Turbulent structure derived from meteorological disturbance with low-frequency fluctuation affects different characteristics for the distribution of wind pressure and the peak value of concentration in comparison with the inflow condition using turbulent boundary layer (TBL) over smooth or rough surface in usual cases. However, it is impossible to resolve turbulent field around buildings explicitly and difficult to reproduce high frequency component by only the mesoscale meteorological model. Thus, several researches show the connection method between mesoscale meteorological model with LES. However, when connecting mesoscale meteorological model with LES for turbulent field around buildings, tasks still remain in making inflow condition by adding the appropriate scale of fluctuation with high frequency to meteorological model results. Therefore, this study presents a method to add fluctuation components which are extracted by spatial filtering and rescaling technique to a velocity field of mesoscale meteorological model. In the presented method, the fluctuation with high frequency which is generated in the driver region is decomposed physically to the scale of original inflow and residual fluctuation. Then, the residual fluctuation is rescaled and imposed to inlet plane at an appropriate scale of fluctuation.
Then, the proposed method is validated by a priori test applying the presented method to the filtered TBL. This study compares the calculation result and the original TBL before a filter operation. As a result of the comparison, it is confirmed that high-frequency components of fluctuation component are added to the filtered TBL at appropriate scale of fluctuation. Also, the turbulent intensity and power spectrum density which are lost in the filtering process are recovered by the presented method.
Next, this study carries out WRF-LES for idealized atmospheric boundary layer and compares results between WRF-LES and TBL by Lund et.al. In the results of WRF-LES, mixing of vertical direction is strong and the fluctuation of velocity remains until the height of 800m. The comparison shows that the power spectrum density of WRF-LES in low-frequency component which decays in TBL by Lund et.al corresponds to that of the Karman type spectrum without decaying. Also, the turbulent intensity near the ground is enough and it corresponds that of AIJ recommendation. This is because a periodic condition is employed in the WRF-LES in this study. However, the comparison of power spectrum density reveals that high-frequency component decays in the results of WRF-LES even in the case using spatial resolution of 25m.
Finally, the presented method is applied to the WRF-LES results. In the results of the presented method, the turbulent intensity of v’ and w’ is slightly decayed because the vertical profile of u above the height of 100m is almost constant. However, the results show that the power spectrum density of low-frequency region is maintained. Also, the fluctuation of high-frequency which decays in WRF-LES is generated by the presented method. In the case where the presented method is applied, the range of frequency reproducing the fluctuation of high frequency is extended to the range almost same as that of TBL by Lund et. al.
