Application of spectral decomposition of LIDAR-based headwind profiles in windshear detection at the Hong Kong International Airport

Abstract

In aviation, rapidly fluctuating headwind/tailwind may lead to high horizontal windshear, posing potential safety hazards to aircraft. So far, windshear alerts are issued by considering directly the headwind differences measured along the aircraft flight path (e.g. based on Doppler velocities from remote-sensing). In this paper, we propose and demonstrate a new methodology for windshear alerting with the technique of spectral decomposition. Through Fourier transformation of the LIDAR-based headwind profiles in 2012 and 2014 at arrival corridors 07LA and 25RA of the Hong Kong International Airport (HKIA), we study the occurrence of windshear in the spectral domain. Using a threshold-based approach, we investigate performance of single and multiple channel detection algorithms and validate the results against pilot reports. With the receiver operating characteristic (ROC) diagram, we successfully demonstrate feasibility of this approach to alert windshear by showing a comparable performance of the triple channel detection algorithm and a consistent hit rate gain (07LA in particular) of 4.5 to 8 % in quadruple channel detection against GLYGA, which is the currently operational algorithm in HKIA. We also observe that some length scales are particularly sensitive to windshear events which may be closely related to the local geography of HKIA. This study serves to open a new door for the methodology of windshear detection in the spectral domain for the aviation community.

Keywords

• aviation meteorology
• windshear
• LIDAR
• spectral decomposition