Résumé:
This research demonstrates how urban design can enhance urban resilience to flood risks, particularly under current and future climate change conditions, and develops an illustrative case for an urban resilience experiment in an Algerian city, the city of Tebessa, which occupies a strategic position in the extreme east of Algeria, a crossroads city on the border of the desert and Tunisia. Over the past two decades, flood risk management has evolved from structural defenses to a broader approach based on resilience. The latter emphasizes the ability to resist, recover, adapt and transform in response to flooding events.
This thesis is divided into three parts: Firstly, the analysis of the urban environment of Tebessa city using machine learning models (XGBoost, Random Forest, KNN) to predict vulnerability to flooding; Secondly, the study of environmental climatology to forecast future rainfall and water levels in the Wadis using GIS; Thirdly, to highlight the relationship between urban expansion and flood risk over the period 1990 to 2023. This research also simulates urban development scenarios in the area covered by land use plan no. 9A. The final results show that the integration of modularity, redundancy, diversity, efficiency and connectivity in urban design considerably enhances the resilience of Tebessa city, the subject of this research, to flood risks.
