On the turbulence structure over highly complex terrain: key findings from the MAP-Riviera project

In this contribution we try to summarize the most salient findings (so far) from the MAP-Riviera project in order to stimulate discussion on whether these might be general (similarly observed in other circumstances) or very specific to the sites in the Riviera Valley.

During MAP SOP in summer/fall 1999 very detailed observations were made in the Riviera Valley in southern Switzerland. These included detailed turbulence measurements on a cross-section through the valley as well as research flights accompanied by radio soundings, tethered balloons, temperature profiling etc.

In fair weather conditions (valley-wind development) no breakup of the inversion is found and the layer of significant turbulence is confined to a few hundred meters from the ground. The boundary layer height does not correspond to the height of the inversion top. This thermodynamic structure is supported by a secondary cross-valley circulation, which is due to the complicated inflow conditions into the valley.

Large spatial inhomogeneity is found in the surface forcing (net radiation) and consequently also in the near-surface turbulent fluxes across the valley. For example the surface heat flux assumes a maximum that can vary by a factor of three in magnitude and the maxima occur at different times. Consequently, the simple surface energy balance equation is not closed due to neglecting advective terms. This has possibly important consequences when estimating latent heat fluxes from, e.g. the Bowen ratio method as done in many hydrological models.

The interaction between slope and valley winds leads to the occurrence of directional shear in addition to and at least equal in magnitude as the frictional shear. Numerical models, in which traditionally the momentum flux is diagnosed from friction alone, seem therefore to be prone to underestimating the total momentum exchange and hence also turbulent exchange of scalars.

Profiles of turbulent kinetic energy and other turbulence statistics throughout the valley atmosphere show an astonishing degree of similarity when scaled with the appropriate surface fluxes. It appears that the energetically most active surface within the valley rather than the ‘surface point beneath the profile’ hereby drives the turbulence structure within the entire valley.