How to Improve the Resilience of Cities in a Disadvantaged Environment?
An Example in Haiti

Haiti is particularly prone to natural threats and disasters. Its mountainous terrain (more than 75 % of the area), where huge deforestation and subsequent soil erosion are observed, is subject to rapid propagation of floods in small coastal plains. Moreover, Haiti is on the route of major hurricanes, such as Hurricane Matthew that killed more than 500 people in October 2016, made more than 175,000 people homeless and affected 2 million Haitians. Moreover, it is one of the poorest countries in the world with a dramatically weak GDP per capita of 1200 US$ and limited government budget, resulting in poor infrastructures and disorganized administration, among others regarding hydrology and river management.

This paper presents a collaborative research between University of Louvain (Belgium) and the State University of Haiti aimed at developing local expertise and know-how in river management. A pilot site was selected, the Cavaillon River that is regularly flooding the city of Cavaillon, as an example for the whole process of flood management and mitigation.

Due to lack of accurate geographic database, a mix of classical GPS-aided surveys and original aero-photogrammetry by drone was used to build a Digital Terrain Model with a 15-cm mesh and 2-3 cm accuracy in elevation, able to give dense information even in inaccessible parts of the flood plains. Drone use revealed to be accurate and rather cheap. Using these data, two-dimensional numerical simulations were run to investigate the extent of the inundated areas for different flood scenarios with the aim of constructing flood maps to be used for management and mitigation purposes.

Haitians are subject to so many catastrophic events that a kind of resignation settled in the population, with a trend to accept the risks without operational reaction. So, a part of the work was to organize contacts with the local populations and decision-makers, considered as sometimes naïve but often full of common sense regarding the origin of floods, to understand their priorities in river training for flood mitigation. This common sense was confronted to sophisticated risk assessment methods like the Analytic Hierarchy Process (AHP) of Saaty, which is a combination of mathematics and psychology for organizing and analyzing complex decisions. Use of notions as vulnerability, hazards and risk mapping in front of a mistrustful population regarding the technical approach was challenging but interesting and this finally let us to improve the flood maps, and make these hopefully useful for the decision-makers.