Characterizing the convective boundary layer turbulence with a High Spectral Resolution Lidar

Abstract

High Spectral Resolution Lidar (HSRL) aerosol backscatter coefficient observations collected in Norman, Oklahoma, USA, between April and July 2012 were analyzed to characterize the turbulent fluctuations in the convective boundary layer (CBL). Seventeen cloud-free 2 h periods, during which the CBL was quasi-stationary in the late afternoon, were selected. Autocovariance techniques were applied to each 2 h period to derive vertical profiles of statistical moments and integral scale. The low random noise level in the HSRL allowed meaningful estimates of the variance, skewness (S), and kurtosis (K) of the aerosol backscatter field to be computed. The S profiles showed a transition from negative to positive just below the top of the CBL (zi); this suggests that there are relatively few narrow plumes of aerosol-laden air being lofted from the CBL into the free troposphere (yielding positive S above zi) and similarly that there are some plumes of cleaner air descending from the free troposphere into the CBL (negative S below zi). Vertical profiles of K indicated that the distribution of aerosols transitioned from leptokurtic to platykurtic just above zi. However, the profile of K is not symmetric around zi; the turbulence is more intermittent above zi and below 0.8 zi, but the turbulent motions are relatively constant between 0.8 zi and zi. The relationship between S and K in the upper portion of the CBL and through the interfacial layer indicated that the distribution of turbulence is not Gaussian (i.e., S is 0 and K is 3) at any height.