COEFFICIENTS OF EDDY DIFFUSION OF MOMENTUM AND HEAT IN ATMOSPHERIC BOUNDARY LAYER: NUMERICAL STUDY

The changes in turbulent eddy mixing in the atmospheric boundary layer (ABL) are investigated with the use of the mesoscale RANS turbulence model with account for effects of internal gravitational waves, which support momentum transfer under condition of very stable stratification. A focus was put on analysis of behavior of the coefficients of vertical eddy diffusion of momentum and heat. The behavior of the turbulent eddy mixing parameters was found to be consistent with measurements in the laboratory and atmosphere. In particular, the flow Richardson number () during the transient flow to a strongly stable state can behave nonmonotonically, growing with increasing gradient Richardson number () to the state of saturation at a certain gradient Richardson number ( Ri@ 1 ), which separates two different turbulent regimes: strong mixing and weak mixing.

Numerical Simulation of Coherent Structures in the Turbulent Boundary Layer under Different Stability Conditions

Coherent structures in the turbulent boundary layer were investigated under different stability conditions. Qualitative analyses of the flow field, spatial correlation coefficient field and pre-multiplied wind velocity spectrum showed that the dominant turbulent eddy structure changed from small-scale motions to large- and very-large-scale motions and then to thermal plumes as the stability changed from strong stable to neutral and then to strong unstable. A quantitative analysis of the size characteristics of the three-dimensional turbulent eddy structure based on the spatial correlation coefficient field showed that under near-neutral stability, the streamwise, wall-normal and spanwise extents remained constant at approximately 0.3 δ , 0.1 δ and 0.2 δ ( δ , boundary layer height), respectively, while for other conditions, the extent in each direction varied in a log-linear manner with stability; only the spanwise extent under stable conditions was also independent of stability. The peak wavenumber of the pre-multiplied wind velocity spectrum moves towards small values from stable conditions to neutral condition and then to unstable conditions; thus, for the wind velocity spectrum, another form is needed that takes account the effects of the stability condition.