Road networks represent critical interfaces where human systems and geomorphic processes interact. In regions with steep landscapes, roads are commonly associated with an increased incidence of landslides. Yet, despite their socio-economic significance, systematic information on the frequency, type, and spatial distribution of road-related landslides remains largely absent. This data gap limits the ability of local authorities to allocate resources effectively, operationalize mitigation efforts, and conduct risk-sensitive infrastructure planning. In this study, we propose an innovative operational citizen-based method that aims to inventory with great detail how road construction, drainage modification, and associated terrain disturbance influence landscape morphology and its related hazards. Focusing on the highlands of Southwestern Uganda, a densely populated tropical region highly exposed to geo-hydrological hazards. We first conducted a detailed systematic baseline survey between November 2025 and January 2026 for road sections of 250-300 m along 254 km of roads of various types across different natural and human-influenced settings. From a total of 937 road section observations, preliminary results reveal various conditions dominated by road cut failure/soil/rock deposit from uphill (22%), active erosion (19%), fresh road cuts (19%), blocked roadside ditches, stone/soil extraction and quarrying of the road cut (7%). To have a detailed systematic temporal and spatial information of these roads conditions, we have established, together with local stakeholders, a network of 15 trained motorcycle-based citizen scientists who, for the next three years, will (i) generate a temporal inventory through systematic bi-monthly monitoring of the roads, and (ii) also report on landslide event occurrences along the road networks. Processes including road cut failures, surface sedimentation, drainage obstructions, pavement cracking, and proximal landslides, are being inventoried. We present here the first results of this operational participatory monitoring framework for understanding a human-influenced hazard in a data-scarce mountainous environment context.

Abstract. Urban gullies (UGs) are a rapidly expanding but under-recognized human-induced hazard in cities across the Global South, causing widespread damage to housing and infrastructure. Despite their growing importance, their economic impacts remain largely unquantified. Here, we assess the costs associated with UG expansion in Kinshasa (D.R. Congo), the world’s most affected city and a critical case to evaluate the potential magnitude of this hazard. Using an integrated approach combining field surveys and semi-structured interviews with 802 households, real estate data, and spatial analysis, we quantify direct damages, household-level costs, and system-wide economic impacts. Households experiencing damage (n=666) reported average material losses of 4,320 USD, while most respondents (n=606) contributed to bottom-up stabilization efforts, investing on average 335 USD. These costs are clearly substantial given that most households report monthly incomes below 500 USD. In addition, UGs significantly reduce property values, with total losses in Kinshasa estimated at 979 million USD. These findings reveal the massive and multi-dimensional economic burden of UGs and suggest that their impacts are likely severely underestimated in rapidly urbanizing regions. Urban gullies should therefore be recognized as a major and growing component of urban risk, requiring urgent integration into disaster risk reduction and urban planning strategies.

Abstract Citizen science (CS), the participation of laypeople in scientific research, is arguably a new concept in low‐ and middle‐income countries (LMICs). Consequently, formal CS guidelines and support infrastructure are largely non‐existent in these regions. This study seeks to contribute to this discussion by examining the factors that shape CS implementation and success in LMICs by analysing the strategies employed in the Action Towards Reducing snail‐borne Parasitic diseases (ATRAP) project. ATRAP aimed to prevent snail‐borne diseases in the Democratic Republic of Congo (DRC) and Uganda by establishing CS networks in selected communities. The project empowered community members to become “local snail experts” and communication “ambassadors” by collecting data on snail vectors. This study compared ATRAP's participant recruitment strategies and interests in DRC and Uganda, juxtaposing them with CS principles and practices observed in high‐income countries (HICs). The findings indicate that although project implementation followed similar strategies in both countries, the outcomes varied. Specifically, more participants were nominated in Uganda, while in DRC, female participants worked alongside their spouses. However, participant interests were consistent across both countries, with tangible benefits (financial compensation and smartphones) and social recognition observed as major motivating factors. These findings, among others, deviate from typical HIC‐based CS models that emphasize voluntary engagement, underscoring the need for context‐specific adaptations. Rather than simply replicating “standard” CS practices, future (LMICs‐focused) initiatives should prioritize proper community entry, diverse recruitment strategies, and strong support systems to address participation barriers within the communities they aim to serve.

Urban gullies (UGs) are an increasingly urgent concern in many cities of the Global South. Rapid and largely unplanned urban expansion, combined with inadequate drainage infrastructure, erodible soils, and intense rainfall, have led to the formation of thousands of large UGs —often several tens of meters wide and deep and extending over hundreds of meters— in cities across the Democratic Republic of the Congo. These gullies cause loss of life, destroy housing and critical infrastructure, and further exacerbate the vulnerability of already marginalized populations. The situation is particularly severe in Kinshasa, where more than 800 UGs have already been recorded, threatening over one million people.A wide range of initiatives has been implemented to stop UG expansion. These include large-scale engineering interventions led by the state or non-governmental organizations, such as concrete reinforcement of gully heads and canalizing the gully channel. However, many measures are expensive and/or often fail.Nevertheless, emerging evidence highlights promising strategies for urban gully prevention and control. A key principle is to prevent rainwater from leaving individual parcels by installing water retention structures, as the accumulation of runoff along roads is a primary driver of gully initiation and expansion. A critical requirement for success is that a majority of households actively participate in such initiatives. Improving risk awareness and creating synergies between UG control and water accessibility will therefore be crucial to achieving this.Here we aim to demonstrate the effectiveness of such a strategy. For this purpose, we installed water retention structures in a representative catchment in Kinshasa affected by UGs. This is done in close collaboration with local stakeholders. By monitoring and studying participation rates as well as the resulting hydrological effects (e.g., through the involvement of local students), we will develop actionable guidelines to this growing problem in Kinshasa and elsewhere, thereby enhancing both urban resilience and water security in vulnerable neighborhoods.

Studying natural hazards in the context of human-induced landscape transformation is complex, especially in regions with limited information. The narratives of the elderly can play a role in filling these knowledge gaps at the multi-decadal timescale. Here, we build upon a citizen-based elderly approach to understanding natural hazard patterns and landscape transformation in a tropical mountainous environment, the Kigezi Highlands (SW Uganda). We engaged 98 elderly citizens (>70 years old) living in eight small watersheds with different characteristics. Through interviews and focus group discussions, we reconstructed historical timelines and used participatory mapping to facilitate the interview process. We cross-checked the information of the elderly citizens with historical aerial photographs, archives, and field visits. Our results show that major land use/cover changes are associated with a high population increase over the last 80 years. We also evidence an increase in reported natural hazard events such as landslides and flash floods from the 1940s until the 1980s. Then, we notice a stabilization in the number of hazard events per decade, although the two most impacted decades (1980s and 2000s) stand out. Despite this new information, an increase in natural hazard frequency due to land use/cover change cannot yet be quantitatively validated, especially when the probable modulator effect of climate variability is considered. Nevertheless, the increase in the exposure of a vulnerable population to natural hazards is clear, and population growth together with poor landscape management practices are the key culprits that explain this evolution. This study demonstrates the added value of historical narratives in terms of understanding natural hazards in the context of environmental changes. This insight is essential for governments and non-governmental organizations for the development of policies and measures for disaster risk reduction that are grounded in the path dependence of local realities.

Large urban gullies cause damage in many tropical cities across the Global South1,2. They can result from inappropriate urban planning and insufficient infrastructure to safely store and evacuate rainfall in environments that are already highly sensitive to soil erosion1,3,4. Although they can cause large destruction and societal impacts such as population displacement1,2,5, the magnitude of this geo-hydrological hazard remains poorly documented and understood6,7. Here we provide an assessment of the extent and impact of urban gullies at the scale of the Democratic Republic of the Congo (DRC). Through mapping, we identify 2,922 urban gullies across 26 cities. By combining their formation and growth rates with population density data8, we estimate that around 118,600 people (uncertainty range: ± 44,400 people) have been displaced by urban gullies over the period 2004-2023. We find that average displacement rates increased from about 4,650 persons yr-1 (pre-2020) to about 12,200 persons yr-1 (post-2020). Between 2010 and 2023, the number of people living in the potential expansion zone of urban gullies doubled from 1.6 (±0.6) to 3.2 (±1.3) million, with more likely to be exposed due to urban sprawl9,10 and climate change11. We suggest that there is a need for tools and strategies to prevent and mitigate this hazard.

Improving understanding and awareness of risks associated with natural hazards among the population at risk and DRR managers is essential for achieving the objectives of the Sendai Framework. This is particularly crucial in contexts where natural hazard risk knowledge is scarce and poorly disseminated, while the frequency of disasters and the severity of their impacts are high. Highly interactive, educational games are an engaging method for exposing players to disaster risk situation by allowing them to observe and acquire knowledge, train their problem-solving and decision-making skills, and test different disaster risk reduction (DRR) strategies, while experiencing the consequences of disasters in a safe and entertaining environment. Such an approach based on educational games is experimented in eastern DRC, with the Hazagora and Chukuwa games. Hazagora is a board game originally designed for secondary school children. It is used not only as a knowledge-building tool, but also to raise awareness regarding the potential impacts of disasters and how to reduce them, through active engagement of participants in discussion on DRR strategies. As such, this approach sits at the science-policy-practice interface, involving not only children, but also teachers, scientists, civil society organisations and civil protection representatives. Building on this experience, the Chukuwa card game was developed as a disaster risk awareness tool for primary school children, whose ability to take their new understanding back to their families is recognized as a vector for disseminating knowledge. After several years of experimentation, some practical limitations linked to the contextualisation and institutionalisation of these games have however been identified. Based on the lessons learned, adaptations of the Hazagora game are being considered, as is the translation of the Chukuwa card game into local languages, alongside the strengthening of the involvement of secondary and primary education authorities and the integration of these tools into school (extra-)curricula. Educational games are therefore an effective learning tool for introducing participants to the concepts of natural hazards, risks and disasters, as well as for actively and sustainably engaging them in discussions and reflections on DRR strategies conducive to strengthening the risk culture within the community.

North-Western Rwanda's hilly and mountainous topography, high elevation, frequent torrential rainfall, and high population density render it highly susceptible to landslides and floods. A comprehensive understanding of community vulnerability to these hazards is crucial for effective risk assessment and mitigation strategies. To address data scarcity in the region, this study is based on a household survey approach that incorporates hazard-specific variables to compare vulnerability across three hazard categories: landslides, floods, and a combination of both. The survey encompasses 904 households across 50 cells (local administrative units), purposively selected according to hazard susceptibility distribution. Principal Component Analysis (PCA) was applied to derive a contextualized Social Vulnerability Index (SoVI). Five principal components accounting for 73.2% of the variance were identified. The first component, contributing 23.4%, highlights the vulnerability associated with unplanned settlements and low income. The second component, representing 19.5% of the variance, emphasizes demographic and social factors. The third component (12.6% of the variance) points to the vulnerability of households solely reliant on agriculture for their income. The fourth component (9% variance) is associated with land ownership, with households lacking land assets experiencing lower vulnerability. The fifth component (8.7% variance) underlines the relevance of household structure variables, indicating the high vulnerability of single-person households. SoVI scores classified 19 cells in the very high or high vulnerability category, predominantly those prone to landslides. These highly vulnerable cells are concentrated in the Northern Province, emphasizing the need to prioritize interventions in this region, such as effective land use planning and livelihood improvement strategies. This study provides a comprehensive vulnerability assessment and valuable insights for prioritizing interventions. The inclusion of hazard-specific variables and a comparative vulnerability approach across areas susceptible to landslides, floods, and both hazard types enhances the specificity and applicability of the findings. These insights are invaluable for local policymakers and disaster prevention and management authorities, enabling them to develop context-specific strategies to improve community resilience and reduce vulnerability to natural hazards.Keywords: Community Vulnerability, Landslides, Floods, Noth-Western Rwanda, Social Vulnerability Index

Many regions of Africa are exposed to a large variety of geo-hydrological hazards such as earthquakes, volcanic eruptions, landslides, floods, karst collapses and large urban gullies. Despite the soaring impacts on population, infrastructure and the environment associated with the occurrence of these hazard risks, most regions are under-studied. In addition to this lack of information, stakeholders, policy makers and the public at large remain relatively poorly aware of the hazard and risk problems, whether it is about their causes, their impact, and/or their mitigation. This overall lack of knowledge and awareness is associated with an aggravation of the impacts as the growing and vulnerable population of these regions, in search for new settlements and opportunities, is often moving towards areas that are more prone to natural hazards. This is in this context that UNESCO supports the preparation and dissemination of a guide booklet on geo-hydrological hazards for stakeholders, policy makers and the general public. The booklet targets ten African countries (Angola, Burundi, Cameroon, Central African Republic, Chad, Democratic Republic of the Congo, Equatorial Guinea, Gabon, Republic of the Congo, São Tomé and Príncipe) that are covered by the UNESCO regional office of Yaoundé. The aim of this work is to raise collective awareness of the need to prevent natural hazard risks at local, regional and national levels in order to ensure the protection of populations and promote the sustainable development of territories. In this way, UNESCO aims to guide and advise the ten African countries by providing them with useful and practical information.

Citizen science (CS) has gained recognition as a useful tool for monitoring and facilitating sustainable development transitions. However, CS initiatives are only emerging in the Global South, leaving many unknowns, like the factors influencing participation. This article contributes to the literature by examining the factors driving and limiting participation in two CS networks established in Uganda: Action Towards Reducing Aquatic Parasitic diseases (ATRAP), which monitored freshwater snails; and the Geo-observer (GO), which monitored natural hazards. Building on the theory of planned behaviour and the volunteer functions inventory, a questionnaire was administered to the participating individuals or citizen scientists and a control group that consisted of candidate citizen scientists, through group and individual interview settings. Motivations for participation were strikingly similar across the CS networks, respondent groups, and interview settings. The main drivers for participation were the desire to gain new skills and knowledge (understanding) and contribute to the community’s well-being (values), while the influence of others (social) and opinions or expectations of significant others (subjective norms) played lesser roles. Although the control group in both networks consistently expressed higher levels of positive responses, the importance of the motivational factors generally declined in both respondent groups over time. Financial compensation and favorable working conditions, like flexibility, contributed to sustained participation among the citizen scientists, while the major barriers to participation embodied external and internal factors, like bad weather and sickness. This study provides valuable insights to guide future CS recruitment initiatives toward alignment with the aspirations of individuals in similar contexts.

In many administrative entities, there is little information on the accessibility of critical resources such as food, water, healthcare and fuel. Yet the lack of accessibility to critical resources can be a risk-mitigating factor. Indeed, people who are not physically threatened by a hazard can be put at risk by the fact that they lack critical resources. In the specific context of an area threatened by lava flows and characterized by a scarcity of accessible and reliable data, we assessed the accessibility of critical resources in Goma, in the Democratic Republic of Congo (DRC), a densely populated city located at the foot of the Nyiragongo volcano. Our study shows that by mapping the location of critical resources and characterizing their accessibility, we can systematically identify underserviced urban areas that are consequently more vulnerable to risks, and in our study to volcanic risks.

This qualitative study takes a translocal perspective by considering migrants’ views on environmental change, migration decisions and translocal practices in favour of environmental adaptation. This article addresses the following question: To what extent do migrants’ perceptions of environmental change in their region of origin influence their translocal practices in favour of adaptation to socio-environmental change? Our data show that while environmental change may not be the primary reason that people migrate, nor do they perceive it as such, their translocal practices may have very concrete impacts in environmentally fragile areas, especially with respect to non-migrants in the place of origin. Most practices (e.g. family economic remittances) are spontaneous and unintentionally adaptive to environmental change. In contrast, collective projects initiated through hometown associations, especially in Senegal and Morocco, often have a more intentional and proactive character, resulting in translocal adaptive socio-environmental dynamics.

Assessing volcanic hazards in locations exposed to multiple central volcanoes requires to consider multiple potential eruption sources and their respective characteristics. While this is common practice in ashfall hazard assessment, this is generally not considered for topography-controlled volcanic flow processes. Yet, in volcanic areas with closely spaced volcanic systems, eruptions fed from several contrasted volcanic systems might threaten one given area. Considering the case of the Nyiragongo and Nyamulagira volcanoes in the Virunga Volcanic Province (D.R.Congo), we present a method to produce a combined lava flow inundation susceptibility map that integrates both volcanoes. The spatial distribution of the probability of vent opening for the next eruption is separately constrained for both volcanoes based on the mapping of historical and pre-historical eruptive vents and fissures. The Q-LavHa lava flow probability model is then calibrated separately for each volcano, considering several historical lava flows of Nyamulagira (2004, 2006, 2010) and Nyiragongo (2002). The maps for the two volcanoes are thereafter integrated based on a weighted sum of both individual lava flow inundation probability maps, assuming historically-based relative eruption frequency of the two volcanoes. The accuracy of this probabilistic susceptibility map for the most active volcanic region in Africa was unfortunately validated by the May 2021 lava flow produced by Nyiragongo. This map was discussed and validated in 2019 with local scientists, as well as representatives of disaster management and urban planning institutions, but was not included in the regional contingency plan ahead of the 2021 eruption crisis. Updating the volcanic crisis and evacuation management plans with this lava flow probability map could contribute to reinforce risk awareness among the population and inform the future development of the city of Goma.

Citizen science (CS) is gaining global recognition for its potential to democratize and boost scientific research. As such, understanding why people contribute their time, energy, and skills to CS and why they (dis)continue their involvement is crucial. While several CS studies draw from existing theoretical frameworks in the psychology and volunteering fields to understand motivations, adapting these frameworks to CS research is still lagging and applications in the Global South remain limited. Here we investigated the reliability of two commonly applied psychometric tests, the Volunteer Functions Inventory (VFI) and the Theory of Planned Behaviour (TPB), to understand participant motivations and behaviour, in two CS networks in southwest Uganda, one addressing snail-borne diseases and another focused on natural hazards. Data was collected using a semi-structured questionnaire administered to the CS participants and a control group that consisted of candidate citizen scientists, under group and individual interview settings. Cronbach’s alpha, as an a priori measure of reliability, indicated moderate to low reliability for the VFI and TPB factors per CS network per interview setting. With evidence of highly skewed distributions, non-unidimensional data, correlated errors and lack of tau-equivalence, alpha’s underlying assumptions were often violated. More robust measures, McDonald’s omega and Greatest lower bound, generally showed higher reliability but confirmed overall patterns with VFI factors systematically scoring higher, and some TPB factors—perceived behavioural control, intention, self-identity, and moral obligation—scoring lower. Metadata analysis revealed that most problematic items often had weak item–total correlations. We propose that alpha should not be reported blindly without paying heed to the nature of the test, the assumptions, and the items comprising it. Additionally, we recommend caution when adopting existing theoretical frameworks to CS research and propose the development and validation of context-specific psychometric tests tailored to the unique CS landscape, especially for the Global South.

Impacts induced by natural hazards are increasing globally. Some of these hazards, such as volcanic eruptions, cannot be prevented. Thereby, mitigating impacts is crucial, especially in densely populated areas, like in Goma city in the eastern Democratic Republic of Congo, which is exposed to volcanic threats from Nyiragongo. Mitigation requires the population to be prepared to face volcanic eruptions, by having a high state of awareness, anticipation, and readiness to take appropriate decisions during eruption crisis. Therefore, this study used a mixed-method approach (questionnaire survey, focus group, and written composition) to assess the impact of two educational tools: the Hazagora serious game and a museum of Virunga volcanoes. 402 students from 12 different schools participated in the evaluation, two third as participants in the educational activities and others as control group. Results indicated thatboth educational tools have, in a complementary way, a positive impact on student's volcanic disaster understanding, and their implication in DRR initiatives. The museum improves more the knowledge of the volcanic processes while the Hazagora game increases more the perceived self-capacity and mitigation skills. A key finding is that both educational tools significantly increase the willingness of implementing protective measures, but not the risk perception.

During a volcanic eruption, evacuation of exposed population is one of the most effective mitigation actions. However, if the eruption is not preceded by precursors nor guidelines by authorities, such as the 2021 Nyiragongo eruption, inhabitants living near the volcano need to rapidly decide by themselves whether to evacuate or not; and if so, where, when and how. On the evening of 22 May 2021, a red glow on the south flank of Nyiragongo alerted the population of Goma (East DRC) that an eruption had started. There were no instructions from the authorities, but on that night, part of the population self-evacuated and returned the following day as the lava had stopped flowing. In the following week, intense seismic shocks affecting the city led the North Kivu military governor to order, on May 26, an evacuation of the eastern neighbourhoods of the city. Using qualitative interviews associated with a quantitative survey of 1137 adults living in both mandatory and non-mandatory evacuation zones, this study analyses individual evacuation decision during both volcanic and seismic phases of the 2021 Nyiragongo eruption. The results indicate that the population in the north-eastern neighbourhoods self-evacuated on the night of the eruption without prior knowledge of their destination; women being more likely to evacuate than men. The decision to evacuate was not influenced by an individual risk assessment, but by the family safety. Following the May 26 evacuation order, 56% of participant households evacuated but were wishing to rapidly return to protect their home. Living in the east neighbourhoods was a major factor of evacuating during the two evacuations. Those who did not evacuate considered the risk to be lower and manageable, in comparison to the 2002 disaster. Heuristic bias plays a strong role in inhibiting the evacuation intention. This unique analysis of evacuation behaviour during a complex volcanic crisis provides insights about elements influencing the evacuation decision, which might be of relevance for other volcanic crisis worldwide.

Risk perception is an essential element to consider for effective risk management at the time of eruption, especially in densely populated cities close to volcanoes like Goma in the east of the Democratic Republic of the Congo, which is highly exposed to volcanic hazards from Nyiragongo. The perception of volcanic risk involves the processes of collecting, selecting and interpreting signals about uncertain impacts of volcanic hazards. Using a questionnaire survey, this study describes the spatial differences and factors influencing the individual volcanic risk perception of 2224 adults from eight representative neighbourhoods of Goma before the May 2021 Nyiragongo eruption. A composite risk perception indicator was built from the perceived severity and perceived vulnerability. Statistical analysis of the survey's results shows that the risk perception was high (mean = 3.7 on a five-point Likert scale) and varies less with demographic and contextual factors than with cognitive and psychological factors. Volcanic hazards were perceived to be more threatening the city and its functioning than the individuals themselves. The spatial analysis shows that respondents from the eastern neighbourhoods, affected by the 2002 eruption, demonstrated a significantly higher level of risk perception than participants living in the western neighbourhoods. This study will help to improve volcanic risk awareness raising in Goma.

<p>The expansion of informal and uncontrolled urban landscapes commonly overlooks the natural constraints from the environment. This is particularly true for urban environments affected by landslides. Landslide risk assessment relevant for urban planning and disaster risk reduction (DRR) strategy requires highly spatially-resolved datasets and approaches. It also requires that both physical and social local aspects of risk are studied in an interdisciplinary manner. Such assessment of hazard risk remains challenging and under-researched in many regions, especially in low- and lower-middle-income countries in the tropics, as it usually requires large and diverse datasets that are frequently unavailable or unreliable. In addition, specifically in urban contexts, human-induced environmental change impacts slope stability. Under these conditions of data-scarcity and land transformation, reliable and detailed landslide risk assessment encompassing the physical and societal aspects in an operational approach strongly relies on expert knowledge.</p> <p>In this research, we assess the risks associated with landslides in Bukavu, a city located in the eastern DR Congo where urban sprawling is high and the problem of landsliding is particularly acute. Firstly, we compiled a comprehensive multi-temporal landslide inventory covering several decades using remote sensing, archives, field survey and interviews with key informants. From this inventory, we derived three hazard zonations with multiple scenarios that allow to consider the interactions between various landslide processes and the role of human activities. Secondly, we obtained detailed socio-economic data from a sample population survey in morphological areas determined by remote sensing. Within two months, 10 specifically-trained local interviewers counted and located nearly 44,000 inhabitants living in about 6,580 households, and collected socio-economic baseline data over 10,880 people from 1,614 households. These demographic data were used to determine the variations in population density (exposure) in the city. These data were also key for the vulnerability assessment. For this, we designed a contextualised vulnerability index capturing the various dimensions of vulnerability with a set of selected indicators aimed at facilitating understanding, replicability and updating of the data collection. By combining hazard, exposure and vulnerability, we produced three risk zonation maps at a very high spatial resolution with the potential to be used operationally: one for shallow landslides, another for deformation within landslides and one for reactivation of deep-seated landslides. The development of these maps, as well as the collection of field-based information were carried out in close interaction with the city authorities and various stakeholders (e.g. civil protection, local community leaders) involved in DRR. A specific effort of awareness raising was also made through the organisation of dedicated workshops and radio programmes, and the implementation of a disaster risk information centre in Bukavu.</p> <div id="ConnectiveDocSignExtentionInstalled" data-extension-version="1.0.4"></div>

In the Tropics, disasters associated with natural hazards (intense convective rainfalls, floods, landslides) occur regularly. However, the general scarcity of reliable and accurate data collected on these events does not allow for a complete picture of their frequency and magnitude, thus hindering effective Disaster Risk Reduction (DRR). Such situation is observed in the Kivu region, in the eastern part of the DR Congo. Recurrent insecurity, long distances to travel, poor communication networks and the lack of financial resources to reach the affected areas are the main challenges faced by the Congolese Civil Protection in building a database that would allow for a better knowledge of these phenomena, in view of an appropriate disaster response and, in the long term, efficient DRR.Based on this observation, a group of 20 citizen observers was set up to collect data on six different types of natural hazard disasters (floods, landslides, wind storms, hail storms, lightning, and earthquakes) using smartphone technology connected to an online platform. This new approach, based on citizen science, makes it possible to significantly improve the documenting and understanding of the spatial and temporal occurrence of these disasters that affect the provinces of North and South Kivu. Since the establishment of this network in December 2019, more than 700 events have been recorded.If the data collected by this network of citizen observers constitute above all an unprecedented amount of information on the disasters occurring in such a tropical environment, they also allow for the compilation of a WebGIS and quarterly reports illustrated with maps and graphs, disseminated by Civil Protection to key DRR stakeholders active in the region, for a more tailored response, its planification, and, to some extent, the anticipation of such events. Scientists from universities and research centers in Bukavu and Goma are associated to that data collection and analysis. Moreover, citizen observers position themselves within their communities as key actors in raising awareness about disaster risks. However, although this type of approach has proven to be effective in the short term, the motivation on the long term of citizen observers, as volunteers, has been identified as a weakness to be addressed.

<p>In urban areas, networks and life-support systems, such as roads, power lines and drinking water pipes, are key infrastructures which contribute to the functioning of the city. Their damaging or destruction can lead to harmful consequences for the population. Such situation has already occurred in several African cities following the impact of a hazard of natural or human-induced origin. Bujumbura, the economic capital of Burundi, is no exception to such scenario, particularly as a result of geo-hydrological hazards which regularly affect various neighborhoods of the city: floods, flash-floods, river bank collapse, large urban gullies. For instance, the February 2014 flash flood damaged several key infrastructures in the north of Bujumbura, including the National Road number 1, a medium-voltage line and drinking water pipes.</p> <p>The goal of this study is to show how, in an urban system like the city of Bujumbura, the vulnerability of these key infrastructures to geo-hydrological hazards can lead to the fragility of a large territory. This vulnerability is influenced by weaknesses related to their wear or age, their construction technique, and their dependence on external systems for their effective functioning. In addition to their own weaknesses, these key infrastructures are exposed to several natural hazards that can lead to their damage or destruction. In Bujumbura, the geo-hydrological hazards that regularly impact the city are accentuated by the lack or outdated rainwater drainage facilities and the inadequacy of the urban development plan.</p> <p>After listing, locating and characterizing all the infrastructures that play a role in the life of the population in the urban environment, a classification, based on a participatory approach, was developed to select those on which the vulnerability assessment focused. Six forms of vulnerability (intrinsic, dependency, hazard exposure, control capacity, alternative operation, and crisis management) were assessed for each key infrastructure. To assess each form of vulnerability, interviews were conducted with facility managers (28) to supplement the data from the literature and field collection. Geometric weighting was performed to highlight the most vulnerable infrastructure that could disrupt the functioning of others and, by extension, that of the city.</p> <p>The results of this study show that several road sections in the center and at the exit points of Bujumbura, as well as the drinking water treatment plant and the pumping stations, stand out among the most vulnerable infrastructures. Continued efforts are required to maintain, protect, and strengthen these networks and life-support systems in the face of geo-hydrological hazards, as their damage or destruction due to cascading effects caused by dependence on, or wear and tear of, infrastructure can affect a much wider area than just the hazard impact zone.</p>