Monday, 11 January 2016
The EMC Ensemble Group has been working closely with the ESRL/PSD Forecast Modelling and Development Team to transition a stochastic physics suite into operations for a future upgrade of the NCEP global ensemble (GEFS). The typical pattern of this collaboration is that ESRL produces the bulk of the code and initial testing, and then hands it off to EMC to perform more extensive tests and parameter tweaking. At each step both teams, along with a few outside experts on model uncertainty, assess the results and brainstorm solutions to deficiencies. This then informs future development by ESRL and EMC, both in the short term (accelerating R2O) and long term (pseudo-O2R), repeating the cycle. Here we present a few examples of deficiencies noted in the initial test configuration of stochastic physics for GEFS and how these were addressed by ESRL and EMC to improve the pre-implementation configuration.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner