
The strong correlation between extreme temperatures and water stress underscores the critical role of climate in the degradation of water resources. Furthermore, the significant contribution of rainfall and forest cover reveals the importance of ecosystems in mitigating the impacts of climate change. The classification of countries into different clusters makes it possible to identify distinct vulnerability profiles, paving the way for differentiated adaptation strategies. For example, critical countries require urgent interventions in water management and demographic regulation, while resilient countries must maintain and strengthen their natural assets to continue buffering climate impacts.

IV. Adaptation strategies

  The analysis results indicate the need to implement specific adaptation strategies, tailored to the vulnerability profiles identified. These strategies must be targeted according to each group of countries’ climatic, environmental, and socio-economic characteristics

1. For Critical Zones (e.g., Chad, Niger)

  These countries combine unfavorable indicators: extreme temperatures, critical water stress, low forest cover, and very limited rainfall. «According to the UNDP, Chad faces extreme environmental degradation and is weakened by irregular rainfall, causing droughts, floods, and locust invasions. Forecasts indicate that the country will become increasingly hot and dry by the end of the century. »【11】 «According to the Maplecroft Climate Change Vulnerability Index (2016), Chad ranks as the most vulnerable country in the world. »【12】 «The ND-GAIN index ranks Niger as the second most vulnerable country to the increasing risks of floods, droughts, and extreme heat. The frequency and severity of extreme rainfall and flooding have significantly increased in recent years. »【13】

Recommanded strategies : « National Adaptation Plans supported by the UNDP, these plans aim to integrate climate adaptation into medium- and long-term development planning, particularly in agriculture, livestock, fisheries, and water resources. » 【11】

  • Optimize Irrigation: Deploy solar-powered drip irrigation systems to maximize water use and reduce evaporation losses.

  • Demographic Management: Implement policies to regulate population growth, easing pressure on water resources.

  • Strengthen Hydraulic Infrastructure: Invest in storage and water distribution systems to ensure supply even during dry spells.

2. For Tension Zones (e.g., Burkina Faso, Mali))

  These countries show intermediate indicators, with moderate water stress and moderately exploited natural resources. Recommended Strategies: « To adapt to climate change, farmers must adopt agricultural practices that reduce vulnerability to climate change and promote environmental preservation: the use of improved seeds, the diversification of crops on a plot and the choice of the date of sowing seeds. »[14]

  • Community Agroforestry: Promote agroforestry practices integrating tree planting and sustainable soil management to improve water retention and reduce erosion.

  • Rainwater Harvesting: Install rainwater collection and storage systems, especially in urban areas, to mitigate drought impacts.

  • Improve Energy Efficiency: Promote low-energy irrigation techniques and develop real-time water resource monitoring systems for proactive management.

3. For Resilient Zones (e.g., Cameroon, Nigeria)

  These countries benefit from a more favorable climate with high forest cover and abundant precipitation, but must safeguard and build upon these advantages to avoid future degradation. « According to the World Bank’s 2022 Cameroon Country Climate and Development Report, climate change poses an imminent threat to the country’s development due to its reliance on natural resources» 【15】

Recommended Strategies: « According to the World Bank’s 2022 Cameroon Country Climate and Development Report report identifies four priorities to tackle climate risks and foster a green, resilient, and inclusive future. »【16】

  • Develop climate-resilient agriculture, forestry, and land-use systems that combine mitigation measures and sustainable development.

  • Integrate climate risks into urban planning, infrastructure financing, and design to enhance city resilience and well-being.

  • Invest in sustainable infrastructure to address infrastructure deficits and improve living standards.

  • Adopt a holistic approach to resilience that includes community-led solutions, citizen engagement, local government involvement, and cross-sectoral coordination.

  • Forest Ecosystem Conservation: Strengthen protection and sustainable management of forests to preserve their buffering and hydrological regulation roles.

  • Carbon Credit Mechanisms: Leverage carbon credits to fund forest conservation and encourage sustainable development practices.

  • Local Governance: Implement integrated natural resource management policies to ensure infrastructure sustainability and anticipate climate change.

4. Special Case : Mauritania

  Despite a medium climate-demography index, Mauritania experiences high water stress, mainly due to overexploitation of aquifers. « Mauritania is one of the most arid countries in the world, with one of the lowest levels of water availability. Over 22% of the population lacks access to basic drinking water sources, with stark regional disparities. Climate change exacerbates this with erratic rainfall and prolonged droughts . »【17】 « Rural regions like Assaba face severe water shortages and unfavorable hydrogeological conditions (e.g., groundwater salinization). Women and children are particularly affected, spending hours daily fetching water—impacting their health, safety, and education. . »【17】 « The 2020 ThinkHazard report indicates that Mauritania faces high risk from coastal flooding, drought, extreme heatwaves, wildfires, and general water scarcity. . »【18】

Recommended Strategies:

  • Strict Water Management: Enforce water withdrawal quotas and improve monitoring of water resources to avoid overuse.

  • Desalination Technologies: Invest in solar-powered desalination units to increase freshwater supply in a context of chronic water stress.

  • Regulatory Frameworks: Reinforce water governance policies for effective and sustainable resource management.

Conclusion

  This study shed light on the impact of climate change on the environment using a quantitative approach and an in-depth analysis of data from recognized sources. The Principal Component Analysis (PCA) helped identify the most influential variables, notably water resources, food systems, and extreme temperatures, which are key factors behind environmental disparities across regions. Other factors, such as health, precipitation, and forest cover, also play a significant role in structuring climate dynamics.   The results highlight the complexity of interactions among these various variables and emphasize the urgency of sustainable resource management in the face of climate challenges. However, some limitations must be acknowledged, particularly the reliance on secondary data, which can affect the precision and completeness of the analysis.   Despite these limitations, this study makes an essential contribution to understanding the effects of climate change on the environment and serves as a foundation for future research to deepen this analysis. Broader integration of local data and a diversification of analytical methods would help refine findings and better tailor adaptation strategies to the specific realities of the studied regions.

Figure1 representative map of the climate vulnerability index on water

Figure2 representative map of the climate vulnerability index on health

Figure3 representative map of greenhouse gases

Figure4 map representing the population growth rate

Figure5 representative map of the climate vulnerability index on housing

Figure6 representative map of the climate vulnerability index on food supplies

Figure7 map representing the annual temperature anomaly

Figure8 representative map of forest areas

Figure9 representative map of precipitation

Figure10 representative map of the red list index

Figure11 map representing extreme temperatures

Figure12 map representing the using of water

Household Survey Questionnaire

https://ee.kobotoolbox.org/x/zfddtXyl

Interview Guide for Researchers and Policymakers

https://ee.kobotoolbox.org/x/x8LZIq95

Interview Guide for NGOs

https://ee.kobotoolbox.org/x/NMXd7MTG

Interview Guide for Health Services

https://ee.kobotoolbox.org/x/XZWvZQLY

Interview Guide for Researchers and Scientific Experts

https://ee.kobotoolbox.org/x/x8LZIq95

Référence bibliographie

[1] « Le Sahel face aux changements climatiques ». Consulté le: 2 avril 2025. [En ligne]. Disponible sur: https://www.afrique-gouvernance.net/bdf_document-1407_fr.html

[2] A. D. Bank, « Les changements climatiques en Afrique », Banque africaine de développement. Consulté le: 31 mars 2025. [En ligne]. Disponible sur: https://www.afdb.org/fr/themes-et-secteurs/secteurs/changement-climatique/les-changements-climatiques-en-afrique

[3] Consulté le: 2 avril 2025. [En ligne]. Disponible sur: https://www.cascades.eu/wp-content/uploads/2023/01/CASCADES_Scenarios_Sahel_final-FR-with-back-cover.pdf

[4] « Fiche de résultats : Environnement », World Bank. Consulté le: 31 mars 2025. [En ligne]. Disponible sur: https://www.banquemondiale.org/fr/results/2013/04/13/environment-results-profile

[5] « L’insécurité alimentaire en Afrique de l’Ouest nécessite une réponse adaptée au climat dans un contexte de crises », World Bank. Consulté le: 31 mars 2025. [En ligne]. Disponible sur: https://www.banquemondiale.org/fr/news/feature/2022/09/08/west-africa-food-insecurity-demands-climate-smart-response-amid-multiple-crises

[6] « Adaptation au changement climatique en Afrique », UNDP. Consulté le: 31 mars 2025. [En ligne]. Disponible sur: https://www.undp.org/fr/publications/adaptation-au-changement-climatique-en-afrique

[7] « Les pays du Sahel doivent accélérer la croissance et prioriser l’adaptation climatique pour mieux faire face à la crise climatique et à l’insécurité alimentaire – nouveau rapport du Groupe de la Banque mondiale », World Bank. Consulté le: 31 mars 2025. [En ligne]. Disponible sur: https://www.banquemondiale.org/fr/news/press-release/2022/09/19/sahelian-countries-can-boost-and-diversify-their-economies-to-take-on-the-climate-crisis-and-food-insecurity

[8] « Open Knowledge Repository ». Consulté le: 8 avril 2025. [En ligne]. Disponible sur: https://openknowledge.worldbank.org/home

[9] M. C. W. // U. of N. Dame, « Country Index // Notre Dame Global Adaptation Initiative // University of Notre Dame », Notre Dame Global Adaptation Initiative. Consulté le: 8 avril 2025. [En ligne]. Disponible sur: https://gain.nd.edu/our-work/country-index/

[10] « Our World in Data ». Consulté le: 8 avril 2025. [En ligne]. Disponible sur: https://ourworldindata.org/

[11] « Le Tchad accélère sa course vers l’adaptation au changement climatique. Et après ? », UNDP. Consulté le: 3 avril 2025. [En ligne]. Disponible sur: https://www.undp.org/fr/blog/le-tchad-accelere-sa-course-vers-ladaptation-au-changement-climatique-et-apres

[12] « Environnement et énergie | Programme De Développement Des Nations Unies ». Consulté le: 4 avril 2025. [En ligne]. Disponible sur: https://www.undp.org/fr/chad/environnement-et-energie

[13] « Une approche intégrée de gestion des risques de catastrophe et de développement urbain pour lutter contre les risques climatiques dans les villes nigériennes », Blogs de la Banque mondiale. Consulté le: 4 avril 2025. [En ligne]. Disponible sur: https://blogs.worldbank.org/fr/sustainablecities/une-approche-integree-de-gestion-des-risques-de-catastrophe-et-de-developpement

[14] Consulté le: 4 avril 2025. [En ligne]. Disponible sur: https://documents1.worldbank.org/curated/en/099052223060022893/pdf/P1770410ef216f0e609bbf006330e6838ef.pdf

[15] « Se mobiliser en faveur du climat : l’urgence pour la jeunesse camerounaise », Blogs de la Banque mondiale. Consulté le: 4 avril 2025. [En ligne]. Disponible sur: https://blogs.worldbank.org/fr/youth-transforming-africa/se-mobiliser-en-faveur-du-climat-urgence-pour-la-jeunesse-camerounaise

[16] « Vers un avenir vert et résilient pour les Camerounais », World Bank. Consulté le: 4 avril 2025. [En ligne]. Disponible sur: https://www.banquemondiale.org/fr/news/feature/2022/11/04/towards-a-people-centered-green-and-resilient-cameroon

[17] « Mauritanie : La soif d’une vie meilleure | UNICEF ». Consulté le: 4 avril 2025. [En ligne]. Disponible sur: https://www.unicef.org/mauritania/recits/mauritanie-la-soif-dune-vie-meilleure

[18] « Gestion des risques de catastrophes et changement climatique - Mauritanie | Croix-Rouge française ». Consulté le: 4 avril 2025. [En ligne]. Disponible sur: https://www.croix-rouge.fr/la-croix-rouge-francaise-en-mauritanie/gestion-des-risques-de-catastrophes-et-changement-climatique

## R version 4.4.3 (2025-02-28 ucrt)
## Platform: x86_64-w64-mingw32/x64
## Running under: Windows 11 x64 (build 26100)
## 
## Matrix products: default
## 
## 
## locale:
## [1] LC_COLLATE=French_France.utf8  LC_CTYPE=French_France.utf8   
## [3] LC_MONETARY=French_France.utf8 LC_NUMERIC=C                  
## [5] LC_TIME=French_France.utf8    
## 
## time zone: Atlantic/Reykjavik
## tzcode source: internal
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
##  [1] RColorBrewer_1.1-3   cluster_2.1.8        Factoshiny_2.7      
##  [4] FactoInvestigate_1.9 shiny_1.10.0         viridis_0.6.5       
##  [7] viridisLite_0.4.2    ggrepel_0.9.6        corrplot_0.95       
## [10] factoextra_1.0.7     ggplot2_3.5.1        FactoMineR_2.11     
## [13] readr_2.1.5         
## 
## loaded via a namespace (and not attached):
##   [1] Rdpack_2.6.2         gridExtra_2.3        tcltk_4.4.3         
##   [4] rlang_1.1.5          magrittr_2.0.3       shinydashboard_0.7.2
##   [7] compiler_4.4.3       mgcv_1.9-1           vctrs_0.6.5         
##  [10] crayon_1.5.3         pkgconfig_2.0.3      shape_1.4.6.1       
##  [13] fastmap_1.2.0        backports_1.5.0      labeling_0.4.3      
##  [16] utf8_1.2.4           shinyjqui_0.4.1      promises_1.3.2      
##  [19] rmarkdown_2.29       tzdb_0.4.0           nloptr_2.2.0        
##  [22] missMDA_1.19         purrr_1.0.4          xfun_0.51           
##  [25] glmnet_4.1-8         jomo_2.7-6           cachem_1.1.0        
##  [28] jsonlite_1.9.1       flashClust_1.01-2    later_1.4.1         
##  [31] pan_1.9              broom_1.0.7          parallel_4.4.3      
##  [34] R6_2.6.1             stringi_1.8.4        bslib_0.9.0         
##  [37] car_3.1-3            boot_1.3-31          rpart_4.1.24        
##  [40] jquerylib_0.1.4      estimability_1.5.1   Rcpp_1.0.14         
##  [43] iterators_1.0.14     knitr_1.49           httpuv_1.6.15       
##  [46] Matrix_1.7-2         splines_4.4.3        nnet_7.3-20         
##  [49] tidyselect_1.2.1     abind_1.4-8          rstudioapi_0.17.1   
##  [52] yaml_2.3.10          doParallel_1.0.17    codetools_0.2-20    
##  [55] miniUI_0.1.1.1       lattice_0.22-6       tibble_3.2.1        
##  [58] withr_3.0.2          evaluate_1.0.3       survival_3.8-3      
##  [61] ggpubr_0.6.0         pillar_1.10.1        carData_3.0-5       
##  [64] mice_3.17.0          DT_0.33              foreach_1.5.2       
##  [67] reformulas_0.4.0     generics_0.1.3       hms_1.1.3           
##  [70] munsell_0.5.1        scales_1.3.0         minqa_1.2.8         
##  [73] xtable_1.8-4         leaps_3.2            glue_1.8.0          
##  [76] emmeans_1.10.7       scatterplot3d_0.3-44 tools_4.4.3         
##  [79] dendextend_1.19.0    lme4_1.1-36          ggsignif_0.6.4      
##  [82] colourpicker_1.3.0   mvtnorm_1.3-3        grid_4.4.3          
##  [85] tidyr_1.3.1          rbibutils_2.3        colorspace_2.1-1    
##  [88] nlme_3.1-167         Formula_1.2-5        cli_3.6.4           
##  [91] dplyr_1.1.4          gtable_0.3.6         rstatix_0.7.2       
##  [94] sass_0.4.9           digest_0.6.37        farver_2.1.2        
##  [97] htmlwidgets_1.6.4    htmltools_0.5.8.1    lifecycle_1.0.4     
## [100] multcompView_0.1-10  mitml_0.4-5          mime_0.12           
## [103] MASS_7.3-64