Case Studies of CAT: Evaluation and Verification of New Forecasting Techniques

Clear-air turbulence (CAT) is an important and unsolved problem within the aviation industry and the atmospheric sciences. In this presentation we examine CAT using three case studies of turbulence outbreaks in cyclonic upper-level flow in December 2010, January 2011, and September 2011. High-resolution model output was utilized to create turbulence forecasts using six turbulence indices, including the new Ellrod-Knox index and the unique Lighthill-Ford algorithm. These forecasts were compared to actual eddy dissipation rate (EDR) turbulence reports. Verification statistics and ROC (relative operating characteristic) curves were produced to determine which forecast metric had the most skill during these outbreaks. In terms of the area under the ROC curve, the Ellrod-Knox method performed the best in the December and January cases, for all turbulence intensities and moderate-or-greater turbulence. For the September case, the Lighthill-Ford method performed the best for all turbulence intensities. The implications of these results for operational CAT forecasting will be discussed.