Namibia is vulnerable to various anthropogenic and natural hazards, especially flooding (Reid et al. 2008). Floods are an annual recurrence, worsening every year and gravely impacting the northern and northeastern regions of the country (UNDRR 2019). In the past, flood-fighting initiatives were almost solely centred on relief efforts. The colonially inherited South African Civil Defence Act and Ordinances, or the Civil Defence Act 1966 as it was known in Namibia, governed all disaster management matters until the promulgation of the Disaster Risk Management Acts (DRMA) in 2012 (GRN 2013; Republic of South Africa 1966, 1969).

The Act postdated the undeniable need for policy-driven hazard risk reduction strategies following the devastating impacts of the 2009–2011 flood disasters in the country (Hosseini-Boroujeni 2019) and is supported by two other pieces of legislation, the National Disaster Risk Management Plan (2011) and the National Disaster Risk Management Policy (2009) (GRN 2011, 2013). The development of this framework (comprising these three pieces of legislation) presented a step towards the global paradigm shift from disaster response to an integrated multi-hazard strategy focusing on hazard risk reduction (Van Niekerk 2008). Aligning well with several relevant international agreements, that is, the SFDRR, the African Regional Strategy for Disaster Risk Reduction (ARSDRR) and the Kyoto Protocol, the framework outlines a well-structured approach to DRR (African Union 2004; UNDRR 2015; United Nations 1998), even highlighting the relevance of and promising the immediate establishment of early warning systems (GRN 2013). However, several reports suggest that this is yet to manifest empirically, as several communities have continued to bear the seemingly unmitigated brunt of flood events (Mabuku et al. 2018; Mashebe 2015; Shaamhula, Smit & Van der Merwe 2021).

The main objective of the Flood Early Warning System (FEWS) in Namibia is to provide at-risk communities with information and alerts of impending hazards and directives for evacuation (Lumbroso 2018). To this end, several organisations operating in the managerial and technical facets of disaster preparedness collaborate to provide civil society with flood information (Mandl & Frye 2012). These systems require an accurate and timely forecasting system, robust and continuous communication network, reliability and collaborative action to meet this objective (Cools, Innocenti & O’Brien 2016). However, floods have persisted in frequency and impact, with failed evacuations and losses often credited to poor communication and coordination (National Planning Commission 2019).

In the assessment for African floods in 2021, the International Federation of Red Cross and Red Crescent Societies lists Namibia as the country with the highest population ratio exposed to flood risk (IFRC 2021). Approximately $100 million of GDP, corresponding to about 0.9% of the total GDP per annum, is at risk of direct loss by flood impact (UNDRR 2019). Apart from experiencing major flooding from 2008 to 2011, in which over $136m in damages and $78m in direct economic loss were sustained in 2009, Namibia is exposed to high-impact flooding annually, impacting over 54% of the total population (Government of the Republic of Namibia 2009).

The seemingly unmitigated flood recurrence has raised questions about why DRR has yet to receive due attention at all governmental levels in Namibia. Why do developed policy-based institutional and communication networks not deliver on the ground? What are the policies, structures and institutions that hinder effective DRR in the country? To help address these concerns, this paper systematically assesses the roles, responsibilities, and success of the focal disaster management institution, the DDRM, in streamlining DRR across all governmental levels and how this has translated in reference to flood early warning in Namibia.

The study adopted a case study approach, which included the review of literature, the legislative framework (Figure 1) and supporting policies, supplemented with qualitative data obtained through purposive key informant interviews (KIIs) and focus group in the Kabbe constituency (Figure 2). Characterised by annual riverine and flash floods and currently categorised as the area most sensitive to flood impact in Namibia, the researches deemed it most appropriate for the assessment. The researchers assessed the DDRM’s establishment of effective institutional coordination and communication networks for flood early warning in Namibia by collecting data on the extent of DRR streamlining through (1) an overview of the technical operationalisation of the FEWS through KIIs with DRR officials and (2) the experiences and perspectives on flood risk and reduction of impacted communities through FGDs. The information was then compared against the strategy outlined in the DRMA. The combination of data collection tools also allowed for cross-validation of information, increasing the study’s reliability.

Acknowledging the composite nature of flood disasters, the selection of nine KII participants comprised a diversity of national to local level government DRR officials, the Namibia Red Cross, which actively collaborates with communities to manage flood risk and community leaders (i.e. village chiefs and church leaders) directly involved in early warning systems operations. The interviewees were selected based on their expertise, experience and involvement at particular levels within the flood early warning process.

The interviews were based on semi-structured questions structured around the DRM Act 10 of 2012 which outlines DDRM’s duties. Some of the questions included: What is the role of the DDRM in establishing stakeholder relations and communication flow among early warning system institutions at all governmental levels? Has the DDRM established DRR institutions for early warning coordination and information management at all levels of government? How does the DDRM ensure community engagement in early warning systems processes across all governmental levels? What is the extent of the DDRM’s implementation of its roles and responsibilities outlined in the DRMA and support policy concerning the flood early warning system?

Assisting communities in successfully combating flood risk was the impetus for assessing the DDRM’s implementation of the DRMA. As such, the study saw it necessary to draw on communities’ perceptions, experiences and opinions to ascertain good practices and challenges and establish how society and its concerns are integrated into the institutional framework. To conduct this analysis, eight FGDs were conducted with residents from the four most flood-affected communities in Kabbe, located both up and downstream in the study area and homes of two different ethnic groups, which represented possible disparity in, for example, risk perceptions, preparedness methods, influence of customs of dealing with floods, among others. A total of 59 participants were selected based on:

The selected participants included members from community-driven organisations such as the Red Cross-established Community Disaster Risk Management Committees (CDRMCs), members of community organisations and regular residents. The diversity within the selected sample provided additional information on the extent of DRR decentralisation by enabling the assessment of the level of community involvement in the planning, communication and response of the early warning process. Some questions asked included: Are residents aware of the national flood early warning system? Are residents aware of what governmental institutions to contact during a flood? How often are the residents engaged in government-led flood DRR initiatives? What are the main sources of flood information in your locality?

The area was selected for two reasons:

Ethical clearance was granted by the General/Human Research Committee at the University of the Free State under Research Project Title: Analysis of the Operational Flood Early Warning System in Kabbe Constituency, Zambezi Region – Namibia. Ethical clearance number: UFS-HSD2022/0400/22.

The study confirmed that at the national level, the DDRM is the country’s custodian for all DRM affairs and is considered the government’s disaster management agency. Assuming its primary function as stakeholder coordination, participants at the national level agree that, at this level, coordination is somewhat effective. However, interviewees suggested the establishment of an effective information management and communication system and a better definition of roles and mandates to strengthen coordination and communication among different agencies. Although not legally mandated, the DDRM is responsible for several sectoral units that feed into the NDRMC flood early warning decision-making process (i.e. NGOs, I/NGOs, private sector). Although these agencies maintain contact during DDRM-led meetings, communication linkages among themselves are often limited:

Furthermore, data sharing among agencies is minimal. However, some institutions produce information, for example, risk maps, that can be used in flood forecasting and monitoring, but these agencies are not formally involved in the FEWS process. In addition, the FGDs coroborate these findings by pointing out the lack of government-led risk awareness and communication efforts supplemented by the observable unavailability of risk maps in regional and local level offices and communities. On the contrary, interviewees agree that existing formal communication networks work relatively well at the national level. However, the same cannot be said for subsequent governmental levels.

The study established that a Regional DRMC (RDRMC), tasked with all DRR efforts, including FEWs initiatives, has been replicated and is fully functional. The KIIs with officials from the Zambezi Regional Council, Local Authority and the Namibia Red Cross from field visits to the Kabbe confirmed that the Regional DRMC was responsible for FEWS activities in Kabbe and that the Regional Disaster Management Committee (RDMC) met occasionally. The KIIs further revealed that the personal capacity of the RDMC was limited to a single permanent staff member, with no designated department and limited resources available for flood early warning activities, including emergency response.

The participants also revealed the ongoing response-driven nature towards flood risk management by the DDRM and inadequate departmental awareness of DRR awareness and long-term preparedness and resilience. This can be attributed to inadequate resources, planning and training and the need to prioritise time and available resources for response rather than preparedness and mitigation:

A major criticism levied by sub-national participants is that the RDRMC comprises Zambezi Regional Council staff primarily and the committee often meets on an ad hoc basis when flood disasters or emergencies occur. As such, ensuring DRR implementation and ownership at these governmental levels is often difficult as very little time is assigned to planning and effecting long-term strategies. Participants further revealed recurring absenteeism and substitutions by unversed proxies at RDRMC meetings, demonstrating that not all committee members prioritise their duties. This is partly attributed to members’ perception of the meetings as an added burden to their ‘regular workload’.

However, good practice was observed in the open and functioning communication lines between the DDRM and RDRMC. The regional council official also praised the DDRM’s involvement in regional DRR matters. On the contrary, a major contingency plan that was the development of the Regional Emergency Operations Centre (REOC) for which preparations began in 2013, is yet to be established. Furthermore, communication networks among regional institutions are non-existent and emergency response coordination often takes the form of personal favours among officials from different institutions or assistance is requested from the DDRM:

Another challenge brought forward at the regional level is poor DRR decentralisation and the red tape that delays FEWS’ functions at the regional level. The DDRM approves sectoral budgets, which are often delayed and contingent upon what issues the government considers priorities. Furthermore, although the regional office is extended the mandate to plan flood DRR initiatives, budgetary constraints limit their scope of implementation to emergency response. Although the DDRM is supposed to ensure that the RDRMCs coordinate and implement flood preparedness, mitigation, response and recovery efforts, the Namibia Red Cross independently steers these efforts:

The KIIs and FGDs revealed significant gaps in local and community-level DRR prioritisation and implementation. Firstly, constituency and settlement DRMCs have yet to be established under the guidelines provided by the DRMA; secondly, DRR functions are the designation of a single official at the constituency level with no personnel at the settlement level, an indication of limited capacity. It is expected that NGOs will compensate for capacity building in this area. Similarly, government-established community DRMCs and local authority volunteers are non-existent. Although communities acknowledge receiving flood warnings and relief from authorities at times, communities are unaware of the existence of a national FEWS. This can be attributed to the lack of DRR ownership at the local administrative level:

Although, in theory, communities are regarded as the most relevant stakeholders in the FEWS process, the FGDs and KIIs reveal that, in practice, this is not the case. Residents and government officials confirm that communities are only engaged within the FEWS chain during the alert and evacuation process. Regional officials attribute this to a lack of funding for conducting any type of outreach but argue that the Red Cross conducts several important functions, for example, risk mapping, evacuation planning, identification of vulnerable areas, among others.

Moreover, the FGDs highlighted poor knowledge of DRR legislation and practice among most residents and a lack of opportunities for community engagement. While acknowledging the sense of ownership provided by the Red Cross’s initiatives, residents simultaneously expressed feeling left out and disconnected from decision-making by national and sub-national government institutions. However, it is imperative that these grievances are contextualised within the general scope of community collaboration in developmental matters in Namibia.

Firstly, participatory natural resource management (involving communities) is an effective and relatively common practice in Namibia and several NGOs support and manage community-based developmental programmes (Naidoo et al. 2011). Secondly, many participants belonged to community organisations, committees and volunteer groups and contrasted their decision-making power within these organisation against their experiences or lack thereof with the government:

Conversely, regional and local authority officials observed bureaucracy, inadequate financial and human resources and slow-paced decision-making as challenges for inducing community participation. Whereas national-level officials promise a change in this direction, regional and local authorities expressed frustration at the inability to execute their own decisions. In addition, the available scope for participatory decision-making is limited to necessity and traditional authorities who may or may not consult their residents.

A key-note feature of the study was the prevalence of female FGD attendees. Although it can be argued that men in rural communities often travel for work, the prevalence of women in the village committees proves that their needs are prioritised. Female participants expressed feeling well-represented in community organisations and were included in decision-making. The KII, with an NGO, referred to women as primary contact points and beneficiaries of flood warnings in rural areas. Finally, the KIIs revealed that the consolidation of national early warning systems was still in the planning stages, explaining some gaps identified in multisectoral communication.

This study presented several limitations:

These limitations were delimited by the fact that: