ID#018

The role of microphysical parameterisations and model grid size on the formation of extensive precipitation events in southern France

Wolfram Wobrock, Claire Saugues, Andrea I. Flossmann
Laboratoire de Météorologie Physique, Université Blaise Pascal/CNRS, Clemont-Ferrand - France

Recently, an initiative of the French ministry of Research was generated to study the risks of severe precipitation and resulting inundation. In the framework of this initiative the basin of the mountainous region "Cevennes-Vivarais(southern France). " was selected as a centre of interest. As this region is repeatedly haunted by severe precipitation events an observatory equipped with rain gauges and a precipitation radar network monitors it.

As a part of the initiative, the severe precipitation event on 13/14 October 1995 was analysed. For this period we performed a 3D cloud resolved modelling with the cloud scale model of Clark and Hall. The observed convective precipitation had a duration of more than 15h, and rain locally accumulated up to 300mm during this period with severe flooding that occurred.

In a first approach we tested the role of grid resolution on the model performance. Here, different scales of nesting and different horizontal and vertical grid increments were used. Furthermore, we applied different microphysical schemes in the same dynamical framework. The main focus here was put on the role of the cold cloud microphysics for this long lasting case of deep convection. Consequently, different microphysical parameterisations of cloud ice and precipitating ice were compared: (1) the scheme of Murray-Koenig, (2) the detailed ice microphysics of Farley and Orville, and (3) the parameterisation of Lin.

The different model outputs were compared with the observations taken during this day by the rain gauge network and the local precipitation radar.

It was found that influence of the applied microphysical ice parameterisation dominated over the role played by the grid resolution (between 300m and 900m).