A dynamically adaptive model has been developed, based on the well known MM5, to improve resolution of user-determined features in atmospheric modeling. The non- hydrostatic equations are transformed to a general moving coordinate system and the NCSU dynamic solution adaptive grid algorithm DSAGA is installed to selectively resolve chosen properties or features of the atmosphere or topography. The first goal of the research is the accurate prediction of optical and clear air turbulence. Neither of these turbulence scales are resolved adequately by the inner nests in standard mesoscale models with typical nest numbers and grid refinement ratios. Furthermore, standard nesting techniques are limited by the need to allow extra domain extent in each nest to allow nest boundary values to adjust to the finer mesh spacing in the nest. Mesh adaptation by DSAGA and a new NCSU state of the art four equation turbulence model are employed to achieve LES scale simulation of gravity waves and their breakdown into turbulent eddies and direct output of  for use in the prediction of optical turbulence. The adaptive MM5 has been used to predict optical turbulence () for cases in which observations and predictions from radiosonde data are available for comparison. Results show a clear advantage over predictions obtained with the unmodified MM5 when the adaptor is used to increase mesh resolution in the stratosphere. The figure included below illustrates one such case in which adaptation within the lowest nest combined with the new turbulence model recovers much of the  structure found in the binned radiosonde observations. In order to improve further the predictions and to explore other applications, DSAGA and the NCSU turbulence model are being installed into WRF-ARW.

The final paper will include examples of the use of this technique to model gravity waves and turbulence to LES scales and will compare the WRF and MM5 adaptive models.