Dynamic downscaling based on weather types classification: An application to extreme rainfall in south-east Japan
By J.F Vuillaume and S Herath
- Optimization of weather forecast systems is crucial for efficient real-time weather predictions.
- This research focuses on optimizing the physical scheme of numerical weather prediction for the south-east coast of Japan, specifically the Tokyo area.
- Weather types are used to cluster and optimize the physical scheme.
- Calibration and validation of physical schemes were done using observed rainfall at gauge stations near Tokyo, categorized by weather circulation.
- 24 ensemble members and 20 heavy rainfall events from 4 weather types were used for validation.
- The physical scheme ensemble consisted of micro-physics, cumulus, planetary boundary, and radiative schemes, simulated using the weather research and forecasting model.
- Limited physical scheme variability was observed within a weather type cluster for a station pool, but significant variations were present among stations.
- Rainfall Cumulative Probability Distribution Function curves were computed, showing wide differences due to clustering.
- Results allowed for:
- Selection of optimal physical scheme for each weather type.
- Development of weather type-specific bias correction curves.
- Evaluation of spatial distribution of regional bias correction.
