It is well established that turbulence in the stable planetary boundary layer (PBL) is neither steady nor isotropic and often occurs where conventional PBL theory predicts no turbulence. These outbreaks of turbulence in the PBL have yet to be taken into account in forecasting models either for weather or air quality. As a result, the models miss the enhanced vertical mixing of heat, moisture, and air pollutants by these intermittent turbulence episodes. This can lead to inaccurate forecasts of minimum surface temperature, fog, and pollutant concentration. A long-standing ARL research program is now at a stage where descriptions of intermittent nocturnal turbulence are ready for consideration by the modeling community. In other words, transfer to operations is now feasible.

A major source of this turbulence is the breaking of gravity waves launched by small-scale topographic features. Although the amplitudes of these waves might be small compared to similar waves downwind of mountain ranges, the effects of these smaller waves on the PBL are no less important. Ridges and hills as small as 1000 m (horizontally) and 100 m (vertically) can cause gravity waves. These surface features are sufficiently small that they cannot be considered explicitly in typical mesoscale models instead, their effects must be parameterized.

The importance of gravity wave-induced turbulence in the stable PBL was recognized by ARL scientists some time ago, and a program to parameterize this turbulence was then initiated. Collaborators in this research included scientists from Sweden, Germany, Australia, and most recently South Korea. A paper recently submitted to Atmospheric Environment presents a detailed analysis of the effects on predictions of PBL flow and stability when this parameterization scheme is incorporated in mesoscale models. The validity of the scientific underpinning of the scheme has been demonstrated, however work remains to be done to make the schemes suitable for operational use. For this reason, the next stage of the research will be to engage NCEP scientists in a collaborative effort to explore (a) the practical need to operationalize the schemes now developed and (b) the methodology for their implementation in current forecasting models.