The structure of turbulence over flat, homogeneous surfaces and various atmospheric conditions is sufficiently understood nowadays. Measurements of turbulence structure over these areas have led to similarity theories that describe the flow over these surfaces quite well.

However, in recent years, the efforts to describe the turbulent boundary layer have focused on more complex surfaces. Another major concern of boundary layer studies is the very stable regime, where the bulk Richardson number exceeds 0.2.

Areas of complex terrain may not allow a local equilibrium to be formed except very close to the surface, hence the application of similarity parameters requires some empirical justification. On the other hand a major difficulty in determining the similarity functions is the reliability of turbulence flux calculation from observation, specially at night, when turbulent intensity can be largely reduced.

This study is intended to verify the applicability of similarity theory by use of tower surface-layer data in very stable conditions. The experimental site is characterized by peak-through heights of typically 400 m occurring on horizontal length scale 1-2 km. The instrumentation included a 3D sonic anemometer, krypton hygrometer and slow response sensors such as cup anemometer and thermistors. Nocturnal data comprises 50 thirty-minute time series with low wind speed (< 1 m/s).