Spatial Resolution and Parameterization of an Urban Hydrological Model : Requirements for the Evaluation of Low Impact Development Strategies at the City Scale

The impacts of urbanization on the environment are widely acknowledged. Urban development implies the increase of impervious cover that is unable to provide hydrological functions of natural catchments, such as infiltration, evapotranspiration, and attenuation. Thus, increasing imperviousness alters the hydrological cycle, which is seen as increasing runoff volumes, larger runoff peak rates, and more severe pollutant loads. To mitigate these impacts, Low Impact Development (LID) tools have been developed. These aim to mimic hydrological processes of natural catchments reducing runoff and pollutant loads close to the source. Hydrological modelling is one option to evaluate the performance of LID strategies before implementation. However, such an assessment requires an explicit modelling strategy. Such a strategy implies the availability of detailed spatial data for model development and rainfall-runoff data at a sufficient temporal resolution for model calibration. Such data are often not available for larger urban catchments hampering the evaluation of LID strategies at the city-scale. This study presents a methodology for the parameterization of a hydrological model to a large urban catchment where the explicit simulation of various LID tools for urban stormwater management is supported. Several aspects of urban hydrological modelling were investigated under consideration of limited availability of data and with the focus to retain the possibility for explicit LID simulation. These aspects include (i) the implications of surface discretization approaches on simulation results, (ii) the impact of spatial resolution on simulated runoff, (iii) the impact of an automated DEM-based delineation approach on catchment properties and simulation results, and (iv) the parameterization of a large, ungauged catchment. Finally, a green roof model was parameterized to allow its implementation into a large scale urban catchment model. While both a coarser spatial resolution and a DEM-based delineation affect the simulation results, conducted simulations allowed the determination of a suitable threshold for the reduction in spatial resolution that can reasonably well replicate the dynamics of urban runoff. Results concerning the parameter inter-changeability show that the transfer of parameter values calibrated to monitored study catchments using a surface-type based surface discretization is a feasible way to parameterize large, urban catchments.