GLUE Based Uncertainty Estimation of Urban Drainage Modeling Using Weather Radar Precipitation Estimates

Distributed weather radar precipitation measurements are used as rainfall input for an urban drainage model, to simulate the runoff from a small catchment of Denmark. It is demonstrated how the Generalized Likelihood Uncertainty Estimation (GLUE) methodology can be implemented and used to estimate the uncertainty of the weather radar rainfall input. The main findings of this work, is that the input uncertainty propagate through the urban drainage model with significant effects on the model result. The GLUE methodology is in general a usable way to explore this uncertainty although; the exact width of the prediction bands can be questioned, due to the subjective nature of the method. Moreover, the method also gives very useful information about the model and parameter behaviour.