Novel Gust Front Detection with Polarimetric WSR-88D Using Neuro-fuzzy System

Gust front is the leading edge of the cold outflow of a convective storm, which is often accompanied with extreme changes of wind speed and direction, pressure and temperature. Typically, the gust front exhibits as a thin line of arc shape. The strong wind and shear along with the gust front can endanger air terminal operations especially during landing and takeoffs. Previous gust front detection algorithm such as the Lincoln Laboratory's Machine Intelligent Gust Front Algorithm (MIGFA), searching for the line features in Doppler products, has been working reasonably well with the single-polarized terminal Doppler weather radar. The upgrade of the national network of the Weather Surveillance Radar - 1988 Doppler (WSR-88D) to dual-polarization is underway, which provides additional features for gust front detection. In this work, a novel Neuro-fuzzy gust front detection (NFGFD) was developed based on both Doppler and polarimetric measurements from the upgraded WSR-88D. NFGFD is a fuzzy system that is optimized by a training procedure with neural network. Statistical analysis of the inputs to the algorithm as well as the architecture of the algorithm will be discussed. The algorithm was tested and demonstrated using 5 cases of polarimetric WSR-88D data downloaded from the National Climate Data Center (NCDC). Preliminary results have shown that gust fronts with different orientations can be detected reliably using NFGFD.