Comparison of flood zones and evaluation of water depths of buildings for the flood of July 2021 in the Vesdre watershed

Abstract

This master thesis addresses the mapping of flood extents and water depths in the Vesdre watershed, Wallonia, in response to unprecedented floods in July 2021.

The study emphasizes the significance of comparing outcomes from a sophisticated 2D model (WOLF2D) with a simplified conceptual approach implemented by the SPW, which involves superimposing diverse data sources. The SPW cartography integrates digitalisation primarily based on aerial photos and a modeling created with an inverse distance weighting interpolation utilizing water depths from buildings measured by SPW employees.

The comparison, facilitated by coefficients of correlation considering the size of evaluated wet and dry areas, reveals satisfactory global and local similarities at the Vesdre watershed scale. However, notable discrepancies emerge, particularly in highly urbanized areas where WOLF exhibits complexity, and in rural areas distant from measured water depths for the SPW modeling. This modeling tends to overestimate flood extents and water depths globally, struggling with dynamic flood effects and relying predominantly on topography. In Verviers, the complexity of the built-up environment contributes to poor correlation coefficients. Indicators comparing measured water depths of SPW buildings and WOLF-simulated water depths demonstrate varying results. A nuanced analysis, adjusting parameters such as minimal water depth consideration, zone dimensions around buildings, and riverbed inclusion, permits a closer alignment of the WOLF water depths with the on-site measurements. Pepinster and Eupen both located on a confluence of two important streams present results with high errors.

Lastly, through a literature review of 0D methods also called simple conceptual approaches compared to 2D models, the results of SPW modeling are considered quite favorable.