PRELIMINARY UNCERTAINTY ANALYSIS - A PREREQUISITE FOR ASSESSING THE PREDICTIVE UNCERTAINTY OF HYDROLOGIC-MODELS

Physically-based urban rainfall-runoff models are mostly applied witho ut parameter calibration. Given some preliminary estimates of the unce rtainty of the model parameters the associated model output uncertaint y can be calculated. Monte-Carlo simulation followed by multi-linear r egression is used fbr this analysis. The calculated model output uncer tainty can be compared to the uncertainty estimated by comparing model output and observed data Based on this comparison systematic or spuri ous errors can be detected in the observation data, the validity of th e model structure can be confirmed, and the most sensitive parameters can be identified. If the calculated model output uncertainty is unacc eptably large the most sensitive parameters should be calibrated to re duce the uncertainty. Observation data for which systematic and/or spu rious errors have been detected should be discarded from the calibrati on data This procedure is referred to as preliminary uncertainty analy sis; it is illustrated with an example. The HYSTEM program is applied to predict the runoff volume from an experimental catchment with a tot al area of 68 ha and an impervious area of 20 ha. Based on the prelimi nary uncertainty analysis, for 7 of 10 events the measured runoff volu me is within the calculated uncertainty range, i.e. less than or equal to the calculated model predictive uncertainty. The remaining 3 event s include most likely systematic or spurious errors in the observation data (either in the rainfall or the runoff measurements). These event s are then discarded from further analysis. After calibrating the mode l the predictive uncertainty of the model is estimated. Copyright (C) 1996 IAWQ. Published by Elsevier Science Ltd