Extreme precipitation is globally one of the most significant natural disasters, both in terms of economic costs and the number of affected people. Belgium is no exception to this, and densely populated urban areas are particularly vulnerable due to their high percentage of impermeable surfaces. Furthermore, climate change exacerbates the severity of extreme precipitation events. This research profile aims to provide the scientific basis for impact-based warning systems covering various time scales while capturing associated uncertainties. It will achieve this through a unique integrative approach that considers the entire chain from observations to risk assessment for end-users, fully leveraging probabilistic information at each step. Ultimately, the goal is to provide information that end-users can integrate with exposure and vulnerability maps to thoroughly assess risks and their probabilities.