The correct depiction of the atmospheric boundary layer (ABL) in numerical models continues to be inaccurate in certain weather patterns. Since the ABL is the lowest part of the atmosphere, problems in its depiction have a direct impact not only on temperature, humidity and wind near the ground, but also on the vertical stability of the whole atmosphere, thus affecting the development of clouds and precipitation.

Numerical simulation of the atmosphere has grown tremendously over the last few decades. Nevertheless, simulation problems remain, affecting the accuracy and usefulness of the simulations. Simulation errors have recently been documented in which the differences between the real atmosphere and the simulation were found entirely within the ABL. Since the ABL is not explicitly modeled in most models, we need to look closely at the ABL parameterizations as a likely source of these errors.

We will present output from the MM5 mesoscale model concentrating on the differences made by using the five different boundary layer parameterization schemes. Most of the cases were run at the University of Washington using a doubly nested, 12km grid centered on the Pacific Northwest. The development of the differences in the boundary layer structure and the extent of these differences will be shown.