Our cities concentrate the bulk of wealth and human population. Conversely they produce 70-80% of greenhouse gases and consume about the same proportion of energy. These urban concentrations go far beyond average assessments! A city is a specific hydrological object, which nevertheless should be placed not only in its basin, but also in its changing climate. Then why urban hydrologists are still focusing on “methods to characterize hydrological process scales”, instead of better exploring the multiscaling of hydrological processes? Consequently, the mutual interactions of space-time rainfall variability with urban catchment spatial heterogeneity on the hydrological response remain largely unexplored. Throughout examples of theory-advanced multiscale monitoring and modelling of urban systems in Paris area, this communication proposes a considerable extension of the multifractal theory to be taken into account to better control the complexity of the water cycle, including its urban components, accounting for the strong heterogeneity of both geophysics and urban objects. This enables not only to significantly improve our understanding of so-called “characteristic scales” for urban hydrology, but also to promote principally new models, like Multi-Hydro. It has been developed as an interactive multi-modules software, by continuously integrating in it new functional modules, coupling them in a smart manner. Each module is a physically based, fully distributed model that represents an additional physical process and/or environmental feature. From an implementation point of view, Multi-Hydro is well adapted to take advantage of the multifractal theory to reproduce hydrological responses over highly heterogeneous watersheds; this heterogeneity is out of reach of operational, semi-distributed models and essential to increase the operational benefits.