The automatic and non-supervised detection of the planetary boundary layer height (<i>z<sub>PBL</sub></i>) by means of lidar measurements was widely investigated during the last years. Despite the considerable advances achieved the experimental detection still present difficulties either because the PBL is stratified (typically, during night-time) either because advected aerosol layers are coupled to the PBL. The coupling uses to produce an overestimation of the <i>z<sub>PBL</sub></i>. To improve the detection even in these complex atmospheric situations, we present a new algorithm, called POLARIS (PBL height estimatiOn based on Lidar depolARISation). POLARIS applies the wavelet covariance transform (WCT) to the range corrected signal and to the perpendicular-to-parallel signal ratio (δ) profiles. Different candidates for <i>z<sub>PBL</sub></i> are chosen and the attribution is done, based on the WCT applied to the RCS and the δ. We use two ChArMEx campaigns with lidar and microwave radiometer (MWR), conducted on 2012 and 2013, for the POLARIS' adjustment and validation. POLARIS improves the <i>z<sub>PBL</sub></i> detection thanks to the consideration of the relative changes in the depolarization capabilities of the aerosol particles in the lower part of the atmospheric column. Taking the advantage of a proper determination of the <i>z<sub>PBL</sub></i> determined by POLARIS and by MWR under Saharan dust events, we compare the POLARIS and MWR <i>z<sub>PBL</sub></i> with the <i>z<sub>PBL</sub></i> provided by the Weather Research and Forecasting (WRF) numerical weather prediction model. WRF underestimates the <i>z<sub>PBL</sub></i> during daytime but agrees with the MWR during night-time. The <i>z<sub>PBL</sub></i> provided by WRF showed a better temporal evolution during daytime than during night-time.