During the Mesoscale Alpine Programme flights with the NCAR Electra were, at times, focussed on the study of strong mountain wave activity over the Alps, with a focus on breaking gravity waves. In one compelling case, which occurred on Sept 20, 1999, early in MAP, clearly showed a failure by a numerical model to accurately predict wave breaking in the lee of the Hohe Tauern Range of the Northern Alps (east of Innsbruck). This case was a false positive result - where wave breaking was predicted by a numerical model using 3.5 km horizontal resolution, yet there was none found in several flights by the Electra. However, strong, persistent lee waves were sampled in multiple flight legs.

It is postulated that in addition to grid spacing as a factor in this failure, the turbulence scheme also played a role. Thus, in this study focus is made on a few sensitivity studies of in the grid spacing/turbulence parameters matrix. Using the CSU - Regional Atmospheric Modeling System, the turbulence parameterization coefficients, type and also grid spacing (in the range 4.0 km to 0.5 km) are systematically altered to learn about the factors influencing more accurate forecasts of gravity wave evolution. Not surprisingly, it is found that smaller grid spacing enables lee wave activity to be represented more closely, as other studies have show. However, it is interesting to note the influence of turbulence parameterization coefficients on accuracy, where, with fixed grid spacing lee waves are obtained with certain settings, but not with others. The implications for numerical prediction of gravity waves is discussed.