We present a study of a case of nontornadic supercell that took place over France on 30 may 1999. That day, strong gust winds were produced over Paris and its suburbs, leading to several casualties. Such storms, particularly common in the Great Plains and Midwestern regions of the United States, have been less studied under other environmental context and especially over France. However, occurrence of such severe storms is not so rare, in France, even if few of them produce tornado (not more than two F2 per year). Hence, a better understanding of these thunderstorms in the setting of the French topography and climatology is of critical importance for improving the forecast of such events.

Using a mesoscale numerical model, we simulate this 30 may 1999 supercell case. The simulation is made from a three dimensional non-hydrostatic mesoscale numerical model, called Meso-NH. It uses a two way interactive grid nesting with two nested domains (10 km and 2.5 km horizontal meshes). The initial state is provided by the French operational data assimilation system. Starting from this non-homogeneous large scale initial state, the model succeeds in simulating a storm splitting that leads to a right-moving storm. This is encouraging for the future generation of operational numerical weather prediction (NWP) models which are planned to run at few kilometers resolution in the next decade.

After having compared the simulation with the observations, we made use of the simulation in order to better understand the storm dynamics. First, we identified the triggering mechanism of the initial cell. A vorticity analysis, with emphasis on stretching and tilting terms of the midlevel vertical vorticity, was carried out. Backward trajectories allowed us to underline this vorticity source. Finally, we studied the interaction of the low level mesocyclone with a thermal boundary. Most of the results compare well with previous results on American great Plains supercells obtained from theoretical studies with idealized homogeneous base state.