This study proposes a method for analyzing the convective boundary layer (CBL) structure using Haar discrete wavelet  transform  (DWT)  and  the  covariance  relationship  between  the  potential  temperature  ()  and  water  vapor  mixing ratio (r). In this study, the Haar DWT was selected to detect the position of rapid change in the vertical profiles of  and r in the CBL. The approximation and detail coefficients at various scales were calculated for each variable using the Haar DWT.  Among  the  various  scales,  this  study  utilized  the  smallest  scale,  specifically  the  scale  with  a  segment  length  of approximately  1.5 m,  which  was  smaller  than  the  segment  length,  including  the  top  of  the  surface  layer  and  the  mixed layer. The typical covariance of  and r in the CBL had positive values in the surface layer, zero in the mixed layer, and negative values in the entrainment zone. Based on the covariance structure in the CBL, the top of the surface layer and mixed layer were estimated for 32 sunny days, and by comparing the estimated heights with the vertical profiles of  and r, a method using Haar DWT and in -r covariance was proposed.