Weather forecasting is increasingly relying on high-resolution and ensemble numerical weather prediction (NWP) model data. Efforts are underway to develop convective weather products utilizing high-resolution model ensembles for strategic planning of the National Airspace (NAS). The FAA Aviation Weather Research Program (AWRP) is currently funding the development of Consolidated Storm Prediction for Aviation (CoSPA) which is a collaborative (MIT-LL, GSD, NCAR) research and development effort designed to produce rapidly updating 0-8 hr forecasts of summer and winter storms using a combination of nowcasting techniques and NWP. One of the products of CoSPA will be a weather avoidance field. Studies have shown that pilots tend to avoid areas of high storm intensity and echo tops exceeding 35 kFt – this is known as an aviation weather hazard. Echo top height and storm intensity can be computed from the model-predicted vertical distribution of condensed water; however, intensity tends to be underestimated if the model prediction does not include enough graupel or hail. The use of a high-resolution model is critical for resolving the dynamics that drive storm organization, vertical structure and evolution—information required for aviation planning.

In this study, we assess the skill of ensemble, high-resolution model in predicting the joint distribution of echo top height and storm intensity (derived from vertical integrated liquid, VIL) as a function of forecast area for a set of length scales ranging from 50 to 500 km. The model data is obtained from the NOAA/GSD High Resolution Rapid Refresh (HRRR) which was run for CoSPA at 3 km resolution with one hour cycling over a domain spanning ORD, NYC and DCA airports). The joint distributions obtained with the model are compared with observed values obtained from post-processed WSR-88D radar data available from MIT-LL and the National Severe Storms Laboratory (NSSL). Several cases with large aviation impacts in the summer of 2008 are analyzed. The convective weather aviation hazard forecasts generated from the model ensemble are compared with information contained in the Collaborative Convective Forecast Product (CCFP) that is currently used for strategic planning of the NAS. The potential benefits to Air Traffic Management (ATM) of generating convective weather products from model ensembles are discussed.