Weather and Radar Processor (WARP) Optimal Mosaic Evaluation

The Weather and Radar Processor (WARP) supports the en-route environment of Air Traffic (AT). It is currently deployed at all 21 Air Route Traffic Control Centers (ARTCCs) and the Air Traffic Control System Command Center (ATCSCC). Utilizing a direct port to every Next Generation Radar (NEXRAD), WARP generates radar mosaic images from raw radar products and provides these mosaics to Center Weather Service Unit (CWSU) and Air Traffic (AT) users. The initial release of WARP (Release 7.4) used a process called “Highest Contributor” to determine which radar data would be incorporated into the mosaic when more then one radar was available. This method brought with it substantial non-weather echoes.

Release 8.0 was a major milestone in the lifecycle of WARP, providing for the first time the capability to disseminate radar mosaics from WARP to the Display System Replacement (DSR) for use by air traffic controllers. This build was also the first attempt to implement the mosaic process called “Optimal Contributor” where supporting information from overlapping radar coverage was used to determine the existence/nonexistence of weather. Unfortunately, testing revealed that the Optimal Contributor process also had a tendency to reduce displayed precipitation intensities when terrain blockage was present. This reduction in displayed precipitation levels could possibly mislead users, and in turn lead to faulty decisions on the part of controllers and pilots.

Engineering Change Proposal (ECP) 20 was funded to remedy the issues identified in the previous release. WARP now generates terrain maps for all radars and implements a new Optimal Mosaic (OM) model. In all, Optimal Mosaic generation takes into consideration terrain elevation, radar coverage quality and supporting data from neighboring bins/radar to determine radar contribution for a specific bin. WARP Release 8.6 was the official release to incorporate the changes of ECP 20. Because of the recent identified shortcomings of the Optimal Mosaic model, it was tasked to conduct a formal evaluation of this model in producing the DSR mosaic product.

Evaluation and assessment of the Optimal Algorithm for the DSR_CROP mosaic product was a very complex issue that required a quantitative, measurable and repeatable success criterion. Unfortunately, due to the lack of a baseline “truth” (i.e., do not know for sure what is real weather and what is non-weather), the evaluation technique was subjective in nature. In an attempt to include a quantitative quality to the process, a three step approach for Optimal Algorithm Evaluation was defined that included quantitative scoring and qualitative assessment activities. The three steps were run in sequence and were as follows: • Step 1: Meteorological (MET) Quantitative • Step 2: MET Qualitative • Step 3: AT Qualitative

Additionally, as a final activity, a Keysite Demonstration was conducted to aid in the operational deployment decision.