A real-time numerical weather prediction system is described whose sole purpose is to predict the potential for clear-air, mountain, and convective aviation turbulence. The Real-Time Turbulence Model (RTTM), which is designed to support NASA-Langley B-757 research flights, consists of a data ingest system, mesoscale numerical model, and a postprocessor. The data ingest system involves the NWS ETA analyses, rawinsondes, and surface observations which are synthesized into an initial state as well as time-dependent lateral boundary conditions. The numerical model consists of a contemporary version of the Mesoscale Atmospheric Simulation System (MASS) hydrostatic model. The model is integrated twice-daily for 24 hours of forecast time at three horizontal grid resolutions, i.e., 60 km, 30 km, and 15 km over a matrix of 90 x 90 grid points. Initial conditions and time-dependent lateral boundary conditions for the nested grids are derived from the next coarser grid simulation. The three simulations employ 45 terrain-following vertical layers as well as the Blackadar planetary boundary layer and Kain-Fritsch convective parameterization schemes. These simulations are continuously generated in real-time on a university workstation and the products displayed on the internet.

The post-processor provides model-simulated turbulence products at all three scales of motion over which the model is integrated and centered primarily on the eastcentral United States. These products are available on a public web site on height surfaces ranging from 18,000 - 46,000 feet in elevation at 2,000 feet intervals every 3 hours on the two coarsest resolution grids and every 90 minutes on the finest resolution grid. The product suite is designed to support forecasts of regions of turbulence potential that are hazards to aviation including clear-air, mountain, and convective turbulence. The centerpiece of the product suite is an original turbulence potential index based on the concept that vortex generation by strong solenoidal tendencies in the vertical vorticity equation is a reliable predictor of fine-scale regions of intense turbulence independent of turbulence category. Preferred turbulence regions occur where along-stream pressure perturbations become orthogonal to regions of deformation-induced frontogenesis producing index maxima. Additionally the product suite includes: three other predictive indices, wind isotachs and barbs, flow streamlines, Richardson numbers, turbulence kinetic energy, convective precipitation, total precipitation, cloud mass fluxes, cloud top elevation, two convective stability indices, high terrain Froude number profiles, and numerous thermodynamic soundings with detailed stability/wind shear products.

The presentation will include examples of turbulence potential index forecasts that were available 6-12 hours in advance that accurately discriminated among quiescent flow and fine-scale regions of moderate-severe clear-air, mountain, and convective turbulence and were compared to observed pireps, airmets, and sigmets.