Author(s): Qiang Dai; Dawei Han; Miguel A. Rico-Ramirez
Linked Author(s): Dawei Han
Keywords: Ensemble generator; Radar precipitation estimation; Rainfall uncertainties; Copul
Abstract: Ensemble generation of probable true rainfall is an elegant solution to characterize the uncertainty of radar precipitation estimation. In this study, we propose an ensemble generator to model the radar rainfall uncertainty based on copula and autoregressive filter. We firstly build an error model based on the distribution of gauge rainfall conditioned on radar rainfall (GR|RR). A number of realistic error perturbation fields with given spatial correlation are generated using copula. Then the autoregressive filter is used to impose the desired temporal correlation. The fact that the simulated uncertainty band of rainfall encompasses most of the reference rain gauge measurements and a fairly good agreement between the simulated and observed spatial and temporal dependences indicates that the proposed scheme is a statistically reliable method in ensemble radar rainfall generation.
Year: 2013