The threat of severe weather (e.g. thunderstorms and snowstorms) is a major cause of delays and financial loss to the aviation industry. The Research Applications Program (RAP) at the
National Center for Atmospheric Research is being partially funded by the FAA to develop algorithms that provide real-time, automated nowcasts of specific aviation weather hazards. Accurately detecting and forecasting where these severe weather patterns will be in the future can enhance aviation weather forecasts in both the terminal and enroute environments. Currently, most automated algorithms that predict storm locations use either centroid or cross-correlation area tracking techniques on radar reflectivity data. Centroid (or cell) tracking can be very effective when storms have good cellular structure, such as thunderstorms. However, problems may be encountered if the cells merge or split. Problems also occur when individual cells embedded within larger scale features (e.g. squall lines) have different motion vectors than the larger scale systems. Additionally, snowstorms and stratiform rain can't be tracked with the centroid technique because they don't exhibit the required cellular structure. Consequently, cross-correlation area tracking techniques are being implemented into winter and summer weather-related forecasting algorithms at RAP.
The Weather Support to De-Icing Decision Making (WSDDM) System is one integrated display system developed at RAP that utilizes the correlation tracking technique. The WSDDM system provides real-time nowcasts of snowfall rate and was developed specifically for airport ground de-icing personnel. The information from WSDDM allows decision makers to anticipate the onset and termination of snowfall at the airport and surrounding regions. The Automated Thunderstorm Nowcasting System (referred to as the Auto-nowcaster) is another integrated display system developed at RAP. The Auto-nowcaster provides short-term, time and space-specific forecasts of thunderstorm initiation, growth, movement and dissipation. A centroid tracker is used to identify and track existing thunderstorms for forecast periods < 60 minutes. For forecast periods beyond 60 minutes, a cross-correlation area tracker is being used to track the larger scale systems. Using both a centroid and cross-correlation tracker in the Auto-nowcaster will make it easy to compare the two techniques.
This paper will briefly review the pros and cons of the two tracking techniques and a detailed discussion on the cross-correlation area tracking techniques implemented in the RAP algorithms will be provided. Reflectivity image forecasts from the cross-correlation area tracker will also be presented. This will be the first time that image forecasts (opposed to forecast contours) were displayed on the Auto-nowcast system
The 8th Conference on Aviation, Range, and Aerospace Meteorology