Understanding and preparing for extreme events in a warming climate is fraught with significant challenges, particularly when modelling (the ever increasing) flash floods in small- to mesoscale catchments. While top-down modelling approaches that attempt to describe fluxes directly at the scale of the system perform rather well for riverine floods caused by saturation-excess runoff generation, models that rely on bottom-up paradigms have been successful in handling intensity-controlled runoff generation, along with associated preferential flow processes. In this work, we develop a meso-catchment (196 km 2) scale, spatially distributed process-based model based on the gradient conserving approximation of representative hillslopes. The approach helps us balance the required complexity for modelling such coupled flows with the practicalities of scaling our simulation to the mesoscale catchment. The model is used to investigate a severe flood that happened in 1994 in southwest Germany. After validation, the model is used to reconstruct flood magnitudes in the poorly gauged yet severely affected headwater regions of the catchment. The results reiterate the importance of accounting for differing gradients and landuse patterns across headwater regions and their role in runoff generation. To further investigate the implications of these findings for flood hazard assessment and to explore the reliability of design flood values in this area, we conducted additional simulations for a range of precipitation return periods, demonstrating that errors are often most pronounced at smaller spatial scales, largely due to data scarcity. Finally, given the growing interest in nature-based solutions (NbSs) as decentralized, process-oriented alternatives to conventional structural flood protection, we implemented representative NbSs at the hillslope and headwater scales and assessed their downstream effects at the catchment scale Our work has broad implications for modelling and the appropriate characterisation of overland flow storm responses at such crucial intermediate scales, given the increasing frequency of convective extremes.