Identifying disaster risk factors and hotspots in Africa from spatiotemporal decadal analyses using INFORM data for risk reduction and sustainable development

Eze, Emmanuel; Siegmund, Alexander Article — Published Version Identifying disaster risk factors and hotspots in Africa from spatiotemporal decadal analyses using INFORM data for risk reduction and sustainable development Sustainable Development Provided in Cooperation with: John Wiley & Sons Suggested Citation: Eze, Emmanuel; Siegmund, Alexander (2024) : Identifying disaster risk factors and hotspots in Africa from spatiotemporal decadal analyses using INFORM data for risk reduction and sustainable development, Sustainable Development, ISSN 1099-1719, John Wiley & Sons, Inc., Chichester, UK, Vol. 32, Iss. 4, pp. 4020-4041, https://doi.org/10.1002/sd.2886 This Version is available at: https://hdl.handle.net/10419/306090 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by-nc/4.0/ RESEARCH ARTICLE Identifying disaster risk factors and hotspots in Africa from spatiotemporal decadal analyses using INFORM data for risk reduction and sustainable development Emmanuel Eze 1,2,3 | Alexander Siegmund 1,2 1 Institute of Geography & Heidelberg Center for the Environment (HCE), University of Heidelberg, Heidelberg, Germany 2 Department of Geography –Research Group for Earth Observation ( r geo), UNESCO Chair on Observation and Education of World Heritage & Biosphere Reserve, Heidelberg University of Education, Heidelberg, Germany 3 Geographical and Environmental Education Unit, Department of Social Science Education, University of Nigeria, Nsukka, Nigeria Correspondence Emmanuel Eze, Institute of Geography & Heidelberg Center for the Environment (HCE), University of Heidelberg, Heidelberg, Germany. Email: [email protected]-heidelberg.de Abstract The incidence and magnitude of hazards in Africa are escalating. Extant knowledge base of disaster risk (DR) trends, factors, and hotspots is lacking for the continent. Here we applied random forest machine learning regressions, spatial stratified heterogeneity, and hotspot analyses on INFORM data to identify DR patterns, factors and interactions, and notable risk hotspots. We show that although DR is generally decreasing in Africa, the Eastern, Southern, and Western regions record increasing DR. Physical exposure to floods, epidemics, and violent conflicts are hazard drivers of DR in Africa. Other significant DR drivers are mostly clustered under vulnerable groups and poor infrastructural coping capacities. Human hazards interact with other factors, exhibiting the highest influences on DR. Precisely, 19 out of 53 African countries in this study are DR hotspots. Eritrea is identified as a new hotspot. Targeted policies, resilience building, vulnerability reduction measures and comprehensive sustainability-infused solutions are required for DR reduction and sustainable development in Africa. KEYWORDS Africa, disaster risk assessment, disaster risk reduction, resilience, sustainable development, vulnerability 1|INTRODUCTION The prevalence of both natural and human-induced hazards in Africa has shown a notable increase (Aliyu, 2015). Hazards, when compounded by vulnerability or insufficient coping capabilities, often culminate in disasters characterized by substantial loss of lives and livelihoods (Raju et al., 2022). An overlooked yet crucial aspect in understanding disasters revolves around the evolution of disaster terminology. The distinction between ‘unnatural’disasters in earlier literature (O'Keefe et al., 1976; Wisner et al., 2004) prompts a clear differentiation between hazards and disasters throughout this study. Hazards encompass specific phenomena, whether natural or human-induced, capable of instigating adverse socioeconomic and environmental impacts, including fatalities or injuries. On the other hand, disasters manifest as significant disturbances within a community or society, brought about by the convergence and interactions of hazardous events, exposure, vulnerability, and coping capacity (United Nations General Assembly [UNGA], 2016). Thus, efforts to mitigate disasters, must, according to Rahman and Fang (2019), be based on a comprehensive understanding of all these facets. A comprehensive report by Mizutori and Guha-Sapir (2020) reveals a stark rise in both the frequency and severity of global disaster events. They captured disaster events as surging by 74.48% from 4212 events in the period between 1980 and 1999 to 7348 events Received: 20 July 2023 Revised: 13 December 2023 Accepted: 25 December 2023 DOI: 10.1002/sd.2886 This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. © 2024 The Authors. Sustainable Development published by ERP Environment and John Wiley & Sons Ltd. 4020 Sustainable Development. 2024;32:4020–4041. wileyonlinelibrary.com/journal/sd from 2000 to 2019, with attendant losses of human lives, livelihoods, and ecosystems, amounting to trillions of dollars. These disasterinduced losses are inimical to the achievement of global sustainability frameworks such as the Agenda 2030 and the Sendai Framework for Disaster Risk Reduction [SFDRR] (United Nations International Strategy for Disaster Reduction [UNISDR], 2015). The fundamental goal of disaster risk reduction (DRR), as outlined by UNGA (2016), is to prevent the emergence of new disaster risks (DR), mitigate existing ones, and manage any residual risks to improve societal resilience and the achievement of sustainable development. It is noteworthy that Wen et al. (2023) describe the evolution of the concept of DRR through three phases in the 1990s, 2000s, and 2010s, featuring key paradigms of disaster management, risk management, and resilience management and development. Thus, the current paradigm underscores the critical link between enhancing resilience and sustainable development. The concepts of resilience and development remain subjects of contention, necessitating clarification for collective actions (Park, 2024). In this study, resilience is employed as a rallying call, following Park (2024), to denote the outcomes of disaster risk management (DRM) and DRR initiatives. It emphasizes the fortification of a community or society's structures and functional capacities, enabling them to withstand, assimilate, adjust to, adapt, transform, and swiftly recover from hazard impacts (UNGA, 2016). Such framing of resilience by UNGA (2016) is likely to be understood as the absence of vulnerability or susceptibility of a community to damages from hazards. In fact, prior studies, such as Cardona et al. (2012) and Ismail-Zadeh (2022), highlight vulnerability and exposure as primary drivers of disasters, with natural hazards acting as triggers, and climate change intensifying them. Conversely, other studies like Imperiale and Vanclay (2016,2021), elucidate that the absence of resilience is not solely characterized by vulnerability, as even highly vulnerable communities possess substantial resources contributing to their resilience during crises and disasters. Similarly, Lavell et al. (2012) assert that the inadequacy of capacity is but one dimension within the context of overall vulnerability. This warrants a focus on not just the vulnerability levels of communities but also their resilience levels while considering the multidimensionality of these concepts. Multiple dimensions of disasters numbering six have been synthesized by Imperiale and Vanclay (2021) to include the: (i) nature and attributes of the hazard itself; (ii) social aspects of risks and impacts (including distribution, perception, and experiences); (iii) underlying societal conditions and factors preceding disasters; (iv) capacity of local individuals to learn from past failures and disasters to facilitate sustainability; (v) underlying principles, objectives, and approaches embedded within DRM; and (vi) efficacy of social processes, services, and support systems available to a community prior to and postdisasters. The complexities inherent in disasters stem from a complex interplay of social, environmental, and governance factors, significantly amplifying their impact. Such amplification often leads to disparities in both the distribution and potential impacts of disasters. Previous research, as highlighted by Imperiale and Vanclay (2021), and findings from Machlis et al. (2022), attribute these disparities to the profound social dimensions of vulnerability and varying levels of coping abilities prevalent across diverse regions. Africa, in particular, stands exceptionally vulnerable to the ramifications of natural hazards. Bari and Dessus (2022) identify a total of 13 African countries out of the 15 most vulnerable countries globally. Hazardous events like droughts and floods have further exacerbated poverty in the continent, causing a decline in GDP by 0.7% and 0.4% respectively (Bari & Dessus, 2022). The escalation of poverty suggests a likely simultaneous increase in vulnerabilities across the region. Moreover, a report from The Africa Center for Strategic Studies (2022) reveals a consistent rise in the population affected by disasters, with projections indicating a potential doubling by 2050, reaching nearly 2.5 billion disaster-affected individuals. This concerning trajectory is closely linked to urbanization in hazard-prone regions and increased hazard frequencies driven by climate change (Yaghmaei, 2019). Therefore, tailored DRR strategies are imperative, with a focus on reducing vulnerability dimensions and intentionally improving resilience across different regions within Africa. Given disparities in disaster distribution across regions, it's imperative to evaluate these variations across countries within Africa, the continent most vulnerable to such challenges. Surprisingly, no prior study, to our knowledge, has delved into this crucial area. Hence, our study pioneers the application of multiple spatial statistical analyses, examining a decade's worth of data encompassing 53 African countries. This comprehensive investigation offers fresh perspectives on national-level DR patterns, trends, and focal points, employing spatiotemporal approaches such as spatial stratified heterogeneity (SSH), optimized hotspot and emerging hotspot analyses. Utilizing a meticulously curated INFORM dataset, we consider three fundamental dimensions: hazard & exposure, vulnerability, and coping capacity, consisting of six categories and 18 components—largely aligned with dimensions outlined by Imperiale and Vanclay (2021) barring one aspect (i.e., social aspects of risks and impacts). Our method hinges on empirical analysis but bears practical implications. By presenting nuanced insights into historical DR trends, we aim to furnish datadriven recommendations conducive to directing DRR efforts across the region. Moreover, our exploration of emerging hotspots serves as robust evidence to identify areas necessitating intervention priorities for mitigating DR. 2|CONCEPTUAL FRAMEWORK OF THE STUDY 2.1 |Disaster risks, DRR and sustainable development Natural phenomena, including volcanic eruptions, precipitation patterns resulting in floods and droughts, epidemics, and climatic trends, often transform into so-called natural disasters. While these events, termed hazards, according to (Kelman et al., 2016) are frequently essential for the environment and society, their catastrophic outcomes arise from human-induced vulnerabilities, eroding the natural EZE and SIEGMUND 4021 essence of disasters. Hence, vulnerabilities created by social processes render these events disastrous, necessitating a deeper enquiry. Bailey (2022) simplifies the disaster concept, highlighting hazards and the social system as its fundamental variables. Disasters, stemming from natural hazards, exert profound impacts on development. They exacerbate vulnerabilities, weaken resilience, and strain coping mechanisms (Pal et al., 2021). Furthermore, Bendimerad (2003) enumerates extensive damages to the environment, economy, and human capital, leading to disruptions in developmental programs. Disaster risks (i.e., different kinds of potential losses), emerge from the convergence of hazards, exposure, vulnerabilities, and coping capacities (Imperiale & Vanclay, 2021; Raju et al., 2022; Trogrli c et al., 2022), necessitating comprehensive consideration in DRR strategies. As natural hazards are uncontrollable, focusing on vulnerabilities and coping capacities becomes imperative (Ginige, 2011). Sometimes, vulnerability concepts are insufficiently covered in DR literature (Zhou et al., 2015). Vulnerability, according to Birkmann et al. (2006), comprises susceptibility and coping capacity, which embodies weaknesses affecting community well-being, often intensified by social risks (Imperiale & Vanclay, 2021; Samaraweera, 2024). Coping capacities, vital for recovery, entail mechanisms that shape or challenge recovery in the aftermath of disasters (Birkmann & Wisner, 2006). Thus, lack of coping capacity is also vulnerability. However, rectifying vulnerabilities doesn't guarantee increased coping capacity, as existing strengths within vulnerable communities are often overlooked (Samaraweera, 2024). Recent literature emphasizes vulnerability as the underlying cause of disasters. Kelman (2015) advocates for acknowledging vulnerability as the principal cause of disasters and highlights its significance in global policy frameworks. Moreover, Ginige (2011) views vulnerability as a controllable element within disasters, suggesting that managing vulnerabilities is crucial in disaster mitigation. However, Samaraweera (2024) criticizes the framing of vulnerability, which often equates it with weakness, incapability, and poverty, arguing that such a portrayal overlooks its multifaceted nature, encompassing socio-economic, political, and cultural dimensions. Therefore, addressing vulnerability comprehensively requires a holistic approach considering these diverse factors. Drakes and Tate (2022) delineate social aspects determining vulnerability, including demographics, land tenure, living conditions, and socio-economic status. Hence, specific aspects of vulnerability to be tackled in any community, region or state are known. Also, measures to address vulnerability primarily involve social interventions, albeit supplemented occasionally by technical measures (Kelman et al., 2016). Holistically addressing factors contributing to vulnerability becomes crucial, particularly in policies and actions (Bendimerad, 2003). Unfortunately, there is a lack of studies presenting the trends and levels of vulnerability in African countries. The attempt to assess vulnerability in Africa by Ahmadalipour and Moradkhani (2018) was limited to droughts. Drakes and Tate (2022) confirm that current research examining social vulnerabilities in lower-income nations tends to focus on specific localized areas, limiting the identification of vulnerability factors across wider spheres. Furthermore, insights derived from scientific studies conducted in middle- and high-income countries may not be directly applicable or transferable to lowerincome nations (Drakes & Tate, 2022). This creates a need for a comprehensive understanding of vulnerability in low-income nations, such as Africa, to provide a research representation necessary for designing sustainable and impactful interventions towards vulnerability reduction. Lower-income economies can cope less with hazards, as they lack resources and infrastructure for preparedness (Abdel Hamid et al., 2020). Moreover, disaster losses extend a vicious cycle of poverty (Hallegatte et al., 2020; Salvucci & Santos, 2020). Coping capacities are integral prerequisites for reduced vulnerability and resilience. Recommending the improvement of living standards and protection from hazards for increased resilience, Naheed (2021) defines resilience as the stability and persistence of social systems, involving the enhancement of adaptive capacities through the provision of relevant assets and resources for individuals, communities, or states. Thus, initiatives that strengthen societies to withstand the impact of and cope in the face of disasters will reduce their susceptibility. These initiatives should include capacity building and resource provisions (Acharibasam & Datta, 2024). As a crucial factor for disaster recovery and adaptation, Marin et al. (2021)advocate for the inclusion of such resilience components in DR assessments. Notably, acute shortage of skilled practitioners have been identified as a source of delay in translating good policies into actions (Bang, 2014,2022; Becker & van Niekerk, 2015). Furthermore, Keating et al. (2017) suggest that resilience provides an opportunity to address interconnected social and ecological aspects of DR and development. Designing effective resilience initiatives should be preceded by a deep understanding of various needs, concerns, contexts, and vulnerability components. Subroto and Datta (2024) present such understanding as a prerequisite for disaster governance and resilience. In addition, McBean and Rodgers (2010) advocate for the establishment of national systems, infrastructure, and social networks to bolster resilience, enabling countries to better absorb the impact of natural hazards and thwart their progression into severe disasters. Therefore, the preparedness of nations to mitigate hazards and prevent their escalation into disasters requires strengthening resilience. Consequently, resilience efforts should not be regarded as singular actions but as ongoing processes. Kapucu et al. (2013) urge for continuous, coordinated local planning for resilience as vital for ensuring sustainability. Local planning plays a crucial role in building resilience, as Chipangura et al. (2017) highlight that communities are the focal point of DR concerns. Consequently, reducing vulnerabilities and exposure levels at the community level stands out as the most effective approach to DRR (Jafari et al., 2018; Zhou et al., 2015). Notably, prior research, such as Bang (2014) and Aka et al. (2017), critiques topdown hierarchical structures for mitigating disasters, advocating instead for community-driven bottom-up approaches. Additionally, Jiménez-Aceituno et al. (2020) suggest that proper DRR governance approaches can simultaneously advance both DRR and sustainable development objectives. 4022 EZE and SIEGMUND The literature provides evidence that implementing DRR measures, such as reducing hazard exposure and vulnerability levels, can bolster resilience, subsequently advancing the attainment of sustainable development. According to Bello et al. (2021), disasters' consequences are so profound that development cannot occur sustainably without resilience being an inherent component of development policies. The benefits derived from resilience in managing disasters are intrinsically linked to development. For instance, as highlighted by Naheed (2021), resilience leads to a decrease in fatalities during disasters, diminishes property damages, and safeguards socio-ecological systems. These observed advantages align with the economic, social, and environmental aspects of sustainable development. Furthermore, Keating et al. (2017) previously conceptualized resilience as the capability of a system, community, or society to pursue its social, ecological, and economic development objectives while effectively managing DR in a mutually reinforcing manner. Thus, the widely recognized pillars of sustainable development are captured within the concept of resilience and underscore the synergy between DRR and sustainable development across various domains, encompassing social, economic, and environmental dimensions. Imperiale and Vanclay (2024) establish an explicit connection between DRR, sustainable development, and resilience while emphasizing wellbeing improvement, resources, services, capacities, and skills enhancement as the key to achieving DRR, resilience, and sustainable development. Similarly, Naheed (2021) contends that strategies and investments in DRR are significantly tied to reducing poverty and equitably enhancing social foundations, bearing in mind the needs of future generations. Alike, Akanle et al. (2022) describe poverty as a major concern for development and the success of the SDGs. Moreover, Ginige (2011) asserts that reducing vulnerability stands as the central objective of DRR within the framework of sustainable development. The intricate relationship between DRR, vulnerability reduction, resilience building, and sustainable development requires a careful policy approach to avoid detrimental outcomes. Previous research like Thomalla et al. (2018) highlights the failure of DRR research to acknowledge and address how developmental processes contribute to the fundamental causes of disasters. Thus, questions arise as to what development would solve, rather than exacerbate, the problem of disasters. For instance, earlier research by Naheed (2021) established a connection between unplanned urbanization and heightened vulnerability levels. Therefore, inappropriate development practices tend to amplify, rather than diminish, vulnerability to DR. Acknowledging this reality, Thomalla et al. (2018) advocate for transformative changes towards equitable, resilient, and sustainable development. They propose a need to challenge prevailing values and objectives in existing development practices, critically evaluate the deficiencies in development and DRR approaches, and advocate for radical policy shifts and systemic changes in social systems to mitigate risks and counteract unsustainable development. Therefore, policy gaps emerge concerning DRR and sustainable development. According to Kelman (2015), comprehensive DRR frameworks ought to effectively balance the concepts of hazards, vulnerability, and resilience. Similarly, Imperiale and Vanclay (2024) suggest that development policies, plans, programs, and projects should acknowledge, involve, and empower the processes of social learning and sustainability transformation, which are fundamental to fostering proactive resilience. Advancing research, policy, and practice in addressing vulnerability and resilience for sustainable development mandates a clear understanding of the complexities involved. Kelman (2015) argue that comprehending the long-term causes and consequences of vulnerability and resilience requires insights gleaned from various sources and contexts to develop diverse intervention strategies. Moreover, policies driven by information are imperative to progress both DRR efforts and sustainable development (Chen et al., 2021). Further gaps exist in the lack of long-term analyses of DR in African countries, encompassing the various components involved. This dearth of analysis could potentially hinder the continent's ability to enhance resilience for preparedness and DRR, which are crucial elements for advancing towards risk-informed sustainable development. Previous continental studies in Africa have primarily focused on two major hazards: floods and droughts. Ahmadalipour and Moradkhani (2018) conducted an assessment that evaluated historical drought vulnerability and projected this vulnerability until the end of the century. They identified Egypt, Tunisia, and Algeria as the least vulnerable countries to drought, contrasting Chad, Niger, and Malawi as the most vulnerable in Africa. Similarly, Li et al. (2016) identified Ethiopia, Kenya, Somalia, Tanzania, Nigeria, Libya, and Sudan as countries prone to floods in Africa. They highlighted runoff, per capita GDP, population, and urbanization rate as significant vulnerability factors influencing spatial disparities related to flooding. However, studies encompassing hazards, vulnerability, and coping capacities as interconnected components of disasters are scarce, prompting the necessity for this study. Our research aims to assess historical DR spanning from 2012 to 2022. The objective is to reveal trends, pinpoint significant variables driving these trends, and identify hotspots. Within hotspots, as indicated by Shi et al. (2016), the occurrences of hazards surpass coping capacities, resulting in increased vulnerabilities. Furthermore, Szabo et al. (2016) emphasize that the failure to monitor hotspots could impede progress in both sustainable development initiatives and the achievement of the SDGs. 2.2 |The index for risk management (INFORM) framework The INFORM is a risk model that encompasses three crucial dimensions of hazards & exposure, vulnerability, and lack of coping capacity, consistent with disaster literature. According to the Joint Research Centre [JRC] (2023), physical exposure and severity of specific natural and human-induced hazards are equally weighted and aggregated to compose the Hazard and Exposure dimension. Under the vulnerability dimension, socio-economic vulnerability and vulnerable groups are the components considered in the determination of the final DR index (Figure 1). Lastly, the lack of coping capacity dimension incorporates EZE and SIEGMUND 4023 DRR programs, emphasis on mitigation and preparedness, emergency response, and recovery capabilities of countries' governments. Therefore, the Joint Research Centre (JRC) of the European Commission developed the composite INFORM model comprising 53 indicators to gauge factors of global humanitarian crises and DR, in support of initiatives like the SFDRR, 17 SDGs and the global resilience agenda (Marin-Ferrer et al., 2017). The components that make up the three dimensions of the INFORM model have been carefully selected and curated based on scientific literature justification, robustness, transparency, reliability, global consistency, and scalability (JRC, 2023). The multiplicative equation (Equation 1) emphasizes the equal treatment of the three dimensions—hazards & exposure, vulnerability, and lack of coping capacity—within the INFORM model to generate the resulting DR index, usually overall risk scores out of 10 per country (JRC, 2023). Risk ¼Hazard&Exposure1 3Vulnerability1 3Lack of coping capacity1 3ð1Þ As the first global, open-source, and regularly updated tool providing an evidence-based approach to risk analysis, the INFORM model creates an opportunity for developing a comprehensive understanding of DR across 191 countries (Marin-Ferrer et al., 2017). The utility of the INFORM model as a long-term data pool lies in its potential to stimulate national-, regional-, or global-level proactive DRR initiatives, the priority allocation of resources and improving disaster preparedness based on historical trends. However, the absence of regional or national-scale studies employing the INFORM model within Africa at the time of writing this paper (in 2023) is a significant gap, warranting immediate attention to harness its potential benefits in the continent's DRR strategies. Previous comparative studies, such as those by Beccari (2016), and Visser et al. (2020), have examined the INFORM model alongside other indicator-based assessments. Analyses by Visser et al. (2020) evaluated various criteria including definition precision, handling of uncertainty, data completeness, temporal alignment, and aggregation methods. The findings consistently suggest that the INFORM model demonstrates robustness, validity, reliability, and higher performance in comparison to other models in similar domains. Therefore, the robustness and superior performance of the INFORM model, as indicated by comparative studies across various reliability criteria, endorse its suitability and credibility as an effective tool for comprehensive risk assessment and informed decision-making for DRR, hence its use in this study. The components, categories, and dimensions of the INFORM model (Figure 1) depict the comprehensive coverage of the dataset used in its development. The six categories (i.e., natural hazards, human hazards, socioeconomic vulnerability, vulnerable groups, institutional capacity, and infrastructure capacity) that make up the three dimensions (i.e., hazard and exposure, vulnerability, and lack of coping capacity) perfectly align with concepts incorporated in DR assessments. It has been widely acknowledged that hazards function as triggers for risks, and their damaging impact can vary significantly based FIGURE 1 Components of the INFORM model. Source: Modified from Marzi et al., (2021); Marin-Ferrer et al. (2017). 4024 EZE and SIEGMUND on the level of exposure and vulnerability present in the affected regions (Marin et al., 2021). In the African context, the incidence of both natural and human-induced hazards has escalated (Aliyu, 2015). Natural hazards in Africa, while less frequent, possess the potential to yield higher fatalities and trigger greater population displacement compared to human hazards, as highlighted by Bang (2022). Within the INFORM model, physical exposure to the specified hazards is used for the computation of the natural hazard category. The INFORM model, focusing on the human hazard aspect, incorporates present and anticipated risks associated with societal threats such as violent conflicts encompassing civil wars, unrest, high-intensity crime, and terrorism (Marin-Ferrer et al., 2017). The inclusion of violent conflicts as human hazards is warranted, especially in light of projections by the World Bank (Corral et al., 2020) indicating that by 2030, around two-thirds of the world's extremely impoverished population will inhabit regions characterized by fragility and conflict. This projection poses a significant challenge to the global sustainability agenda encapsulated in the Sustainable Development Goals (SDGs) and threatens to impede the progress already achieved. A recent decadal analysis by Anderson et al. (2021), spanning from 2009 to 2019, reveals that violent conflicts in sub-Saharan Africa, notably in regions like South Sudan and Nigeria, exacerbated food insecurity to a greater extent than droughts or locust swarms. Similarly, Adaawen et al. (2019) found widespread drought-related farmer–herder conflicts and water tensions to precede violent conflicts in the Sahel, Eastern, and Southern Africa. Consequently, it becomes evident that violent conflicts intensify vulnerability levels as well. Furthermore, as indicated by Caso et al. (2023), the co-occurrence of violent conflicts with disasters leads to severe consequences. They present a consistent increase in the occurrence of countries facing both armed conflict and concurrent disasters, underscoring the compounded challenges posed by these dual crises. The INFORM model considers two primary vulnerability categories: socioeconomic vulnerability and vulnerable groups. These categories encompass aspects such as income, inequality, overall wellbeing, and specific groups of individuals more prone to disasters due to inherent or external circumstances. Social factors play a pivotal role in the realm of disasters and risk intensification, as asserted by Bailey (2022) and supported by studies from Samaraweera (2024) and Imperiale and Vanclay (2021). Moreover, the multiplier effect of climate change, poverty, and social insecurity, according to Scheffran et al. (2019), worsens humanitarian crises. Thus, this study's incorporation of these social factors is validated by their potential to exacerbate risks associated with disasters. Notably, the indicators under both vulnerability categories closely correspond to the social aspects, excluding land tenure, outlined as critical determinants of vulnerability by Drakes and Tate (2022). The institutional and infrastructural components constitute the lack of coping capacity dimension within the INFORM model. Marin-Ferrer et al. (2017) integrated governmental initiatives aimed at enhancing resilience, disaster management, and mitigation through organized activities and existing infrastructure. Studies, including the work of McBean and Rodgers (2010), advocate for the establishment of institutional systems and robust infrastructure as crucial elements in fostering resilience and preventing the escalation of hazards into disasters. Incorporating all variables from the INFORM model offers crucial insights that significantly contribute to disaster education and knowledge in Africa. As highlighted by Zhang and Wang (2022), there's currently a limited presence of disaster education publications or research from African countries, signaling a need for enhancement in this area. Recognizing the potential of disaster education, as suggested by AlQahtany and Abubakar (2020), holds promise in elevating disaster awareness and preparedness. This, in turn, can foster essential positive behaviors conducive to DRR and subsequently advance sustainable development. 3|METHODS 3.1 |Data source We obtained the INFORM risk model output for the year 2023 from https://drmkc.jrc.ec.europa.eu/inform-index/. Relying on data spanning from 2012 to 2022, this study extracted component- and category-level data for analysis, consisting of 29 variables per annum (Table 1). Meanwhile, the overall risk index of the INFORM model served as the dependent variable for relevant analyses conducted in this study. The INFORM model is esteemed among the top three indexes in climate risk research, known for its adherence to standardized methodologies and scientific conceptual frameworks (Bornhofen et al., 2019; Garschagen et al., 2021). Additionally, prior research has extensively evaluated the reliability and validity of the INFORM index in DR assessment. For instance, Egawa et al. (2018) underscored the credibility of the INFORM data as a representative DR index. Similarly, Birkmann et al. (2022) affirmed the dataset's reliability, highlighting a remarkable internal consistency and indicator reliability exceeding 90%, hence our adoption of this data for this study. 3.2 |Data preparation Data was obtained in clean XLS format. We then prepared two separate sets of data for the intended long-term analyses. Firstly, the mean value was calculated across the entire dataset for each variable per country within the timeframe of 2012–2022. This was used for the decadal mean analyses in the study. Subsequently, the complete dataset was compiled for each country for the decadal analyses, maintaining the original structure per variable. 3.3 |Data analyses 3.3.1 | Trend analyses To examine the trends in DR, our study employed a combination of descriptive chart plots and the Mann-Kendall monotonic trend analysis method (MK). The descriptive chart plots, generated using the EZE and SIEGMUND 4025 ggplot2 package in R version 4.2.1, were designed as scatter plots with a smoothed line to visually represent the pattern in the data. These visualizations allowed us to illustrate the observed trends effectively. In addition to the chart plots, we conducted the MK using the Kendall package of McLeod and McLeod (2015) to rigorously assess the presence and nature of monotonic trends in the DR data. The MK, introduced by Kendall (1975), is a non-parametric statistical test specifically designed to evaluate whether values in a dataset demonstrate consistent trends of increase, decrease, or stability over time. This method is particularly advantageous as it does not rely on assumptions regarding the distribution of the data, ensuring robustness in trend analysis. Hence, for the DR trend analyses for Africa for the period 2012– 2022, we followed two major steps: first, generating descriptive chart plots through the ggplot2 package in R to visually depict patterns in the data. Secondly, we performed the MK using the Kendall package in R to statistically assess the presence and nature of monotonic trends in the DR dataset. The p-values were computed at a 5% significant level to highlight significant trends, if any. This combined approach allowed for both visual and statistical analyses, providing a comprehensive understanding of DR trends without imposing stringent assumptions on the data's distribution. R scripts used for trend analyses are contained in the Data S1. A positive value from the output means that the trend is increasing, while a negative value means that the trend is decreasing. 3.3.2 | Variable importance analyses This study extensively explored variables' importance through the application of random forest regression analyses. To achieve this, we TABLE 1 Variables used in the study. Variable role Variable name Variable type Purposes Dependent variable INFORM risk Overall risk level MKTA, VIDA, VIDMA, HSA, SSHq Independent variables Physical exposure to earthquakes Natural hazard VIDA, VIDMA Physical exposure to floods Natural hazard VIDA, VIDMA Physical exposure to tsunamis Natural hazard VIDA, VIDMA Physical exposure to tropical cyclones Natural hazard VIDA, VIDMA Physical exposure to droughts Natural hazard VIDA, VIDMA Physical exposure to epidemics Natural hazard VIDA, VIDMA Projected conflict risk Human hazard VIDA, VIDMA Current highly violent conflict intensity Human hazard VIDA, VIDMA Development & deprivation Socioeconomic vulnerability VIDA, VIDMA Inequality Socioeconomic vulnerability VIDA, VIDMA Economic dependency Socioeconomic vulnerability VIDA, VIDMA Uprooted people Vulnerable groups VIDA, VIDMA Health conditions Vulnerable groups VIDA, VIDMA Children U5 Vulnerable groups VIDA, VIDMA Recent shocks Vulnerable groups VIDA, VIDMA Food security Vulnerable groups VIDA, VIDMA Other vulnerable groups Vulnerable groups VIDA, VIDMA DRR Institutional coping capacity VIDA, VIDMA Governance Institutional coping capacity VIDA, VIDMA Communication Institutional coping capacity VIDA, VIDMA Physical infrastructure Infrastructural coping capacity VIDA, VIDMA Access to healthcare Infrastructural coping capacity VIDA, VIDMA Category variables Natural hazards Hazard MKTA, SSHq Human hazards Hazard MKTA, SSHq Vulnerable groups Vulnerability MKTA, SSHq Socio-economic vulnerability Vulnerability MKTA, SSHq Infrastructure Coping capacity MKTA, SSHq Institutional Coping capacity MKTA, SSHq Abbreviations: HSA, hotspot analyses; MKTA, Mann-Kendall trend analyses; VIDA, variable importance decadal analyses; SSHq, spatial stratified heterogeneity q-statistics; VIDMA, variable importance decadal mean analyses. 4026 EZE and SIEGMUND employed the randomForestExplainer package, recognized for its robustness in explaining variable importance within random forest models, as developed by Paluszynska et al. (2020). The analyses were conducted using R version 4.2.1, and the R scripts utilized for these analyses are provided in the Data S1 for reference. Our methodology encompassed a two-fold investigation, using both decadal and decadal mean data of independent and dependent variables in Table 1for all African countries for the period 2012– 2022. To identify the pivotal factors influencing DR, an unsupervised Random Forest (RF) regression consisting of 500 trees was trained. Subsequently, we assessed the distribution of minimal depth derived from the constructed forest through the application of the min_d- epth_distribution function. Furthermore, various other crucial importance measures, including the number of nodes, accuracy decrease (MSE increase), Gini decrease (node purity increase), number of trees, times_a_root, and p-value, were derived. To delve deeper into the statistical aspects of the methodology we employed to determine top important DR variables in this study, Paluszynska (2017) provides extensive analysis worth exploring. In summary, the methodology used for ascertaining important variables for both decadal and decadal mean DR factors through the randomForestExplainer package encompassed the following steps: Step 1: Training of the data with a randomForest classifier comprising 500 trees. Step 2: Utilizing the created forest with the min_depth_distribu- tion function to acquire the distribution of minimal depth. A total of seven measures of importance –mean minimal depth, number of nodes, accuracy decrease (MSE increase), Gini decrease (node purity increase), number of trees, times_a_root, and p-values are obtained from our analyses and used in the next step. Step 3: Plotting the top ten important variables alongside the distribution of minimal depth. Step 4: Visualizing multi-way importance with mean squared error post-permutation and the increase in the node purity index (y-axis). 3.3.3 | Spatial stratified heterogeneity (SSH) models Spatial Stratified Heterogeneity (SSH) models serve to unravel associations between geographical attributes by comparing variations in regional and global data (Wang et al., 2010; Wang et al., 2016). We conducted our analysis leveraging the Geographical detector tool, a widely acclaimed method for assessing the power of determinants and attributions for SSH (Song, 2021). We sought to identify and measure the heterogeneity with spatial strata of our DR datasets, assess the coupling between factors Y and X without assuming linearity of their association, and lastly determine the interaction between explanatory factors X1 to X6 (corresponding to the six category-level variables of DR, i.e., natural hazards, human hazards, vulnerable groups, socio-economic vulnerability, Infrastructure, and institutional) and a response factor of Y (DR index). Three major steps were followed in the SSH model performed in this study: Step 1: mapping response variable Y in strata according to X using the risk detector function; Step 2: using the factor detector q-statistic to measure the degree of spatial stratified heterogeneity of a variable Y if Y is stratified by itself; and the determinant power of an explanatory variable X on Y if Y is stratified by X; Step 3: employing the interaction detector to reveal whether the risk factors X1 to X6 have an interactive influence on a response variable Y. Using the decadal mean value of the six category-level variables as X1 to X6 and the INFORM DR model as Y, we obtained three crucial results: the risk detector, factor detector (q-statistic), and interaction detector. The risk detector illustrates how the response factor Y varies across strata defined by X; the factor detector assesses the level of spatial stratified heterogeneity within a factor Y when Y is stratified by itself. Furthermore, it quantifies the determinant power of an explanatory factor X on Y when Y is stratified by X. The interaction detector scrutinizes whether risk factors like X1 to X6 exhibit an interactive influence on the response factor Y. Readers seeking indepth statistical insights into the SSH are directed to Wang et al. (2016) for further details. 3.4 |Hotspot analyses For our analysis, we employed two fundamental techniques, namely the optimized hotspot analyses and emerging hotspot analyses, using ArcGIS Pro version 3.0.2. The Hot Spot Analysis tool in ArcGIS employs the Getis-Ord Gi* statistic (Getis & Ord, 1992) to compute z-scores and p-values for each feature in a dataset, revealing spatial clusters where features with either high or low attribute values tend to cluster together (Esri, 2023a). Additional results from the emerging hotspot analyses incorporate hotspot bin classification based on the MK accounting for temporal dimensions of the data. Consequently, these complementary methods contribute to identifying significant spatial patterns of DR in Africa. They highlight areas with high or low values concerning neighboring features and describe the trends of DR per country based on the DR index spanning from 2012 to 2022. This combined approach provides a comprehensive understanding of the spatial distribution and temporal trends of DR within the dataset. These methodologies are widely acknowledged for their efficacy in pattern analysis (Khan et al., 2022) and we utilized them to unravel the spatial and temporal variations, patterns, and emerging hotspots of DR across Africa. While the optimized hotspot analysis serves to identify spatial relationships and statistically significant clusters of DR hotspots and cold spots within the African region, the emerging hotspot analysis adds a temporal dimension by scrutinizing the evolving patterns of DR per country during the study period (2012–2022). The temporal aspects of the emerging hotspot analysis enabled us to capture the dynamic nature of DR hotspots and cold spots, providing insights into the changing patterns over time in the period under consideration. Together, both analyses reveal countries that demand focused attention and urgent intervention measures. EZE and SIEGMUND 4027 assessment highlighted Chad and Niger as highly vulnerable countries in Africa, correlating with our findings of increasing DR in these regions and their designation as hotspots. Similarly, Li et al.'s (2016) Africa-wide flood disaster assessment identified Ethiopia, Sudan, and United Republic of Tanzania as among the most affected countries, consistent with our identification of these nations as experiencing rising DR and as hotspots. The varying levels and trends of DR identified in our study align with the risk distribution dimension expounded by Imperiale and Vanclay (2021). The variability observed in DR levels and hotspots across different regions and the fluctuating trends over time resonate with the concept of risk distribution as an essential component in understanding the spatial and temporal dynamics of risks. The declining trend in infrastructural coping capacity might signify improvements in infrastructure development, potentially attributed to increased investments or initiatives targeting infrastructure resilience, possibly influenced by global frameworks like the UN's Agenda 2030 and SFDRR. These initiatives could be translating into specific TABLE 5 Results of stratified detected risks. Core disaster risk factors Very low (0) Very low (1) Low (2) Medium (3) High (4) Very high (5) Human hazard Continental 1.5 2.86 3.14 3.43 4.55 5 Central 2 2.5 4 4 4.67 5 Eastern 1 3 3.33 3 4.75 5 Northern *3.5 2 3 3.5 * Southern *2.71 *3.33 ** Western *2.89 3 4 5 5 Infrastructure Continental *1.5 2 2.7 3.39 4.05 Central ***2 3.67 4.75 Eastern *1.5 ** 3.75 4.43 Northern **2 3.25 4 * Southern ** 2.67 2.75 3.33 Western **2 3.33 3.63 Institutional Continental **2.91 3.42 3.33 4.09 Central ** 3 3 4.4 Eastern **2.75 4 4.25 4.33 Northern ***3.25 2 4 Southern **3 2.75 3 * Western **3 3.86 3 3 Natural hazard Continental *2 2.33 3.6 3.64 * Central *2 2 4.29 * Eastern ***3.63 4 * Northern ***3 3.33 * Southern **2.33 3 3.33 * Western **2.5 3.54 * Socio-economic vulnerability Continental **2.5 2.89 3.56 4.5 Central ** 2.5 4 5 Eastern **1.5 4 5 Northern **33 4 * Southern ***2.5 2.86 4 Western ***3.25 3.33 4 Vulnerable groups Continental *1.67 2.4 2.94 3.44 4.72 Central **2 3 4 4.6 Eastern *1.5 *3.5 3.33 4.83 Northern *2334* Southern ***2.33 3 4 Western **2.5 3 3.75 5 *No strata membership. 4034 EZE and SIEGMUND improvements in physical infrastructures and access to healthcare, as captured in the INFORM model. Conversely, the escalating lack of institutional coping capacity could mirror challenges in governance, resource allocation, corruption, or the efficacy of disaster management policies. However, our study lacks additional data to delve deeper into the specific reasons for these trends beyond the indicators within the INFORM model. Additionally, it's important to note that the data analyzed in this paper pertains to 2012–2022, a period preceding recent disasters occurring between April and September 2023. These disasters, stemming from natural hazards such as the floods in Libya, DR Congo, and Rwanda, the earthquake in Morocco, wildfires in Algeria, and the cyclone in Mozambique, are not accounted for in our analysis. These recent events might have influenced the evolving landscape of DR in Africa, potentially altering the trends and dynamics observed in our study. Their omission underscores a limitation in our analysis, indicating the need for further assessment and updated data to comprehensively capture the contemporary scenario of DR in the region. 5.2 |Disaster risk drivers in Africa Our findings pinpointed key drivers of DR, including epidemics, flood exposure, projected conflict risk, development indicators, displacement, child health, food security, governance, infrastructure, healthcare access, and communication. Approximately 43% of the countries assessed fell within high and very high clusters of DR levels. Our continental-wide findings revealed that vulnerable groups and humaninduced hazards collectively accounted for more than 70% of the explanation for the DR index. Evidently, the interaction between human-induced hazards and factors in other categories exerted the most significant influence on DR across the continent. The interaction between human-induced hazards, encompassing current highly violent conflicts and projected conflict risks, emerged as the primary amplifier of DR across the African continent. This concurs with projections made by Corral et al. (2020), highlighting a grim outlook should violent conflicts persist in the region. Notably, violent conflicts not only impede sustainable development but also exacerbate DR by escalating TABLE 6 Results of SSH q-statistics and factor significance. Regions Human hazard Infrastructure Institutional Natural hazard Socio-economic vulnerability Vulnerable groups Central 0.90 0.88** 0.38 0.70* 0.44 0.90* Eastern 0.83* 0.66** 0.30 0.02 0.68 0.88*** Northern 0.65 0.74 0.74 0.05 0.29 1*** Southern 0.28 0.28 0.05 0.53 0.53 0.77 Western 0.92* 0.21 0.23 0.16 0.07 0.72 Continental 0.70*** 0.42*** 0.14 0.20 0.28 0.75*** Note: Significance of q: *** above 0.001 level, ** at 0.01 level, * at 0.05 level. TABLE 7 Interaction results of core factors of disaster risk. Interacting factors Regions Central Eastern Southern Western Continental Human hazard \Infrastructure 1 0.95 0.46 1 0.86 Human hazard \Institutional 1 1 0.46 0.96 0.80 Human hazard \Natural hazard 0.95 0.98 e-n 0.64 0.96 0.84 Human hazard \Socio-economic vulnerability 0.91 0.94 0.74 0.96 0.85 Human hazard \Vulnerable groups 0.95 0.94 0.79 0.96 0.85 Infrastructure \Institutional 0.96 0.74 0.59 e-n 0.46 0.52 Infrastructure \Natural hazard 0.89 0.70 0.84 0.30 0.55 Infrastructure \Socio-economic vulnerability 0.95 0.73 0.61 0.36 0.50 Infrastructure \Vulnerable groups 1 0.93 0.84 0.80 0.80 Institutional \Natural hazard 0.75 0.49 e-n 0.84 e-n 0.41 0.35 Institutional \Socio-economic vulnerability 0.79 0.76 0.61 0.70 e-n 0.54 e-n Institutional \Vulnerable groups 0.94 0.98 0.84 1 0.82 Natural hazard \Socio-economic vulnerability 0.91 0.73 0.84 0.31 0.45 Natural hazard \Vulnerable groups 0.91 0.93 1 0.78 0.82 Socio-economic vulnerability \Vulnerable groups 0.92 0.89 0.84 0.84 0.78 Note: The model returned no results for the computation of factors' interaction effects for Northern Africa. e-n =factors have an ‘enhance, non-linear’ interaction, all others are bi-enhance interaction. EZE and SIEGMUND 4035 FIGURE 5 Map of African countries showing: (a) DR classes based on decadal mean analyses of INFORM risk index (b) optimized Hotspots analyses results (c) emerging hotspots of DRs. 4036 EZE and SIEGMUND displacement, adversely impacting vulnerable groups, and intensifying food insecurity, as illuminated by Anderson et al. (2021). Among the 11 key DR drivers in Africa identified in our study, only two are associated with hazards, while the rest pertain to vulnerability and lack of coping capacity dimensions. This aligns with Bailey's (2022) conceptualization of disasters as comprising hazards and the social system. These dimensions encompass weaknesses capable of exacerbating social risks and affecting the wellbeing of countries, as echoed by Samaraweera (2024) and Imperiale and Vanclay (2021). Encouragingly, the influence of natural hazards on DR in Africa seems lower, providing an opportunity for addressing controllable social drivers emphasized by Ginige (2011). Our findings bridge a gap highlighted by Drakes and Tate (2022), addressing the scarcity of large-scale research into social vulnerabilities in lower-income nations. The identified social vulnerability drivers could guide interventions aimed at reversing these trends. Failure to mitigate vulnerability may lead to increased future disaster losses, potentially perpetuating the poverty cycle, as suggested by Hallegatte et al. (2020) and Salvucci and Santos (2020). This emphasizes the interconnectedness of poverty reduction, equitable resource distribution, DRR, and sustainable development advocated by Naheed (2021), as well as the impact of poverty on impeding development and the success of SDGs, as emphasized by Akanle et al. (2022). Furthermore, our findings provide insights into the long-term causes and consequences of vulnerability, aligning with Kelman (2015) assertion that such insights are crucial for developing diverse intervention strategies to build resilience. Information-driven policies, as highlighted by Chen et al. (2021), are imperative for DRR and sustainable development. Governance, identified as a DR driver, underscores the necessity for capacity building to enhance the effectiveness of human resources (Acharibasam & Datta, 2024), addressing concerns raised by Becker and van Niekerk (2015), Bang (2014,2022), about the deficiency of skilled DRR practitioners in Africa, which leads to delay in the translation of policies into actions. Lastly, the identified DR drivers emphasize the need, as advocated by Imperiale and Vanclay (2016,2021), for deliberate efforts to reduce vulnerability while simultaneously increasing resilience. For example, addressing vulnerability drivers such as child health does not inherently provide healthcare access, a coping capacity driver of DR, highlighting the importance of a multifaceted approach. However, it's essential to note that the coarse national-level data used in this study might not capture community-level variations in vulnerabilities and coping capacities. Therefore, community-level studies across African states and regions are crucial to fully optimize the potential of such studies, enabling the most effective DRR approaches, as advocated by Jafari et al. (2018) and Zhou et al. (2015), and the implementation of robust DRR governance, as highlighted by Jiménez-Aceituno et al. (2020). 6|CONCLUSION In examining DR trends, drivers and hotspots in Africa, our study unveiled various outcomes. The results from our MK trend tests indicate an overall decrease in DR across the continent over the studied period, albeit with regional variations. Specifically, the patterns of DR in Central and Northern Africa align with the continental trend, showing a decline. In contrast, Eastern, Western, and Southern Africa exhibit increasing trends in DR during the same period. Notably, Eritrea emerged as a new hotspot. While Burundi, Chad, Djibouti, Niger, Rwanda, Somalia, and Uganda, exhibited consecutive hotspot trends; Central African Republic and Kenya intensified in DR, and Ethiopia, South Sudan, and Sudan remained persistent hotspots. These findings indicate the evolving nature of DR and highlight specific countries that require more focus and tailored interventions. Additionally, our identification of vulnerability and lack of coping capacity dimensions as predominant drivers of DR corroborates existing research. However, the intricate interplay between humaninduced hazards related to violent conflicts and all other vulnerability and coping capacity factors form part of our primary contribution to understanding DR in Africa and warrants the consideration for multifaceted DRR strategies. Moreover, while recognizing the commendable declining trend in infrastructural deficiencies, persistent vulnerabilities and inadequate institutional coping capacity emphasize the imperative for sustained policy interventions. Targeted strategies should be focused on simultaneously addressing these persisting vulnerabilities and building resilience. Strategies such as poverty reduction approaches, effective community-level bottom-up DRR governance and increased investments in capacity building initiatives are pivotal for mitigating risks and achieving sustainable development. Future community-level studies to complement our national-level analyses are required to enable more tailored and effective DRR strategies. Moreover, findings such as the conflicting outcomes of the smoothened plots and the MK trend tests can be easily avoided. Also, contextual policy actions can be founded on such community-level research. Furthermore, DR assessments ought to be ongoing and updated to ensure evolving and recent risk scenarios are adequately captured. We, therefore, present a compelling call for coordinated efforts with local communities, policies, actions, and further research to address the multifaceted nature of DR in Africa to fortify resilience, and reduce vulnerabilities to achieve sustainable development in the continent. 6.1 |Limitations of the study The study's design was significantly influenced by the limited availability of long-term data on DR dynamics in Africa. Despite the preference for designing community-level DR assessments, the absence of prior empirical assessments at lower scales constrained the scope of this study. Likewise, the scarcity of local studies on DR limits the validation of findings on lower-level scales. To address this, we strongly advocate for in-depth case studies on specific hazard-induced disasters that provide deeper insights into DR at local scales in Africa. Such studies could supplement our findings by offering contextspecific knowledge on social dimensions of risks and resilience, like the insights provided by Imperiale and Vanclay (2016,2021), EZE and SIEGMUND 4037 Samaraweera (2024), Fransen et al. (2024) and other local research captured in this work. Furthermore, our study utilized the INFORM index data, acknowledging the providers' identification of methodological and data limitations (Marin-Ferrer et al., 2017). Methodologically, the use of proxies, such as malaria mortality rates for malaria prevalence, might limit the representativeness of the findings. Additionally, incomplete data, reliance on self-measured Hyogo self-assessment reports from countries, and the static nature of natural hazard data necessitate caution in interpreting the results. It's crucial to recognize these limitations and their potential impact on the study's conclusions. Moreover, our study excludes the recent hazardous events in Africa within 2023, which emphasizes the necessity for ongoing and updated DR assessments to capture the ever-evolving landscape of DR in the continent. Also, while our smoothened trend plots offer insights into DR trends, contrasting outcomes yielded by the MK results warrant careful interpretation. The divergence in results signifies the complexity of trend analyses and underscores the need for cautious interpretation of the study's findings. Despite these limitations, our study offers a distinctive panoramic view of DR trends, influential factors, and hotspots across Africa. By uncovering these insights, our research presents actionable opportunities for DRR strategies and sustainable development initiatives. ACKNOWLEDGMENTS Author Emmanuel Eze extends heartfelt appreciation to the German Academic Exchange Service (DAAD) for their generous scholarship stipends supporting his doctoral research discussed in this paper. Emmanuel Eze also acknowledges conversations with Chinazaekpere Peace, which proved instrumental in honing his ideas. Additionally, open-access publication has been made possible through the University of Heidelberg Project Deal Agreement. Open Access funding enabled and organized by Projekt DEAL. CONFLICT OF INTEREST STATEMENT The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. DATA AVAILABILITY STATEMENT The datasets analyzed during this study are freely accessible at https://drmkc.jrc.ec.europa.eu/inform-index/. ORCID Emmanuel Eze https://orcid.org/0000-0003-2007-2696 REFERENCES Abdel Hamid, H. T., Wenlong, W., & Qiaomin, L. (2020). Environmental sensitivity of flash flood hazard using geospatial techniques. Global Journal of Environmental Science and Management,6(1), 31–46. https:// doi.org/10.22034/GJESM.2020.01.03 Acharibasam, J. B., & Datta, R. (2024). Enhancing community resilience to climate change disasters: Learning experience within and from sub- Saharan black immigrant communities in western Canada. Sustainable Development,32(2), 1401–1411. https://doi.org/10.1002/sd.2677 Adaawen, S., Rademacher-Schulz, C., Schraven, B., & Segadlo, N. (2019). Drought, migration, and conflict in sub-Saharan Africa: What are the links and policy options? Current Directions in Water Scarcity Research, 2,15–31. https://doi.org/10.1016/B978-0-12-814820-4.00002-X Africa Center for Strategic Studies. (2022). Cyclones and more frequent storms threaten Africa. Ahmadalipour, A., & Moradkhani, H. (2018). Multi-dimensional assessment of drought vulnerability in Africa: 1960–2100. Science of the Total Environment,644, 520–535. https://doi.org/10.1016/j.scitotenv.2018. 07.023 Aka, F. T., Buh, G. W., Fantong, W. Y., Zouh, I. T., Djomou, S. L. B., Ghogomu, R. T., Gibson, T., Marmol del, M.A., Sigha, L.N. & Ohba, T. (2017). Disaster prevention, disaster preparedness and local community resilience within the context of disaster risk management in Cameroon. Natural Hazards,86,57–88. https://doi.org/10.1007/ s11069-016-2674-5 Akanle, O., Kayode, D., & Abolade, I. (2022). Sustainable development goals (SDGs) and remittances in Africa. Cogent Social Science,8(1), 2037811. https://doi.org/10.1080/23311886.2022.2037811 Aliyu, A. (2015). Management of disasters and complex emergencies in Africa: The challenges and constraints. Annals of African Medicine, 14(3), 123–131. https://doi.org/10.4103/1596-3519.149894 AlQahtany, A. M., & Abubakar, I. R. (2020). Public perception and attitudes to disaster risks in a coastal metropolis of Saudi Arabia. International Journal of Disaster Risk Reduction,44, 101422. https://doi.org/10. 1016/j.ijdrr.2019.101422 Anderson, W., Taylor, C., McDermid, S., Ilboudo-Nébié, E., Seager, R., Schlenker, W., Cottier, F., De Sherbinin, A., Mendeloff, D. & Markey, K. (2021). Violent conflict exacerbated drought-related food insecurity between 2009 and 2019 in sub-Saharan Africa. Nature Food, 2(8), 603–615. https://doi.org/10.1038/s43016-021-00327-4 Bailey, E. (2022). Disaster risk reduction and management: Recalling the need for paradigm shift in definition. Journal of Geoscience and Environment Protection,10(6), 86–105. https://doi.org/10.4236/gep.2022. 106006 Bang, H. N. (2014). General overview of the disaster management framework in Cameroon. Disasters,38(3), 562–586. https://doi.org/10. 1111/disa.12061 Bang, H. N. (2022). A concise appraisal of Cameroon's Hazard risk profile: Multi-Hazard inventories, causes, consequences and implications for disaster management. GeoHazards,3(1), 55–87. https://doi.org/10. 3390/geohazards3010004 Bari, M., & Dessus, S. (2022). Adapting to natural disasters in Africa. https://www.ifc.org/wps/wcm/connect/775d1c2f-a9f3-4b7d-b0d7- 72738b42e3b8/Working-Paper-Adapting-to-Natural-Disasters-in- Africa.pdf?MOD=AJPERES&CVID=ohpHufW Beccari, B. (2016). A comparative analysis of disaster risk, vulnerability and resilience composite indicators. PLoS Currents,8.https://doi.org/10. 1371/currents.dis.453df025e34b682e9737f95070f9b970 Becker, P., & van Niekerk, D. (2015). Developing sustainable capacity for disaster risk reduction in southern Africa. In Hazards, risks, and disasters in society (pp. 63–78). Academic Press. https://doi.org/10.1016/ B978-0-12-396451-9.00005-6 Bello, O., Bustamante, A., & Pizarro, P. (2021). Planning for disaster risk reduction within the framework of the 2030 agenda for sustainable development. Project documents (LC/TS.2020/108), Santiago, Economic Commission for Latin America and the Caribbean (ECLAC). https://repositorio.cepal.org/server/api/core/bitstreams/ae6fe59f- e288-431b-8edd-7cbe1f760c8d/content Bendimerad, F. (2003). Disaster risk reduction and sustainable development. In World Bank seminar on the role of local governments in reducing the risk of disasters, held in Istanbul, Turkey (Vol. 28, pp. 57–75) https:// 4038 EZE and SIEGMUND citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi= 721929932392f6b5d9bd4861f615d8554dade5e1 Birkmann, J., Hettige, S., & Fernando, N. (2006). Measuring vulnerability in Sri Lanka at the local level. In J. Birkmann (Ed.), Measuring vulnerability to natural hazards: Towards disaster resilient societies (pp. 329–356). United Nations University Press. Birkmann, J., Jamshed, A., McMillan, J. M., Feldmeyer, D., Totin, E., Solecki, W., Ibrahim, Z. Z., Roberts, D., Kerr, R. B., Poertner, H.O. & Pelling, M. (2022). Understanding human vulnerability to climate change: A global perspective on index validation for adaptation planning. Science of the Total Environment,803, 150065. https://doi.org/ 10.1016/j.scitotenv.2021.150065 Birkmann, J., & Wisner, B. (2006). Measuring the un-measurable: The challenge of vulnerability. Studies of the University research counsel education publication series. https://d-nb.info/1029694141/34 Bornhofen, E., Ramires, T. G., Bergonci, T., Nakamura, L. R., & Righetto, A. J. (2019). Associations between global indices of risk management and agricultural development. Agricultural Systems,173, 281– 288. https://doi.org/10.1016/j.agsy.2019.03.006 Cardona, O. D., van Aalst, M. K., Birkmann, J., Fordham, M., McGregor, G., Perez, R., Pulwarty, R. S., Schipper, E. L. F., & Sinh, B. T. (2012). Determinants of risk: Exposure and vulnerability. In: Managing the risks of extreme events and disasters to advance climate change adaptation. In C. B. Field, V. Barros, T. F. Stocker, D. Qin, D. J. Dokken, K. L. Ebi, M. D. Mastrandrea, K. J. Mach, G.-K. Plattner, S. K. Allen, M. Tignor, & P. M. Midgley (Eds.), A special report of working groups I and II of the intergovernmental panel on climate change (IPCC) (pp. 65–108). Cambridge University Press. https://www.ipcc.ch/site/assets/uploads/ 2018/03/SREX-Chap2_FINAL-1.pdf Caso, N., Hilhorst, D., Mena, R., & Papyrakis, E. (2023). Does disaster contribute to armed conflict? A quantitative analysis of disaster–conflict co-occurrence between 1990 and 2017. International Journal of Development Issues,23.https://doi.org/10.1108/IJDI-01-2023-0015 Chen, F., Shirazi, Z., & Wang, L. (2021). Building scientific capacity in disaster risk reduction for sustainable development. Cultures of Science, 4(1), 40–54. https://doi.org/10.1177/20966083211017330 Chipangura, P., Van Niekerk, D., & Van Der Waldt, G. (2017). Disaster risk problem framing: Insights from societal perceptions in Zimbabwe. International Journal of Disaster Risk Reduction,22, 317–324. https:// doi.org/10.1016/j.ijdrr.2017.02.012 Corral, P., Irwin, A., Krishnan, N., Mahler, D. G., & Vishwanath, T. (2020). Fragility and conflict: On the front lines of the fight against poverty. No. 146627 (pp. 1–116). World Bank. https://doi.org/10.1596/978-1- 4648-1540-9 Drakes, O., & Tate, E. (2022). Social vulnerability in a multi-hazard context: A systematic review. Environmental Research Letters,17(3), 033001. https://doi.org/10.1088/1748-9326/ac5140 Egawa, S., Jibiki, Y., Sasaki, D., Ono, Y., Nakamura, Y., Suda, T., & Sasaki, H. (2018). The correlation between life expectancy and disaster risk. Journal of Disaster Research,13(6), 1049–1061. https://doi.org/10.20965/ jdr.2018.p1049 Esri. (2023a). How optimized hot spot analysis works. https://pro.arcgis. com/en/pro-app/3.0/tool-reference/spatial-statistics/how-optimized- hot-spot-analysis-works.htm Esri. (2023b). How emerging hot spot analysis works. https://pro.arcgis. com/en/pro-app/3.0/tool-reference/space-time-pattern-mining/ learnmoreemerging.htm Fransen, J., Hati, B., Simon, H. K., & van Stapele, N. (2024). Adaptive governance by community-based organisations: Community resilience initiatives during Covid-19 in Mathare, Nairobi. Sustainable Development, 32(2), 1471–1482. https://doi.org/10.1002/sd.2682 Garschagen, M., Doshi, D., Reith, J., & Hagenlocher, M. (2021). Global patterns of disaster and climate risk—An analysis of the consistency of leading index-based assessments and their results. Climatic Change, 169(1–2), 11. https://doi.org/10.1007/s10584-021-03209-7 Getis, A., & Ord, J. K. (1992). The analysis of spatial association by use of distance statistics. Geographical Analysis,24(3), 189–206. https://doi. org/10.1111/j.1538-4632.1992.tb00261.x Ginige, K. (2011). Disaster risk reduction and its relationship with sustainable development. In Amaratunga, D., & Haigh, R. (Eds.), Post-disaster Reconstruction of the Built Environment: Rebuilding for Resilience (pp. 287-303.). Wiley-Blackwell. Hallegatte, S., Vogt-Schilb, A., Rozenberg, J., Bangalore, M., & Beaudet, C. (2020). From poverty to disaster and back: A review of the literature. Economics of Disasters and Climate Change,4, 223–247. https://doi. org/10.1007/s41885-020-00060-5 Imperiale, A. J., & Vanclay, F. (2016). Experiencing local community resilience in action: Learning from post-disaster communities. Journal of Rural Studies,47, 204–219. https://doi.org/10.1016/j.jrurstud.2016. 08.002 Imperiale, A. J., & Vanclay, F. (2021). Conceptualizing community resilience and the social dimensions of risk to overcome barriers to disaster risk reduction and sustainable development. Sustainable Development, 29(5), 891–905. https://doi.org/10.1002/sd.2182 Imperiale, A. J., & Vanclay, F. (2024). Re-designing social impact assessment to enhance community resilience for disaster risk reduction, climate action and sustainable development. Sustainable Development, 32(2), 1571–1587. https://doi.org/10.1002/sd.2690 Ismail-Zadeh, A. (2022). Natural hazards and climate change are not drivers of disasters. Natural Hazards,111, 2147–2154. https://doi.org/10. 1007/s11069-021-05100-1 Jafari, A. J., Baba, M., & Dowlati, M. (2018). Disaster risk assessment in health centers of Iran University of Medical Sciences in functional, non structural & structural components during 2015–2016. Iran Occupational Health,15(1), 76–85. Jiménez-Aceituno, A., Peterson, G. D., Norström, A. V., Wong, G. Y., & Downing, A. S. (2020). Local lens for SDG implementation: Lessons from bottom-up approaches in Africa. Sustainability Science,15, 729– 743. https://doi.org/10.1007/s11625-019-00746-0 Joint Research Centre [JRC]. (2023). INFORM methodology disaster risk management knowledge Centre: European Commission. Accessed December 8, 2023. Kapucu, N., Hawkins, C. V., & Rivera, F. I. (2013). Disaster preparedness and resilience for rural communities. Risk, Hazards & Crisis in Public Policy,4(4), 215–233. https://doi.org/10.1002/rhc3.12043 Keating, A., Campbell, K., Mechler, R., Magnuszewski, P., Mochizuki, J., Liu, W., Szoenyi, M., & McQuistan, C. (2017). Disaster resilience: What it is and how it can engender a meaningful change in development policy. Development and Policy Review,35(1), 65–91. https://doi.org/10. 1111/dpr.12201 Kelman, I. (2015). Climate change and the Sendai framework for disaster risk reduction. International Journal of Disaster Risk Science,6, 117– 127. https://doi.org/10.1007/s13753-015-0046-5 Kelman, I., Gaillard, J. C., Lewis, J., & Mercer, J. (2016). Learning from the history of disaster vulnerability and resilience research and practice for climate change. Natural Hazards,82, 129–143. https://doi.org/10. 1007/s11069-016-2294-0 Kelman, I., Gaillard, J. C., & Mercer, J. (2015). Climate change's role in disaster risk reduction's future: Beyond vulnerability and resilience. International Journal of Disaster Risk Science,6,21–27. https://doi.org/ 10.1007/s13753-015-0038-5 Kendall, M. G. (1975). Rank correlation methods (4th ed.). Griffin. Khan, S. D., Gadea, O. C., Tello Alvarado, A., & Tirmizi, O. A. (2022). Surface Deformation Analysis of the Houston Area Using Time Series Interferometry and Emerging Hot Spot Analysis. Remote Sensing, 14(15), 3831. Lavell, A., Oppenheimer, M., Diop, C., Hess, J., Lempert, R., Li, J., Muir- Wood, R., & Myeong, S. (2012). Climate change: New dimensions in disaster risk, exposure, vulnerability, and resilience. In C. B. Field, V. Barros, T. F. Stocker, D. Qin, D. J. Dokken, K. L. Ebi, M. D. EZE and SIEGMUND 4039 Mastrandrea, K. J. Mach, G.-K. Plattner, S. K. Allen, M. Tignor, & P. M. Midgley (Eds.), Managing the risks of extreme events and disasters to advance climate change adaptation. A Special Report of Working Groups I and II of the. Li, C. J., Chai, Y. Q., Yang, L. S., & Li, H. R. (2016). Spatio-temporal distribution of flood disasters and analysis of influencing factors in Africa. Natural Hazards,82, 721–731. https://doi.org/10.1007/ s11069-016-2181-8 Machlis, G. E., Román, M. O., & Pickett, S. T. (2022). A framework for research on recurrent acute disasters. Science Advances,8(10), eabk2458. https://doi.org/10.1126/sciadv.abk2458 Marin, G., Modica, M., Paleari, S., & Zoboli, R. (2021). Assessing disaster risk by integrating natural and socio-economic dimensions: A decisionsupport tool. Socio-Economic Planning Sciences,77, 101032. https:// doi.org/10.1016/j.seps.2021.101032 Marin-Ferrer, M., Vernaccini, L., & Poljansek, K. (2017). Index for risk management INFORM concept and methodology report —Version 2017, EUR 28655 EN. https://doi.org/10.2760/094023 Marzi, S., Mysiak, J., Essenfelder, A.H., Pal, J.S., Vernaccini, L., Mistry, M.N., Alfieri, L., Poljansek, K., Marin-Ferrer, M., & Vousdoukas, M. (2021). Assessing future vulnerability and risk of humanitarian crises using climate change and population projections within the INFORM framework. Global Environmental Change,71, 102393. https://doi.org/10. 1016/j.gloenvcha.2021.102393 McBean, G., & Rodgers, C. (2010). Climate hazards and disasters: The need for capacity building. Wiley Interdisciplinary Reviews: Climate Change, 1(6), 871–884. https://doi.org/10.1002/wcc.77 McLeod, A. I., & McLeod, M. A. (2015). Package ‘Kendall’. London, UK. Mizutori, M., & Guha-Sapir, D. (2020). Human cost of disasters: An overview of the last 20 years (2000–2019). Centre for Research on the epidemiology of disasters (CRED) and United Nations Office for disaster risk reduction (UNDRR), Belgium and Switzerland. https://www. undrr.org/sites/default/files/inline-files/HumanCostofDisasters2000- 2019FINAL.pdf. Retrieved on October 29, 2023 Naheed, S. (2021). Understanding disaster risk reduction and resilience: A conceptual framework. In S. Eslamian & F. Eslamian (Eds.), Handbook of disaster risk reduction for resilience. Springer. https://doi.org/10. 1007/978-3-030-61278-8_1 O'Keefe, P., Westgate, K., & Wisner, B. (1976). Taking the naturalness out of natural disasters. Nature,260, 566–567. https://doi.org/10.1038/ 260566a0 Pal, I., Shaw, R., Shrestha, S., Djalante, R., & Cavuilati, R. A. W. (2021). Toward sustainable development: Risk-informed and disaster-resilient development in Asia. In Disaster resilience and sustainability (pp. 1–20). Elsevier. https://doi.org/10.1016/B978-0-323-85195-4.00001-9 Paluszynska, A. (2017). Structure mining and knowledge extraction from random forest with applications to the cancer genome atlas project. University of Warsaw Master Thesis. Paluszynska, A., Biecek, P., & Jiang, Y. (2020). Explaining and visualizing random forests in terms of variable importance, R Package Version 0.10.1. https://github.com/ModelOriented/randomForestExplainer. Last accessed on December 8, 2023 Park, A. S. (2024). Understanding resilience in sustainable development: Rallying call or siren song? Sustainable Development,32(1), 260–274. https://doi.org/10.1002/sd.2645 Rahman, A. U., & Fang, C. (2019). Appraisal of gaps and challenges in Sendai framework for disaster risk reduction priority 1 through the lens of science, technology and innovation. Progress in Disaster Science,1, 100006. https://doi.org/10.1016/j.pdisas.2019.100006 Raju, E., Boyd, E., & Otto, F. (2022). Stop blaming the climate for disasters. Communications Earth & Environment,3(1), 1. https://doi.org/10.1038/ s43247-021-00332-2 Salvucci, V., & Santos, R. (2020). Vulnerability to natural shocks: Assessing the short-term impact on consumption and poverty of the 2015 flood in Mozambique. Ecological Economics,176, 106713. https://doi.org/ 10.1016/j.ecolecon.2020.106713 Samaraweera, H. U. S. (2024). Exploring complexities of disaster risk and vulnerability: Everyday lives of two flood-affected communities in Sri Lanka. Sustainable Development,32(2), 1376–1385. https://doi.org/ 10.1002/sd.2723 Scheffran, J., Link, P. M., & Schilling, J. (2019). Climate and conflict in Africa. Oxford Research Encyclopedia of Climate Science, 38. https://doi. org/10.1093/acrefore/9780190228620.013.557 Shi, P., Yang, X., Xu, W., & Wang, J. A. (2016). Mapping global mortality and affected population risks for multiple natural hazards. International Journal of Disaster Risk Science,7,54–62. https://doi.org/10.1007/ s13753-016-0079-4 Song, Y. (2021). Introduction to spatial stratified heterogeneity models in R. Subroto, S., & Datta, R. (2024). Perspectives of racialized immigrant communities on adaptability to climate disasters following the UN roadmap for sustainable development goals (SDGs) 2030. Sustainable Development,32(2), 1386–1400. https://doi.org/10.1002/sd.2676 Szabo, S., Nicholls, R. J., Neumann, B., Renaud, F. G., Matthews, Z., Sebesvari, Z., AghaKouchak, A., Bales, R., Ruktanonchai, C. W., Kloos, J. & Foufoula-Georgiou, E. (2016). Making SDGs work for climate change hotspots. Environment: Science and Policy for Sustainable Development,58(6), 24–33. https://doi.org/10.1080/00139157.2016. 1209016 Thomalla, F., Boyland, M., Johnson, K., Ensor, J., Tuhkanen, H., Gerger Swartling, Å., Han, G., Forrester, J., & Wahl, D. (2018). Transforming development and disaster risk. Sustainability,10(5), 1458. https://doi. org/10.3390/su10051458 Trogrli c, R. S., van den Homberg, M., Budimir, M., McQuistan, C., Sneddon, A., & Golding, B. (2022). Early warning systems and their role in disaster risk reduction. In Towards the “perfect”weather warning: Bridging disciplinary gaps through partnership and communication (pp. 11–46). Springer International Publishing. https://doi.org/10. 1007/978-3-030-98989-7_2 UNISDR (United Nations International Strategy for Disaster Reduction). (2015). Sendai framework for disaster risk reduction 2015–2030. http://www.wcdrr.org/preparatory/post2015. Accessed December 8, 2023 United Nations General Assembly [UNGA]. (2016). Report of the openended intergovernmental expert working group on indicators and terminology relating to disaster risk reduction. United Nations General Assembly. Visser, H., De Bruin, S., Martens, A., Knoop, J., & Ligtvoet, W. (2020). What users of global risk indicators should know. Global Environmental Change, 62, 102068. https://doi.org/10.1016/j.gloenvcha.2020.102068 Wang, J. F., Li, X. H., Christakos, G., Liao, Y. L., Zhang, T., Gu, X., & Zheng, X. Y. (2010). Geographical detectors-based health risk assessment and its application in the neural tube defects study of the Heshun region, China. International Journal of Geographical Information Science,24(1), 107–127. https://doi.org/10.1080/13658810802 443457 Wang, J. F., Zhang, T. L., & Fu, B. J. (2016). A measure of spatial stratified heterogeneity. Ecological Indicators,67, 250–256. https://doi.org/10. 1016/j.ecolind.2016.02.052 Wen, J., Wan, C., Ye, Q., Yan, J., & Li, W. (2023). Disaster risk reduction, climate change adaptation and their linkages with sustainable development over the past 30 years: A review. International Journal of Disaster Risk Science,14(1), 1–13. https://doi.org/10.1007/s13753-023- 00472-3 Wisner, B., Blaikie, P. M., Blaikie, P., Cannon, T., & Davis, I. (2004). At risk: Natural hazards, people's vulnerability and disasters. Psychology Press. Yaghmaei, N. (2019). Disasters in Africa: 20 year review (2000–2019). CRED Crunch,56, 1-2. https://reliefweb.int/attachments/c4d1cae0- 8939-3994-9452-807893a8f8d0/CredCrunch56.pdf Zhang, M., & Wang, J. (2022). Trend analysis of global disaster education research based on scientific knowledge graphs. Sustainability,14(3), 1492. https://doi.org/10.3390/su14031492 4040 EZE and SIEGMUND Zhou, Y., Liu, Y., Wu, W., & Li, N. (2015). Integrated risk assessment of multi-hazards in China. Natural Hazards,78, 257–280. https://doi.org/ 10.1007/s11069-015-1713-y SUPPORTING INFORMATION Additional supporting information can be found online in the Supporting Information section at the end of this article. How to cite this article: Eze, E., & Siegmund, A. (2024). Identifying disaster risk factors and hotspots in Africa from spatiotemporal decadal analyses using INFORM data for risk reduction and sustainable development. Sustainable Development,32(4), 4020–4041. https://doi.org/10.1002/ sd.2886 EZE and SIEGMUND 4041