Observation and numerical modeling of an exceptionally intense orographic precipitation event in southern France

On 15 June 2010, the Var department of Provence in south-eastern France was hit by exceptionally intense storms that released up to 400 mm of rain in 12 hours. The tourist town of Draguignan, located in the southwestern part of the Alps, received more than 6 months' worth of rainfall in just a few hours and was devastated by a historical and deadly flash flood. Despite a public advisory and warnings issued by the French Weather Service the day before the event, human casualties caused by these storms were terrible and included dozens of victims.

Observing and understanding the dynamical and microphysical processes occurring during these extreme orographic rain episodes is critical to developing effective flood warning systems and to improve operational weather forecasts of dangerous weather events. With respect to this goal, the French Weather Service has recently begun installing a new network of small gap-filling X-band polarimetric radars to reinforce the radar coverage in the flood-prone Alpine region, which was so far almost entirely hidden from the view of current operational fixed-based radars.

The purpose of this study is to examine the mesoscale organization of the 15 June case, as derived from a combination of radar observations collected by both large conventional Doppler radars and one gap-filling polarimetric radar recently installed in the Mediterranean Alps. The radar-derived 3-D airflow and precipitation structure accompanying this event will be analyzed to investigate and better understand the conditions that had led to such exceptional rain efficiency. Radar output will then be relied upon to assess operational forecasts of Météo-France's new high resolution (2.5 km) models AROME and AROME-WMED - a special version of AROME to be used over the NW Mediterranean basin during the upcoming international HyMeX field experiment in 2012 - as well as to evaluate very high resolution (500m) simulations of these storms performed within the research model MESO-NH.