Modelling flood for urban resilience and disaster risk reduction in Sierra Leone:
A case study

Flood Risk in Sierra Leone

Flooding in Sierra Leone is largely caused by intense, localized monsoon rainfall and, although flood waters generally recede within days, there have been events lasting up to a month (UNDP, 2012).

Flooding poses a significant and pervasive risk to communities in the country – between 1980 and 2010, over 220,000 people in Sierra Leone were affected by flooding and 90% of people affected by disaster in Sierra Leone were affected by floods (EMDAT, 2009; UNDP, 2012). Kroo Bay, one of the largest informal housing settlements in the capital city of Freetown, has flooded almost every year since 2008 due to heavy rains (Figure 1). This is exacerbated by urban expansion onto floodplains and areas at risk from landslide, with floods regularly damaging infrastructure, housing and crops, and causing loss of life and livestock (Africa Research Institute, 2015; UNDP, 2012).

Figure 1: 2015 flooding in Freetown Kroo Bay. Source: SwitSalone.

Project Objectives

Under ACP-EU funding (a development cooperation between the European Union and the countries of the African, Caribbean and Pacific Group of States), the World Bank Global Facility for Disaster Reduction and Recovery (GFDRR) aimed to support development of new hazard and risk information in Sierra Leone. The project targeted three major cities and outcomes would be used to identify priorities for disaster risk management investments for the Sierra Leone Urban Resilience project.

The objectives of the project were outlined to achieve the following:

• Provide support and increased resilience against disasters in Freetown, Makeni and Bo by building the city councils’ capacities in terms of risk knowledge, disaster prevention, and preparedness to strengthen urban communities’ resilience against disasters
• Contribute to urban planning

The Team

An international consortium of four organisations was composed and led by Arup, providing a wide range of technical and management expertise.

Local partner Geo-information and Environmental Management Services (INTEGEMS), alongside the British Geological Survey (BGS), provided data encompassing exposure, vulnerability, geo-spatial information, terrain and landslide susceptibility.

JBA's role was to lead the quantitative flood risk assessment through probabilistic flood modelling. JBA's specialism in flood science enabled Arup to access the catastrophe modelling expertise necessary for the delivery of the project.

Using JBA flood science

JBA developed a multi-city probabilistic model for the cities of Freetown, Makeni and Bo and their surrounding areas, using globally and nationally available datasets in combination with proven JBA modelling methodologies.

The model enabled the quantification of river and surface water flood risk to people, properties and assets in and around these cities, and provided the basis for landslide risk modelling.

The model comprised:

• Generation of an extensive multi-peril event set
• Hazard maps for river and surface water flood providing extent and depth information for return periods up to 1,500 years
• Vulnerability functions relating depth of water to the severity of damage (built environment dataset provided by Arup)

To allow analysis at the highest resolution possible in the study area, 30m cells were created that aligned with the 30m flood hazard data.

Project Challenges

River flow data were required as input to the catastrophe flood model. However, no adequate river flow records were available for Sierra Leone.

In order to overcome this challenge, JBA used a nearest-neighbour approach known as regionalization. Gauged sites are selected from JBA’s global catchment database according to their similarity of catchment attributes. Their parameters are then used as a substitute for estimating river flow within ungauged catchments (Figure 2). The estimated river flows produced by this process were implemented within event sets and rainfall run-off models.

Figure 2: River observation points (orange) and interpolation points (grey) across Sierra Leone.

Positive outcomes

Outputs from the modelling included:

• Economic losses resulting from damage properties and assets
• Number of people affected
• Number of fatalities

JBA’s probabilistic flood modelling enabled the city councils of Freetown, Makeni and Bo, alongside the World Bank, to better understand the scale of flood-related risk in Sierra Leone. Outputs can be used to make recommendations for risk reduction, including informing future strategies around urban planning and disaster preparedness, and for highlighting key areas at risk - as the project was nearing completion, a major landslide occurred in an area that had been identified as at risk in the outputs of the model.