As many global models do, the French spectral global with variable mesh model ARPEGE (Courtier and Geleyn, 1988) doesn't often locate well the tropical cyclones. That is why the CRC (Cyclone Research Cell) of Meteo-France in La Reunion island has been developping two bogussing techniques to improve the initial conditions of ARPEGE and the subsequent tracks forecastings. The first one, following an idea of Jean-François Geleyn, uses ALADIN (the French limited area model) to locate the tropical cyclone at the convenient place : we use the real location of the tropical cyclone to make a limited area centered on this position. If the tropical cyclone forecasted by the model is within this area, we translate the dynamical fields in a non-uniforme manner so that the tropical cyclone reaches its real location and that the border's points remain at the same place. This geometrical transformation allows to easily replace the values of ARPEGE by the values of ALADIN. We only have a few experiments available, but the results are rather good : the new ARPEGE fields are not too distorted and the tropical cyclone track forecastings are better than without bogus.
The second bogussing technique is easier. We simply oblige the initialisation to accept the mean sea level pressure of the tropical cyclone's center. This pressure is calculated by the forecasters of the RSMC (Regional Specialized Meteorological Centre) using the Dvorak method (1984). After that bogus, all the model levels are modified and a good tropical cyclone structure is taken into account. The tropical cyclone track forecasting made after this bogus gives good results.
We intend to apply these two initialisation techniques at the same time and hope to improve significantly the skill of ARPEGE to forecast tropical cyclone tracks and may be its intensity too.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner