Effects of land surface heterogeneities on the boundary structure and turbulence during LITFASS-2003: large-eddy simulations in comparison with turbulence measurements.

The german LITFASS-2003 field experiment covered an area of about 20x20 km and was carried out during may/june 2003 in north-eastern Germany. The main goal of this experiment is to derive area-averaged values of the surface sensible and latent heat flux from a large number of ground-, helicopter-, and satellite-based measurements. The site is typical for central parts of Europe with flat topography and landuse characterized by forest, lakes, and different types of crops.

We carried out large-eddy simulations for some days of the experiment over a 32x40 km domain including the experimental site using horizontal resolutions between 50 and 100 m. Our goals were to study the effect of the surface heterogeneities on the turbulence structure of the convective boundary layer (CBL) and to simulate local turbulence measurements (helicopter-borne measurements, LIDAR, scintillometer) to allow statements about how far they are representative for larger areas.

Initial conditions were taken from wind profiler soundings and radiosonde data. At the lower boundary, the temporal development of the surface sensible and latent heat flux was prescribed for the different types of landuse as given by representative measurements from the respective energy balance stations.

The temporal development and absolute values of the mean variables like mixed layer temperature, humidity, and boundary layer height compare very well with the observations.

Heterogeneity induced secondary circulations have been determined by averaging ensembles of identical LES runs, each started with different initial random perturbations in order to filter out the randomly distributed up- and downdrafts of the CBL. Secondary circulations only develop under weak mean wind conditions and have maximum vertical velocities of about 0.5 m/s, i.e. they are more than one order of magnitude smaller than the maximum values observed for the random convection. Although the secondary circulation contributes up to 15% to the area-averaged total vertical heat flux, the shape (and absolute values) of its profile remains nearly unchanged by the heterogeneities, as proved by additional runs with a mean, horizontally homogeneous heating. However, due to the differential heating from the heterogeneities, the horizontal diffusivity (derived from particle transport simulations) is found to be increased by more than a factor of two.

We also made detailed comparisons between our LES data and the turbulence measurements from the helicopter-borne probe HELIPOD, a ground-based LIDAR system, and a large aperture scintillometer. Effects of secondary circulations on these measurements as well as questions about their representativeness will also be discussed in the talk.