Abstract. Weather Research and Forecasting (WRF) single-column model simulations were performed in the late summer of 2012 in order to analyse the diurnal changes of the planetary boundary layer (PBL). Five PBL schemes were tested with the WRF. From the radiometer and wind-profiler measurements at one station, derived PBL heights were also compared to the simulations. The weather conditions during the measurement period proved to be dry; the soil moisture was below wilting point 85 percent of the time. Results show that (1) simulation-based PBL heights are overestimated by about 500–1000 m with respect to the observation-based PBL heights, and (2) PBL height deviations between different observation-based methods (around 700 m in the midday) are comparable with PBL height deviations between different model schemes used in the WRF single-column model. The causes of the deviations are also discussed. It is shown that in the estimation of the PBL height the relevance of the atmospheric profiles could be as important as the relevance of the estimation principles.