Determining the height of the planetary boundary layer (PBL) is of crucial importance as it is a key parameter in air-quality modelling and weather forecasting. Continuous remote sensing measurements allow to estimate this parameter based on temperature, humidity, turbulence, or aerosol backscatter profiles. In this study, measurements from radio-sounding (RS), lidars, microwave radiometers (MWR) and wind profilers (WP) were coupled to various detection methods (parcel method (PM), bulk-Richardson number method (bR), surface-based temperature inversion (SBI), potential temperature gradients (SBLpt), aerosol scattering ratio (ASR) and signal-to-noise (SNR) ratio gradients) for day-time and night-time detection of the PBL height. An inter-comparison of the results from each set of instrument and method, for a period of 5 years (2016-2020), was performed taking RS with PM as reference. The Raman lidar (RALMO) and the COSMO model showed very good agreements with RS, while MWR underestimated the PBL height, mostly in summer, probably due to an overheating of the instrument by the sun. This study notably exposed the great perspective of using temperature and humidity profiles retrieved with RALMO to estimate the PBL height. WP showed more scattered, and overall underestimated, results as the measured maximum of turbulence did not always correspond to the PBL height. A 5-year climatology resulted in clear seasonal and diurnal cycles, with maximum height attained in summer, during the day between 12:00 and 14:00, and a minimum in wintertime. Clear and cloudy sky differentiation showed a negative correlation between the PBL height and cloudiness. A decrease of the PBL height in June was observed with all instruments and methods in the three stations of interest, with no clear explanation of the phenomenon. During the night, the bR method has been invalidated due to its tendency to detect layers, almost constantly, just above ground with RS and KENDA. For RALMO and MWR, the use of wind speed measurements from WP in the bR method resulted in a positive bias of the results. This was attributed to a large amount of missing WP data points near ground due to ground clutter and weak nightly turbulences bellow the detection threshold. The growth rate of the convective layer during the day showed similar seasonality, with a maximum in summer and a minimum in autumn when using RALMO and KENDA. An under-estimation of the growth rate was observed using MWR, as a consequence of the underestimated convective boundary layer height. The analysis of a 30-year long-term PBL height trend, using RS, resulted in a small positive trend using PM and negative using bR method, with weak statistical significance. Trends with larger magnitude were observed with shorter data sets from 10 to 20 years, suggesting that stronger variations are observed on the decadal time scale, due to climate oscillations. Finally, the restrictions for each instrument and method, due to weather conditions, vertical resolution and accuracy have been exposed and discussed in this study.