The Progress on Application of Ensemble Prediction to Flood Forecasting

Hydrological Ensemble Prediction is a forming and developing branch of hydrology. Its development can be roughly divided into two stages. The first stage is the longer term streamflow predictions stage from 1970 to the end of the 20th Century, and the second stage is to learn the concept of the numerical weather forecast applied to hydrological ensemble prediction at the beginning of the 21st Century. Compared with the single deterministic prediction meteorological, ensemble numerical forecasting can describe the uncertainly quantitatively. In recent ten years, the application of meteorological ensemble numerical forecasting to hydrological ensemble prediction on the warning of flood forecasting attaches great importance.In addition to single ensemble system, multiple ensemble system is gradually applied to hydrological ensemble forecast, even in some small basins. To study potential benefits of using the TIGGE database in early flood warning, an atmospheric-hydrologic-hydraulic coupled cascade system driven by TIGGE ensemble forecasts is set up. Some hydrological ensemble prediction systems use high resolution ECMWF-EPS or limited area EPS as weather forecasts as initial and boundary conditions. Over the past decade, many studies show that forecasts based on hydrological ensemble prediction systems not only can add accuracy but also increase the warning lead time. The European Flood Alert System and the Advanced Hydrologic Prediction Services of NOAA realize real-time prediction in flood forecast, but there are still some problems, such as large amount of calculation and massive data to handle.The various pre-process and post-process of ensemble forecasts of techniques are in the stage of exploration and verification. Besides, there are a great deal of challenges and difficulties in understanding how to make flood warning decisions based on probabilistic forecasts better.