To investigate the applicability and limitations of both physical (wind tunnel test) and numerical (weather prediction) simulations of the atmospheric boundary layer, field measurements from a Doppler SODAR and a wind-profiler were combined to serve as validation criteria when comparing the results from a numerical simulation conducted by the Weather Research and Forecast model (WRF) and from wind tunnel testing. The comparisons focused on the simulation of the typhoon boundary layer, and revealed that a major drawback of wind tunnel testing is the use of an unrealistic approaching wind profile. As a result, the wind tunnel test results should only be considered valid when the measured wind profile is influenced predominantly by the underlying terrain. Meanwhile, the relatively coarse resolution of the underlying terrain model used in the numerical weather prediction system may lead to an inaccurate mean wind speed profile at lower altitudes, especially when the winds are coming from the land fetch.