Influence of land use and land cover datasets on dam-break modeling in the semiarid region of Brazil

Abstract Accurate estimation of Manning’s roughness coefficient is critical for reliable hydraulic modeling of dam-break floods. However, for this type of accident, the lack of historical flood data makes the definition of Manning’s roughness coefficient challenging. This study utilizes high-resolution land use and land cover (LULC) data to determine Manning’s roughness coefficient values for application in dam-breaking studies. This study assesses the influence of a regional high-resolution LULC dataset (MapBiomas) and two high-resolution LULC datasets (Dynamic World and ESRI 10 m Annual) on hydraulic parameters related to flood wave propagation and flood hazard assessment. The simulations indicated substantial variations in flood behavior across the Dynamic World generated predominant regions with elevated Manning values, resulting in expanded flood zones and heightened flow attenuation simulations. Conversely, the ESRI 10 m Annual exhibited predominant regions of lower roughness, leading to simulations with diminished flood areas, reduced propagation times, and decreased attenuation of peak flows. Compared with the other LULC datasets, MapBiomas demonstrated a balanced representation of the Manning coefficient’s domains and yielded intermediate outcomes. These discrepancies highlight the challenges associated with accurately determining Manning’s values to ensure precise outcomes in flood modeling. The quality of this modeling is critical for identifying risks, formulating emergency responses, and implementing effective mitigation strategies in downstream regions.