Experimental Study Of High Reynolds Number Turbulence In Atmospheric Surface Layer

Observations using the observation array of high Reynolds number turbulence, which was arranged on the dry flats of Qingtu Lake, were performed and large amounts of data of high Reynolds number surface flow were obtained. After rigorous screening and pretreatment to the measured data, the steady wind data in the near neutral flow were selected and used to analyze the mean velocity profile, distributions of turbulence intensity and Reynolds shear stress, spectra and turbulence structures. The results show that the statistics characters at Reτ-O(106) agree with their variation with Reynolds number suggested by previous laboratory works and it indicate the near neutral atmospheric surface layer (ASL) behaves like a canonical turbulent boundary layer during the selected hours of data. The main works and innovations of this thesis are concluded as follows:The von Karman constant which is a key parameter in logarithm-law representation of mean velocity profile increase with Reynolds number and is about0.43at Reτ～O(106) in the atmospheric surface layer, higher than its magnitude in low Reynolds number flows. Similar as the vertical distribution of streamwise turbulence intensity, the spanwise turbulence intensity varies log-linearly with height.The large scale motions (LSMs) and very large scale motions (VLSMs) are confirmed in ASL by different ways. The VLSMs are detected clearly in the upper of ASL and their streamwise length scale of VLSMs increase with wall normal distance and their streamwise length scale reach up to10δ. The contribution of VLSMs to the turbulence kinetic energy is more than70%, which is higher than the results of low and moderate Reynolds number flow. Particularly, the pre-multiplied spectrum curves of different heights reverse at low wave-number range. It suggests that the VLSMs in upper ASL come from the outer "large eddy" moving downwards and sheared by surface, i.e., the "Top-Down" mechanism. It is different from the "Bottom-Up" mechanism which believes the VLSMs originate with self-organization of small streaks near the surface.Further, a new streamwise velocity power spectrum representation formula is proposed through analyzing the measured power spectra and their variation. It can characterize the "-1" spectral law range accurately. Based on the analysis results of spectra and the turbulence structures, a velocity prediction model for high Reynolds number flow is established. With this model, the wind velocity at a higher height can be predicted using the measured wind velocity at a lower height, and the statistics characteristics and spectra of predicted velocities agree well with the measured results.Spectrum analysis shows "-1" spectral law region in spanwise velocity power spectra, and the spanwise turbulent kinetic energy varies log-linearly with wall-normal distance.