A new objective method for the determination of the atmospheric boundary layer (ABL) depth using routine vertically pointing aerosol lidar measurements is presented. A geophysical process-based analysis is introduced to improve the attribution of the lidar-derived aerosol gradients, which is so far the most challenging part in any gradient-based technique. Using micrometeorological measurements of Obukhov length scale, both early morning and evening transition periods are determined which help separate the turbulence regimes during well-mixed convective ABL and nocturnal/stable ABL. The lidar-derived aerosol backscatter signal intensity is used to determine the hourly-averaged vertical profiles of variance of the fluctuations of particle backscatter signal providing the location of maximum turbulent mixing within the ABL; thus, obtained mean ABL depth guides the attribution by searching for the appropriate minimum of the gradients. An empirical classification of the ABL stratification patterns into three different types is proposed by determining the changes in the near-surface stability scenarios. First results using the lidar observations obtained between March and July in 2011 at SIRTA atmospheric observatory near Palaiseau (Paris suburb) in France demonstrate that the new attribution technique makes the lidar estimations of ABL depth more physically reliable under a wide spectrum of meteorological conditions. While comparing lidar and nearby radiosonde measurements of ABL depths, an excellent concordance was found with a correlation coefficient of 0.968 and 0.927 for daytime and nighttime measurements, respectively. A brief climatology of the characteristics of the ABL depth, its diurnal cycle, a detailed discussion of the morning and evening transitions are presented.