* Preparations for assimilation of new satellite data including ENVISAT Ozone data, AIRS data MSG data SeaWinds Data Microwave imager rain-rate data
* Improvements in the pre-conditioning of 4D-Var, and in the treatment of the trajectory
* Improvements in observational quality control
* Incorporation of more complete adjoints of the physical parametrization in the inner loops to permit assimilation of rain-rate data and to prepare the assimilation of cloud data.
* Reduced-rank Kalman filter to enable the cycling of the 4D-Var flow-dependent structure functions.
* Developments to cope with the forthcoming vast volume of satellite data.
Currently, the ECMWF Ensemble Prediction System generates a 50-member ensemble of T255 integrations once daily. The paper will present the current status on a number of projects to improve the ECMWF Ensemble prediction system, including inter alia:
* Tropical Cyclone Perturbations
* Humidity singular vectors
* Comparison of a twice-daily 50-member ensemble with a once-daily 100-member ensemble
The paper will discuss the results in the context of the development of a system for forecasting severe (i.e. high-impact) weather.