LIDAR-based F-factor for wind shear alerting: different smoothing algorithms and application to departing flights

摘要

Rapid fluctuations of wind shear hazard factor (or F-factor) calculated from the headwind profile measured by Doppler Light Detection and Ranging (LIDAR) may make it difficult for use in wind shear alerting at an airport such as Hong Kong International Airport (HKIA). Preprocessing by flight simulator or post-processing by averaging over a certain length are attempted in this study to smooth the F-factor and enhance its skills on wind shear alerting. The F-factor so obtained is used for alerting of wind shear (headwind change exceeding 15 knots) for departing flights. Through case studies and statistical studies, the LIDAR-based F-factors are found to have good correlation with the F-factor determined from flight data (the Quick Access Recorder, or QAR, data from the commercial jets). Using the relative operating characteristics (ROC) curves, the LIDAR-based F-factor obtained by smoothing over a window size of 300-500m is found to have good skills in wind shear alerting, and the performance is comparable with, or even better than, the LIDAR-based F-factor using the flight simulator approach. By choosing an optimal alerting threshold specific for each runway corridor, the F-factor successfully alerts 86% of pilot wind shear reports with the percentage of time on alert of 12%, which is found to be satisfactory for real-time use. The performance of the F-factor on stronger wind shear (headwind change of 25 knots or more) is also studied.