J5.1
The Canadian regional data assimilation and forecasting system

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Monday, 24 January 2011: 1:30 PM
The Canadian regional data assimilation and forecasting system
2B (Washington State Convention Center)
Luc Fillion, MSC, Dorval, PQ, Canada; and A. Patoine, E. Lapalme, N. Benbouta, P. Vaillancourt, M. Tanguay, S. Macpherson, M. Lajoie, M. Desgagne, and V. Lee

The current operational regional data assimilation and forecasting system is based on a 3D-Var analysis approach and uses the Canadian GEM global variable resolution model (15 km horizontal resolution over the North American continent). This system is to be replaced by a new system presently in parallel tests at the Canadian Meteorological Center and based on an improved 3D-Var formulation together with a limited-area 15km version of the GEM model.

We present a further upgrade of this latter version where the analysis increments of 3D-Var are performed at higher resolution (65 rather than 100 km) and the resolution of the model is increased from 15 to 10 km.A new vertical extension of the analysis/forecast grids allows increased vertical resolution in the troposphere. A new feature is the "vertical driving" strategy implemented both in the 3D-Var and GEM limited-area model. The driving model being a global 55 km GEM version with stratospheric extent. Ground based GPS data were also included. Extensive testing of this new system for winter and summer cases in view of operational implementation will be presented.