Meeting the FAA's Thunderstorm Forecasting Requirements–Centralized vs. Distributed Processing

In a January 2004 speech, Transportation Secretary Norman Mineta related both good and bad news regarding the state of aviation and air travel. The good news is that passengers are filling commercial airliner seats and air traffic is rapidly approaching pre-September 11, 2001 levels. For the bad news, the National Airspace System (NAS) is again beginning to show signs of stress in terms of having the airspace capacity to meet that demand. The Secretary described ongoing infrastructure improvements for aviation (e.g., additional runways and air traffic control facilities/systems, and more weather radars). However, despite these improvements, weather will continue to impact NAS operations—restricting the use of available jet routes and reducing the throughput (approaches/departures) of NAS pacing airports (responsible for 80%-90% of delays.) Currently, thunderstorms have the most significant, widespread impact on NAS operations (and capacity).

Until recently, pilots, dispatchers, traffic managers, and air traffic controllers could only monitor local, regional, or national weather radar depictions of heavy precipitation to infer where a thunderstorm was located. And lightning data, while an excellent indicator of thunderstorm activity, is not always available and depicts only the current location. For NAS capacity to match demand, traffic managers need a forecast of where thunderstorms will be in the next several hours. In addition, thunderstorm forecasts must be reliably accurate and convey predicted movement (speed, direction, and intensity). Moreover, what non-meteorologist NAS users need is the impact of thunderstorms on operations, including when jet routes will be constrained (become unavailable), what will be the altitude of thunderstorm tops (for over-the-top routing), when will the thunderstorm (and gust front) approach/depart the airport (to optimize runway usage due wind shifts and maximize the number of aircraft that can be handled during thunderstorm passage).

Studies indicate approximately 40% of the impacts from line thunderstorms are avoidable if accurate forecasts are available for routing/re-routing purposes as well as for terminal applications. To that end, FAA weather research has funded several initiatives to develop a thunderstorm forecast capability for both terminal and en route airspace. These include the Terminal Convective Weather Forecast (TCWF) to be implemented in the Integrated Terminal Weather System to enhance capacity at NAS pacing airports, the Corridor Integrated Weather System (CIWS) to ameliorate the impact of thunderstorms in the busiest corridor (extending eastward from Chicago to Indianapolis to Washington DC to Boston) where parallel jet routes are very close to each other and susceptible to migrating thunderstorms, plus efforts to develop the National Convective Weather Forecast product.

Given the austere fiscal environment in which the FAA will be operating for the next few years, implementing three separate initiatives to provide a thunderstorm forecast capability is being carefully scrutinized by the agency. As a result, the FAA is examining alternatives to fulfill this capability [for both terminal and en route applications] and is considering a single (centralized) processing capability, vice a separate (distributed) processing capability for each domain. Details of the thunderstorm forecasting requirements for both terminal and en route domains will be examined, as well as the various alternative solutions. Further cost savings are likely with the centralized processing approach, especially when coupled with a crucial element of the FAA’s weather architecture—the Weather And Radar Processor—to disseminate products to terminal and en route NAS user displays.