Influence of rainfall spatial variability on rainfall-runoff modelling: Benefit of a simulation approach?

- We evaluate the effect of rainfall spatial variability on runoff modelling.
- We evaluate the benefit of proceeding by simulation.
- We compute rainfall variability indexes for each event and each catchment.
- These indexes represent rainfall spatial variability influence on runoff modelling.

SummaryNo consensus has yet been reached regarding the influence of rainfall spatial variability on runoff modelling at catchment outlets. To eliminate modelling and measurement errors, in addition to controlling rainfall variability and both the characteristics and hydrological behaviour of catchments, we propose to proceed by simulation. We have developed a simulation chain that combines a stream network model, a rainfall simulator and a distributed hydrological model (with four production functions and a distributed transfer function). Our objective here is to use this simulation chain as a simplified test bed in order to better understand the impact of the spatial variability of rainfall forcing. We applied the chain to contrasted situations involving catchments ranging from a few tens to several hundreds of square km2, thus corresponding to urban and peri-urban catchments for which surface runoff constitutes the dominant process. The results obtained confirm that the proposed simulation approach is helpful to better understand the influence of rainfall spatial variability on the catchment response. We have shown that significant dispersion exists not only between the various simulation scenarios (defined by a rainfall configuration and a catchment configuration), but also within each simulation scenario. These results show that the organisation of rainfall during the study event over the study catchment plays an important role, leading us to examine rainfall variability indexes capable of summarising the influence of rainfall spatial organisation on the catchment response. Thanks to the simulation chain, we have tested the variability indexes of Zoccatelli et al. (2010) and improved them by proposing two other indexes.