Multi-Model Grand Ensemble Hydrologic Forecasting in the Fu River Basin Using Bayesian Model Averaging

Bo Qu,Florian Pappenberger,Xingnan Zhang,Yuanhao Fang,Tao Zhang

doi:10.3390/w9020074

Abstract

Statistical post-processing for multi-model grand ensemble (GE) hydrologic predictions is necessary, in order to achieve more accurate and reliable probabilistic forecasts. This paper presents a case study which applies Bayesian model averaging (BMA) to statistically post-process raw GE runoff forecasts in the Fu River basin in China, at lead times ranging from 6 to 120 h. The raw forecasts were generated by running the Xinanjiang hydrologic model with ensemble forecasts (164 forecast members), using seven different “THORPEX Interactive Grand Global Ensemble” (TIGGE) weather centres as forcing inputs. Some measures, such as data transformation and high-dimensional optimization, were included in the experiment after considering the practical water regime and data conditions. The results indicate that the BMA post-processing method is capable of improving the performance of raw GE runoff forecasts, yielding more calibrated and sharp predictive probability density functions (PDFs), over a range of lead times from 24 to 120 h. The analysis of percentile forecasts in two different flood events illustrates the great potential and prospects of BMA GE probabilistic river discharge forecasts, for taking precautions against severe flooding events.

Highlights

Ensemble forecasts from a single ensemble prediction system (EPS)/model only account for some of the uncertainties in initial conditions and model physics [1]
grand ensemble (GE) prediction systems have become a basis for probabilistic weather forecasts at many operational centres [3,4,5,6]
The Bayesian model averaging (BMA) model described above is a standard method that does does not take into account the special circumstance in this study, that not take into account the special circumstance in this study, that might might bring about a serious impact on the forecasting results

Summary

Introduction

Ensemble forecasts from a single ensemble prediction system (EPS)/model only account for some of the uncertainties in initial conditions and model physics [1]. There has been a move to integrate this GE of weather forecasts into coupled meteorological-hydrological modelling systems, in order to provide improved early flood warnings [7]. Pappenberger et al [8] applied seven meteorological EPSs to the European Flood Alert System (EFAS), to hindcast the October 2007 flooding event in the Danube basin. He et al [9] extracted daily precipitation data from six different numeric weather prediction systems (NWPs), to drive the Xinanjiang hydrologic model for forecasting river discharges during three summer flood events in

Objectives

Methods

Results

Conclusion