The fine scale MesoNH simulations at 2.5km horizontal mesh are validated with the surface observations, the RAIN product derived from the operational Monte-Lema radar and the reflectivity and wind 3D-fields derived from the MAP-research ground-based and airborne radars. The validation of the two IOPs simulations is good and allows a detailed analysis. We use the Mesonh model as a numerical laboratory by performing sensitivity experiments to quantify the respective role of each echanisms previously identified as a potential cause of this reverse flow: wet or ry drainage flow, blocking upstream the Alps.
The different sensitivity experiments show that these two IOPs obey to two different conceptual schemes: 1) in agreement with the conclusion of Steiner et al (Map special issue,2003) performed with radar and surface data, the reverse flow of IOP8 mainly originates from diabatic effects i.e. the subsidence caused by melting of precipitation particles over the orography. 2) as illustrated in Brig meeting (Asencio et al,2003), the IOP3 reverse flow mainly originates from dynamical effects i.e. the upstream flow intensity and the transition from blocked to unblocked regime over the Alps.