A GIS-based flood risk tool for Jamaica: the first step towards a multi-hazard risk assessment in the Carribean

The Caribbean is known to be one of the most hazard-prone regions in the world. Hurricanes, flooding, storm surges, earthquakes and landslides lead to extensive material, human and economic losses in the region. The growing intensity of these hazards, combined with the consequences of climate change, rapidly increases the concern among decision makers. Although many researchers have succeeded in developing a single-hazard risk assessment that accurately estimates the risk of one type of hazard, the complexity of the relation between the different types of hazards is causing difficulties in the development of a multi-hazard risk analysis. This research aims to develop such a model. In a first step, the consequences of each type of hazard will be assessed individually, starting with riverine flooding. In the next step, the methodology used in this tool will be assessed and modified to fit other types of hazards. Finally, all single-hazard tools will be combined into a generic multi-hazard risk assessment tool for the region. In Jamaica, local governments use a flood risk methodology that is based on building water defence structures to evacuate the water as quickly as possible. This methodology, however, causes bigger damages downstream. Another method, based on minimizing the consequences of the overall flood, is already in use in many countries. In the Flemish region of Belgium, it is implemented in a tool called LATIS and has already proven to decrease losses after a flood event. Therefore, this risk-based methodology is used as the base for developing the Jamaican flood risk tool. The biggest concern during this research is the lack of data in the region. The methodology used, is based on the Flemish flood risk tool and the acquired data is thus very elaborate. During the implementation of the methods for the Caribbean, especially the lack of sufficient rainfall data and adequate damage functions has proven to result in less accurate damage and vulnerability maps