Evaluation of multi-model rainfall forecasts for the national 7-day ensemble streamflow forecasting service
2018
The Bureau of Meteorology is planning to launch its upgraded national 7-day ensemble streamflow forecast service in mid-2019. This service will fulfil the growing demand for ensemble streamflow forecasting and bring a greater level of accuracy and reliability when compared to the existing 'deterministic service'. It will enable end-users to make more confident water management decisions by considering the likelihood of different streamflow forecast outcomes within a risk assessment framework. Ensemble streamflow forecasts are generated using post-processed multi-model Numerical Weather Prediction (NWP) ensemble rainfall forecasts. These streamflow forecasts are subsequently post-processed using a comprehensive multi-stage error correction model. The forecasting system evaluation was undertaken by comparing rainfall and streamflow forecast performance using four different NWP rainfall products for 26 catchments located in various hydro-climatic regions across Australia. It reveals that the post-processing reduces bias and improves the reliability of rainfall forecasts as well as the corresponding streamflow forecasts. Streamflow post-processing further improves the accuracy and reliability of forecasts significantly at shorter lead-times and the impact declines with the lead-time. Overall, the use of rainfall forecasts of the European Centre for Medium-Range Weather Forecasts (ECMWF) presents relatively better accuracy and reliability of streamflow forecasts at the catchment scale in many occasions compared to the other three products evaluated in this study.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
1
Citations
NaN
KQI