Flood risk assessment and participative process in the data-scarce Metuge district of Mozambique: An exportable approach

Low-income countries are the most vulnerable to floods; moreover, the occurrence and intensity of these disastrous events are progressively increasing worldwide. Quantitative flood risk assessment is the first and primary step that can be made to support local decision makers towards effective flood risk management. However, the lack of data that typically characterizes the Global South hinders the implementation of methods developed for data-richer contexts. The present study aims at proposing a comprehensive and exportable methodology for flood risk assessment, responding to the challenge of data and method deficiencies, by referring to global sources, freely available tools and relying on an intensive field survey. The methodology, including hydrological, hydro-dynamic and damage modelling, and a multi-level participatory process, was developed for a flood prone area in northern Mozambique that is crossed by the Muaguide river. The latter presents a bed completely buried by sediment in many stretches, causing extensive floods which have been more and more frequently in the last years, hitting the surrounding region. The obtained results (Annual Average Damage to roads and buildings of about 300,000 USD/year and an average of about 2000 people at risk per year) increase the knowledge of flood risk in the investigated area and can be a useful support for the design and implementation of effective mitigation measures at local and regional scales. • A framework to conduct a fully quantitative flood risk assessment in low-income countries has been developed. • The major challenge faced is the is the lack of data. • A coupled hydrological-hydraulic modelling has been implemented using Global data and freely available softwares. • The field survey is essential to perform a damage assessment at the micro-scale. • A multi-level participatory process is crucial to improve modelling assumptions and validate results in data-scarce contexts.