Multiple Hazards and Governance Model in the Barranquilla Metropolitan Area, Colombia

The Barranquilla Metropolitan Area is exposed and often vulnerable to various natural and anthropogenic hazards. The paper’s main objective is to identify the level of understanding that local and regional institutions have of such a multi-hazard scenario, as well as the effectiveness of governance arrangements in minimizing impacts. Research employed a questionnaire applied to 115 stakeholders from government and a survey of 391 households from four communities in the study area. Four focus groups were held during the update of the Barranquilla Development Plan 2020–2023. The results allowed the identification of an updated set of hazards and the levels of government response capacity. The overall level of capacity and effectiveness of local government to respond to hazards was classified as regular. Seventy-seven percent of epistemic ‘experts considered that the institutions responsible for risk management did not undertake sufficient analysis for identifying and managing multiple hazards. Finally, the research team developed a new model of risk management.


Introduction
Metropolitan cities are increasingly vulnerable to natural and anthropogenic hazards [1][2][3][4]. In many of these cities, large river floods have become a major source of risk [5,6]. Vulnerabilities intensify due to population growth and poor urban development in highly exposed areas [7,8]. At the same time, factors such as the deficient or null provision of relevant urban services by government prior to and after an event further increases the overall quotidian and disaster risks. All the above-mentioned factors affect the efficiency and relevance of risk management [9].
Different types of floods are identified in the literature on the topic. These include coastal floods due to storms, sea-level rise associated with climate change, wave action, and inland storm surge [10,11]. River floods may occur due to heavy rainfall and river overflow in the section of the floodplain closest to the sea, among other causes [12,13]. Two different mechanisms can generate urban flooding: (a) a town or city that occupies a part of the floodplain is flooded by events of external origin, (b) floods generated by internal urban fluvial and wastewater runoff [14].
In the Barranquilla Metropolitan Area (BMA), the potential flooding associated with the patterns of territorial expansion and rapid urban growth was never considered by planning authorities. Floods occur in Barranquilla due to (a) the intensity and duration of tropical rainfall events and high pluvial flow rates [15]; (b) the magnitude of the hazard, urban occupation of stormwater runoff channels and beaches, as well as the waterproofing of absorbent surfaces in the river basins.
The lack of drainage systems in most areas of the city, as well as the occurrence of highly hazardous flash floods, constitute a challenge for urban risk management [16,17]. The Barranquilla District Government has invested heavily in the construction of a robust stormwater runoff channel system. However, scant investment in raising the levels of civic culture and in stimulating the appropriation and empowerment of citizens with reference to the works carried out has impeded the participation of the local population in the protection and conservation of these works. Continuous dumping of solid waste has been observed in channeled and non-channeled streams. Regular flooding has caused accidents, loss of cars, damage to infrastructure, and on average, around two human deaths a year [18,19]. Indirect effects on livelihoods, income, employment, amongst other factors, constitute the systemic risk impacts of primary events. Floods mainly occur in central areas of Barranquilla city and areas bordering the Magdalena River [20].
Although urban flooding is an important hazard in the BMA, it is not the only hazard present. However, low risk perception by impacted communities impedes sufficient consideration and analysis of these.
In this article, risk perception is defined in terms of the level of awareness of local government, community, public and private enterprise organizations, and institutions when faced with multiple hazards and which affects the taking of risk reduction decisions. This is conditioned by the understanding of the existing levels of vulnerability, probability of damage and possible losses.
In Barranquilla, the capital city of the BMA, streams run from and over the city's western slopes. These slopes are occupied principally by low-income families and communities with high levels of socioeconomic vulnerability. The demographic pressure on these territories has increased due to the migration of people displaced by different forms of internal conflict in Colombia and of Venezuelans due to the country's economic and political situation [21].
The first flooding incident occurring in Barranquilla was registered in 1916 [22]. However, it is in the period 2010-2011 that the area was impacted by the historically most severe natural event ever suffered ("La Niña") [23]. This Niña event led to the entire city, particularly neighborhoods on the western slopes, suffering severely from landslides and stream flooding [10,24]. As of that date, a serious search to identify and understand the causes and effects that these hazards represent for the district began, and a firm intention has been expressed to develop actions that increase urban resilience.
Beyond flood impacts, the BMA faces other challenges that limit its sustainable development. Hurricanes in the Caribbean basin and tornadoes' affect the region periodically [25]. Additionally, medium term projections suggest that the BMA will be severely affected by climate change with consequences in terms of coastal erosion, increase in the intensity and frequency of winds and flooding, among others [13]. The lack of prioritization of these hazards in local risk management policies contributes to the existing low level of urban resilience. In principle, this is due to the insufficient appropriation of risk knowledge, a lack of cross-sectoral coordination, different levels of risk tolerance, and the lack of implementation of measures to mitigate the event's consequences.
Resilience is a broad concept and can be defined in different ways depending on how and in what scientific area the term is applied [26]. These include ecological [27], engineering [28], psychological [29,30], community [31], and urban resilience [32,33]. This latter concept is increasingly employed in the international scientific literature and, as risks to the urban population increase, urban resilience gains relevance in the major international development agendas.
Urban resilience refers to the capacity for overcoming the impact of natural and anthropogenic hazards and the associated vulnerabilities of affected populations [23]. Facing risks through an effective institutional and governmental response, with adequate citizen participation can significantly help recover adequate levels of social security, employment, income, and quality of life for citizens, after they have experienced the negative impact of one or more hazards [34,35].
Although the local government of Barranquilla implements actions and attempts to promote a resilient urban model, the BMA is still not part of the City Resilience Profiling Program (CRPP) promoted by UN-Habitat [36]. In fact, no city in Colombia is part of this program. The local government of Barranquilla attempts to respond to the different hazards by promoting governance processes that increase the presence of urban risk management principles and practice. However, they have not yet managed to respond to the social demands and prospective management needs of the BMA [16].
This paper proposes a model for improving public policies based on governance for the integral management of disaster risk. This model must include increases in the levels of institutional knowledge and understanding of urban hazards, the participation of citizens or civil society in evaluating risks, the design, development, and operation of risk reduction and disaster response measures, and in the government's effectiveness to respond to the impacts of hazards and institute risk reducing recovery processes.
The article comprises four sections. The first addresses the updated set of hazards identified by government institutions, as well as the government's capacity to respond to emergencies. The second section evaluates the perception of hazards and risks. This process aims to comprehensively understand how different stakeholders at different levels perceive risk in the BMA. The level of effectiveness of government entities in managing the multiple hazard scenarios is evaluated in the third section. At the same time, the weaknesses and strengths detected in public policy aimed at the governance of the identified hazards, with special attention to urban floods, are synthesized. The fourth and last section presents a novel model for integrated risk management and for increasing the urban resilience of the BMA.

Materials and Methods
To gain greater clarity as to management and action when faced with flooding, it is essential to measure the differential perceptions of stakeholders [37] in relation to risk management. This includes stakeholders from community, government, civil society, and scientific institutions. For the present research endeavor, we identified stakeholder experts considering their epistemic and on the job expertise [38]. Epistemic experts were able to provide justifications for a range of propositions according to their expert domain (professionals in risk management). While on the job, experts were selected considering their empirical practice and skill with issues of hazards and risks. Figure 1 shows the steps followed for the analysis, collection, and processing of information. In the following section, details of each stage are provided.

The Methodological Framework Followed in the Research
First stage: we undertook an in-depth documentary review, consulting and analyzing different bibliographic sources that report on river flooding in the BMA, with a record between 2010 and 2020 [39][40][41] [49], a GIZ sponsored report on Socio-spatial Segregation, Disaster Risk, and it's Management [50] as well as graduate theses [51]. Second stage: this was developed in two phases. The first evaluates the perception of risk in four communities. A questionnaire was conducted with 391 households in four neighborhoods of the BMA (La Cangrejera, Adelita de Char, Nueva Colombia, and La Esmeralda). We considered older neighborhoods with dense populations located in high-risk areas and with a sense of belonging among the communities. These places are characterized by a marked inequality (the most evident social problem in the city), low urban service endowment, poor socioeconomic conditions, and high exposure to hazard. Furthermore, a significant proportion of the population lives on slopes and is affected by landslides, floods, and fluvial runoff. The questionnaire sought to understand the level of knowledge existing as to what to do when facing hazardous events in the neighborhoods; and also, respondents' levels of exposure to hazards and their perception of these.
In the second phase of stage 2, four focus groups or participatory roundtables were held with experts on 5 March 2020 [52,53]. The objective was to measure the perception of stakeholders (on the job, performance experts) as to the effectiveness of the government management of urban risk when faced with the different hazards present in the BMA. To this end, a weakness-strengths matrix was applied. We also collected information regarding actions and recommendations on risk management that experts considered should be included in the Barranquilla Development Plan in the period 2020-2023 [54].
The focus groups included 75 performance experts. These persons were employees, and executives working in disaster risk management (DRM), (see Appendix A). Those included the business and industrial sectors, military and government, relief agencies, carnival operators, the academic community (the heads of Higher Education Institutions such as Universidad de la Costa, Universidad del Norte, and the Simon Bolivar University). Furthermore, officials from the Fluvial Maritime Committee, several companies that supply inputs for risk management, and specialized consultants also participated.
At this stage, the research work was enriched with contributions from the Office of the Mayor of Barranquilla (www.barranquilla.gov.co), specifically from the Office of Risk Management (www.barranquilla.gov.co/gestion-del-riesgo). This is the institution in charge of identifying hazards, vulnerabilities, and managing risk in the BMA. We also consulted employees of leading institutions, such as the Public Environmental Establishment "Barranquilla Verde" (www.barranquillaverde.gov.co) and the "Autonomous Regional Corporation of Atlántico" (www.crautonoma.gov.co).
Third stage: we undertook surveys using a structured questionnaire applied to forty epistemic experts from regional and local government institutions working on risk management in the BMA (see Appendix A). The questionnaire was designed following the methodology developed by Hurtado in 2010 [55]. The questionnaire had five parameters or points of reference: (1) hazards of natural origin, which were sub-classified into geological and hydrological hazards, (2) hazards of technological origin, (3) hazards of anthropogenic origin, (4) hazards of social origin, and (5) hazards of epidemiological and health origin. The questionnaire was answered directly by the stakeholder without the intervention of a third party. The surveys were distributed by email. Participants were also informed of this survey through phone calls. The questionnaire allowed them to freely identify hazards in the BMA, following the Departmental Risk Management Plan for the Atlántico Department [39].
Concerning multi hazard contexts, the epistemic experts were asked to specify the degree of severity of each hazard, considering five variables, namely: (1) Probability: the possibility that the hazard may occur at a certain time and place. This is measured using the following evaluation scale: 5 (totally probable), 4 (moderately probable), 3 (occasionally probable), 2 (rarely probable), 1 (null or improbable). (2) Frequency: temporal recurrence of the hazard in the BMA. This variable was measured according to the following scale: 1 (never), 2 (sporadically), 3 (moderately), 4 (almost always), and 5 (always). At this stage also, the epistemic experts were asked to freely list the governments' weaknesses and strengths in searching to solve the problem of hazards.
All the questionnaires were processed on 3 July 2020, with a response rate of 95%.

Case Study of Barranquilla Metropolitan Area (BMA)
Barranquilla is the fourth largest metropolitan area in Colombia, with 2.7 million inhabitants [54]. The BMA is located in the northern coastal area, with free access to the Atlantic Ocean and the Caribbean Sea, on the banks of the great Magdalena River. The other municipalities that make up the BMA are situated in the intermediate rural periphery between river and sea (See Figure 2). Urban segregation and inequality are prevalent in the BMA. For several years, the study area has been subjected to a development model that facilitated the emergence of significant inequity and environmental problems, putting the most impoverished communities at high levels of latent risk.
During the first decades of the 20th century, the economic dynamic of the city led to accelerated demographic growth, associated with the migratory process from the countryside to the city. The city's expansion to the northwest and south generated two different socioeconomic patterns associated with the business elite and the working population. The northwest expansion occurred with housing and service provision for a population with greater economic resources and to the south with working class housing due to lower incomes [56]. Parallel to these planned developments, land invasions have occurred, especially in the southwest. These are what are termed "sub-normal" neighborhoods lacking many essential public services. As the city grew, issues such as informality, the government's inability to generate development policies for the entire population and corruption led to the lack of opportunities and means for achieving socially more just and equitable investment [56].
Currently, the city faces severe spatial problems related to the lack of planning and the tendency to privilege dominant economic groups. Regional economic elites have played a decisive role in the adaptation of the territory to their needs using the control of state powers, territorial planning, infrastructure construction, among other processes, to achieve this [57]. As a result, an important part of the city's inhabitants occupy territories unsuitable for urban habitation, in many cases, perpetuating conditions of poverty and vulnerability in social, economic, and environmental terms [58]. The area of study has significant relevance for analyzing multi-hazard, governance, and risk management scenarios.

Identification of Hazards, Government Capacity to Respond to Emergencies, and the Perception of Hazards and Risks by Communities in the BMA.
Identification of the multi-hazard scenario in the BMA was achieved through a literature review and surveys and revealed 36 different hazards (See Tables 1 and 2).
Volcanic mud and gas flows   [39] is the instrument that identifies and analyzes the largest number of events. The remaining plans and documents consulted [24, 40,47] only considered the presence of between 8 and 16 hazards in the study area. None of the documents analyzed identified hazards of a social, epidemiological, or health nature. This probably reflects the interpretation of hazard in disaster studies existing prior to SENDAI when a larger number of additional hazards, beyond the so-called natural and socio natural, were placed on the table, including health-related and epidemiological hazards. The most frequent hazards present in the documents are landslides, coastal erosion, and stream flooding. The three are hierarchically similar. Table 2 registers the degree of probability, occurrence, frequency, and exposure to each of the 36 identified hazards assigned by epistemic experts. Contrary to the analysis developed in the documentary review, in this second stage, 100% of the experts consulted uniformly identified stream flooding and gale force winds as the natural hazards with the highest incidence in the BMA.
Only 37.5% of the experts recognized coastal erosion as a hazard with a high degree of probability, frequency, exposure, and coverage of affected territory. The other 62.5% did not consider that this hazard has a high level of incidence in the BMA.
COVID-19 stood out as the most important hazard for the BMA. Hazards of social origin, such as extortion, poverty, and drug addiction were also recognized by experts. These hazards endanger the lower social strata in particular. At the same time, electrical storms were identified by 18 surveyed experts. Landslides were not relevant for 55% of the experts consulted.
In addition to the set of hazards listed in Table 1, 29 experts recommended that offshore hydrocarbon exploitation be considered a serious hazard.
According to the perceptions as to the efficiency of local government response, epistemic'experts identified stream flooding as the hazard where the government exercises greater response capacity. This is due to the high exposure to, and percentage of territory that historically has been affected by flooding (See Figure 3). The government and mayor's office have implemented different response mechanisms in order to face the emergencies caused by this hazard. Implementing hydraulic plans and the construction and cleaning of stream channels are part of the local government measures to minimize the risk (See Figures 4 and 5).   Regarding the perception of risk in the four studied communities of the BMA, the questionnaire conducted in 391 households addressed local people's knowledge of hazards and their response strategies. Only 11% of respondents were part of a social organization and 9% reported having been involved in risk prevention projects.
In terms of response capacity, we found that 50% had telephone contact with the police authority and 60% were aware of the emergency care lines. Concerning the knowledge as to what to do in an emergency, 55% of respondents said they did not know what to do. Seventy-one percent of the participants reported knowledge on the prevention of vector diseases.
One outstanding aspect relates to the respondents' exposure to hazards. Results suggest that the most hazardous aspects were perceived to be anthropogenic and related to citizen security, as opposed to natural and socio-natural events. This reflects a now well-known fact that poor population are more concerned with everyday risk than with events that occur sporadically, are temporally dispersed and are the subject of attention by local or national governments. Although respondents rated most 'environmental' hazards, such as fires, with low values, in the "Adelita de Char" and "La Cangrejera" neighborhoods, the highest rating was for stream flooding and landslides. This result relates to the high levels of exposure of the inhabitants to these hazards historically and corroborates findings of previous research that identified high levels of awareness about the risk of urban stream floods [39,46,47,58] and landslides [45,58].
The results highlight the fact that the BMA has high socio-cultural vulnerability derived from society's low organizational capacity, inequality in life conditions, and the general ignorance of the population in terms of hazards and risk management issues. Educational strategies are lacking, and there are no community systems of preventive alerts based on strategic communication. In the BMA, vulnerable communities are mostly composed of people living in extreme poverty or marginality due, amongst other things, to the inadequate provision of public services, such as those related to health. These disadvantaged areas lack opportunities for income generation, signifying constant social degradation and insecurity. A better understanding of which people and systems are vulnerable to specific hazards and perceptions about them are necessary to improve risk management governance. We suggest a deeper analysis of this aspect in future studies.

Level of Effectiveness of Government Entities in Managing Multi-Hazard Scenarios
This section presents the weaknesses and strengths detected in governance. The level of effectiveness of BMA government entities in the face of hazards was also analyzed. Results were derived from the opinions of surveyed epistemic experts and community members who participated in the four focus groups. The aspects analyzed had an institutional, territorial, and socioeconomic focus. These elements can be considered for improving public governance and for facing multi-hazard scenarios in the BMA (See Table 3). 3. Poor identification of the hazard hierarchy in planning and risk management plans for BMA.4. An increasing number of low-income households in BMA. 5. Poor performance by public officials in charge of disaster risk management in BMA. 6. Increase in migrant settlements towards the western slopes in the capital of BMA due to deficient territorial and land use control. 7. Increase of the areas exposed to the risk of flooding, such as the land surrounding the Magdalena River wetlands. 8. Little awareness and empowerment of citizens regarding biotic and abiotic resource management, (especially water). 9. Low level of government response to some hazards. 10. Little attention to families living in precarious conditions. 11. Increased land invasion in high-risk areas that are not "liberated" but reoccupied by resettled migrants. 12. High densification in specific city sectors and restricted use of green areas, open spaces, and recreational areas. 13. BMA transformed by anthropic interventions. 14. Little implementation of corrective and preventive measures to relocate populations from high-risk areas, improving their livelihoods, employment rate, access to services. 15. Insufficient promotion of urban agriculture. 16. Weak government management to minimize mitigable environmental risks and hazards such as atmospheric pollution. 17. Proliferation of commercial activities in residential areas. 18. Unjustified delays in the relocation processes affecting populations due to hazards like landslides and granular flows. 19. Few urban planning initiatives to change the use of liberated territories and convert them into recreational or productive public space. 20. Occupation of fragile local ecosystems and vulnerable areas in the BMA. 22. Serious problems due to poor solid waste management inside the newly built rainwater runoff channels, caused by the lack of an environmental culture in the population living in BMA (See Figure 6a-d). 23. Corruption and centralization are considered as the most significant problems to overcome for facing hazards on the territory. Some measures require the authorization of the national government to obtain economic resources. 24. Limited infrastructure and personnel capacity to attend emergencies, especially in the health system. 25. Poor adoption of measures for adapting to climate change in planning and risk management plans. 6. Government strategies include funds for the construction of artificial conduits, retention or detention works, channelling and sediment control, as well as new collectors and water storage works, which improves the macro-drainage network of the streams (See Figure 4). 7. Government approval of investments to improve the micro-drainage network in the capital city of BMA, which guarantees efficient underground and surface drainage (See Figure 5). 8. Development of government measures for maintenance and cleaning of canals, as well as other prevention and management actions related to raising public awareness and control of impacts derived from final disposal of solid urban waste (See Figure 7). 9. High level of human capital formation. 10. Small territory that allows execution of integrated strategies. 11. Academic sector willing to contribute to solving problems. 12. Private sector ready to enter public-private alliances, as well as to address identified and characterized risk scenarios, especially technological risk.  In this phase, research revealed 25 weaknesses and 12 strengths among the measures designed to face identified hazards. Epistemic experts in risk management concluded that although planning and risk management officers recognize hazards due to floods and landslides, there are still insufficient management options given the magnitude of the problem. Sixty three percent of experts stated that greater attention should be paid to the continuous migration to and settling of population on urban slopes as well as areas affected by floods, like the perimeter areas of the Magdalena River wetlands.
Forty-five percent of experts mentioned government strengths in reducing problems related to floods. However, other important hazards for the region are not addressed with the same hierarchy, such as landslides, granular flows, coastal erosion, and social hazards.
Although the impact of flooding originating in stormwater runoff channels has been reduced due to management actions by the government, some canalized stormwater runoff channels continue to be dangerous. This is due to inadequate waste management practices resulting in obstruction of channels (See Figures 6 and 7).
We undertook a detailed review of Law 1523, 2012, which created the so-called National System for Disaster Risk Management [59], in order to analyze and evaluate the level of effectiveness of government entities in the BMA in managing multi-hazard scenarios.  [39].
The roles of these institutions were analyzed by 75 performance' experts in the four focus groups, as well as 40 epistemic experts. Results obtained are recorded in Table 4.

(Very Good)
Mayor's Office 23 10 Table 4 shows the result of the assessment of the effectiveness of government management of urban risk in BMA. With 77.4% of the total weighted value, effectiveness was classified as regular. Most experts consulted consider that several institutions from CDGR in the Atlantic Department have not achieved an adequate level of hazard management.

Model of Governance to Face Disaster Risk Management in the BMA
This third and final section presents a governance model for managing multiple hazards and increasing urban resilience in the BMA. The proposal for this governance model is presented in Figure 8. The approach of local governments to risk management in the BMA has focused on floods and conservation of nature more than on adaptation measures to face multiple hazards [61]. Considering that disaster risk materializes when there is vulnerability to hazards of natural, socionatural, technologic, or anthropogenic origin, our proposed model aims to intervene in the causes that have perpetuated exposure to hazard in the territory and to the existence and growth of vulnerability.
The model is based on the integration of comprehensive government actions to guarantee efficient and timely attention when implementing corrective and prospective measures (Corrective refers to the reduction of existing risk factors. Propective refers to the anticipation of future risk avoiding the creation of new risk factors). Our proposal reinforces management actions favoring improvements in risk knowledge and understanding, which are still insufficiently dealt with in Colombia [9,23,58,62]. The proposal is also based on the legal framework for disaster risk management in Colombia [59]. In Article 2, on liability, the law promotes the exercise of Governance, granting responsibilities to public, private, and community entities for risk management within the framework of their competencies and as components of the National System of Disaster Risk Management.
For achieving proper risk management, the active participation of civil society [58,63], as well as the training of expert technicians in the prevention and control of risks is crucial [64] (p. 1). Civil society must contribute to the design and implementation of public policies, improving their quality, guaranteeing their integrity and security according to principles of sustainable development [65,66]. Because of this responsibility, the Governance Model for Disaster Risk Management proposed in this research integrates procedures, protocols, and indicators described in Law 1523 [59] with actions that local government and civil society must undertake.
The model is based on four steps: (1) Strengthening knowledge and understanding of multi-hazard scenarios. Actions to raise the awareness and knowledge base of the communities and entities with responsibilities for the different hazards; (2) Strengthening knowledge on integrated vulnerabilities. This includes actions to identify impacts and consequences of hazards, as well as degrees of exposure and vulnerability of the territory, analyzed using a participatory approach; (3) Strengthening of risk reduction and control measures. This includes actions for identifying and evaluating possible responses to emergencies. The use of sustainable development indicators to minimize risks is a relevant contribution of the model at this stage; and (4) Recovery of an environmental culture in the population. Actions aimed to implement measures for continuous monitoring of hazards, vulnerabilities, and risks with a SDG approach are also incorporated.
The recommended minimum scale for information collection and analysis is 1: 2000. A systematic inventory of data can be carried out in this model. The model allows for a consideration of indicators such as streams contaminated by waste, number of people lacking environmental culture, among others, all impeding the fulfillment of the 2030 Agenda. As many authors [67] suggest, when monitoring is developed in the short term (each year), it is possible to detect trends.
The four phases of this model are based on active citizen participation.

Relevant Aspects Regarding the Identification of Hazards and Public Perception of Hazards and Risks by Communities in the BMA
In Table 1, it can be appreciated that hazards of anthropogenic origin, environmental degradation, and air pollution, were the most representative in the BMA. Pollution has intensified in recent years. Recent research shows that atmospheric pollutants have increased with impacts not only on the population but also on the historic built heritage of the study region [68,69].
Regarding stream flooding, the result obtained in the review of risk management instruments in the BMA coincided with the results of research undertaken by INVEMAR [61]. BMA and the Department of Atlántico are the territories in the Caribbean area where the highest numbers of natural and anthropogenic events occur. The main hazard is flooding by the Magdalena River and its feeder streams and swamps.
From the 2020 Atlantic Vision Agenda [48], the hazards considered to have the highest incidence on risk in the BMA are landslides, granular flows, erosion, flooding, forest fire, and environmental pollution. Nevertheless, POMCAS [42-44] indicates landslides, erosion, and floods as the main hazards. They are followed in importance by forest fires, extortion, contamination of water bodies, atmospheric pollution, and spills of flammable substances. The land use planning and risk management tools consulted are not homogeneous in terms of the hierarchy levels and priority given to hazards.
With regards to coastal erosion, Table 2 shows several differences as compared to the results of research published by INVEMAR in 2018 [61]. In their research publication, 55% of the coast of the BMA was assigned a certain level of damage due to erosion, with high and moderate levels identified on low coastal lands such as beaches. Another scientific publication that asserts the importance of this hazard for the BMA was published by Rangel et al, in the same year [70]. The authors conclude that the coast of Barranquilla has presented high rates of coastal erosion in the last twenty years.
The presence of COVID-19 and the highest level of impact assigned to this for the BMA is understandable because, when applying the survey, the Atlantic department reported 491 cases of contagion with a sustained increase in new infections and deaths [71]. In fact, the cases continue increasing today [72].
Although electric storms were identified by some experts, and it is in fact an important natural hazard for Colombia [73], this event is not reported as being relevant in the risk management tools analyzed [39,41].
In the BMA, landslides were not relevant for the experts consulted. This analysis may have an implicit individualistic cultural bias. The interviewees stated that they did not fear the occurrence of this hazard. Many experts and community members in fact have personal experience with this type of event but still do not consider it that important. The result corresponds with literature about perception of risks amongst western slopes inhabitants of Barranquilla when faced with landslides [45]. According to that research, it is, however, necessary to include the risk of landslides in management strategies.
The consulted experts also reported gales. This result corresponds with the information obtained in stage 1, where seven risk management plans [24, [39][40][41]45,46] recognize this hazard in the study area.
The exploitation of hydrocarbons offshore was identified by experts but is not present in the risk management tools analyzed. This anthropogenic hazard begins to concern some researchers because it is a potential risk on the BMA coast [74,75].
According to the analysis regarding perception of risk of hazards by communities in the BMA, the research revealed the inhabitants' capacity to respond to multiple hazards and risks. Social perception governs how people understand their physical and social world [10,34,76,77]. Results published in [44,45] indicate that perception of hazards is related to social conditioning factors and highlights the fact that people's perceptions of hazards are both socially constructed (i.e., contextual) and subjective [76,78]. Therefore, social perception with regard to multi-dimensional hazards in the local context is important and is highlighted in our research.

Relevant Aspects Regarding the Model of Governance to Face Disaster Risk Management in the BMA
In the proposed model, governance for risk management is the key. It incorporates an urban resilience perspective based on knowledge and reduction of risk through disaster risk management and addresses the local context.
Our proposal differs from the methodology for assessing hazard, vulnerability, and disaster risk, including climate risk and cost-benefit assessment, for projects in pre-feasibility (phase II) and feasibility (phase III) phases [79] as carried out by the Colombian ministries of economy, finance, planning, and environment. One of the differences in our proposal is based on the scale. The aforementioned method uses 1:25,000 scales, while our proposal promotes hazards and risk assessments at a larger scale of 1:2000.
The model takes up on the weaknesses identified by the experts and converts them into opportunities for the generation of comprehensive policies, in which risk management is an instrument of planning, land use, and development. This also differs from the methodology designed to perform ecological risk assessments in coastal areas and focused on the ecosystem's health, prioritizing ecosystems' risks based on only four natural hazards [80].
The proposal considers the concept of "governance" because it is more encompassing than governability [81]. Environmental governance is in many ways like the concept of sustainable development because it provides a degree of cross-cutting integration of various interests, perspectives, and approaches, that may mean different things to different people [82,83].
Considering that governance refers to the capacity for implementing and enforcing decisions made by government authorities and stakeholders, governance is an exercise of power in a broader sense. It implies certain modalities of allocating resources, con-trol, and coordination, where government actors are not necessarily the most relevant participants [84,85]. For this reason, the model promotes these processes, linked to public participation.
Many weaknesses identified in the methodologies and manuals analyzed indicate the need to reduce inequalities [86,87]. It is important to create conditions for the population to gain access to participation, as well as education that allows greater control regarding their environment, the hazards in their territory, and mechanisms for participation in the construction of local development. In the proposed model, knowledge of risk must be understood from an appropriation of territory perspective and not only through knowledge of hazards.
Another relevant aspect of this model is the relationship with the stakeholders. Local governments, relief groups, and the armed forces play a relevant role in the implementation of measures, as well as in the monitoring of hazards, response, identification, and evaluation of risks. Therefore, the model articulates the role of local governments (governor and mayor), with relief groups, armed forces, among others, and with academic institutions, communities, industry, and public and private companies. Finally, this model responds to the UN's SDGs and fulfillment of the 2030 Agenda. The model contributes to the SDG 11 entitled "Sustainable cities and communities" [88]. Specifically, to target 11.B based on Implementing Policies for Inclusion, Resource Efficiency, and Disaster Risk Reduction.

Conclusions
The BMA suffers from natural and anthropogenic hazards that are not sufficiently recognized by government entities with jurisdiction in the territory. Electrical storms and offshore hydrocarbon exploitation are amongst these. Atmospheric pollution and the COVID-19 pandemic were identified as the most important hazards of anthropogenic origin. Experts agree that the hazard with the greatest recurrence in the BMA is flooding. This coincides with the most relevant hazards identified in all risk management plans consulted.
Channeling of the streams in the BMA has been implemented satisfactorily. However, poor urban discipline among the population and the lack of a civic culture to handle solid waste jeopardize these efforts. Several streams have been clogged due to an excess of solid waste. Therefore, the risk continues to increase.
The government's management of hazards in the BMA is classified as regular. The importance of monitoring multi-hazards and vulnerabilities following a holistic approach is highlighted, as well as the particpatory analysis of risks. A valid governance model has been designed not only for the BMA but also for other Colombian districts. The model incorporates the governance process as a key for increasing urban resilience and decision making. The model will be tested in the BMA by local institutions.

Institutional Review Board Statement:
The study did not involve humans or animals.

Informed Consent Statement:
The study did not involve humans or animals.
Data Availability Statement: Data supporting reported results can be found asking directly of the first author.