Observation and Modelling of Atmospheric Boundary Layer Wind and Turbulence Structures of Hazardous Windstorms

Windstorms are the most catastrophic of natural hazards and are responsible for over 70% of the total insured losses worldwide. Due to climate variability, storm-related hazards could become more frequent and severe in the future. Therefore, it is of great socio-economic significance to gain a comprehensive knowledge of the wind and turbulence structures of windstorms in the atmospheric boundary layer (ABL) to facilitate wind-resistant structural design and disaster prevention. The ABL structure of synoptic winds (e.g., extratropical cyclones and monsoons) has been reasonably understood and modelled since the 1960s, and these models are still used by wind and structural design codes and standards. Nevertheless, no well-established models are available for non-synoptic windstorms, such as tropical cyclones (TCs, or typhoons/hurricanes), thunderstorms, and tornadoes, due to their complex kinematic and thermodynamic structures and limited observational records. To provide a full range of meteorological observations, the Hong Kong Observatory (HKO) has been operating across the territory a dense monitoring network, which incorporates Global Positioning System (GPS) dropsondes, radiosonde balloons, Light Detection and Ranging (LiDAR) systems, Sonic Detection and Ranging (SoDAR) systems, Doppler radar wind profilers, microwave radiometers, Radio Acoustic Sounding System (RASS), Doppler weather radars, and meteorological stations, buoys, and airplanes equipped with various meteorological sensors. In addition, the 356-m-high Shenzhen Meteorological Gradient Tower (SZMGT) and the weather station network in Shenzhen will provide supplementary measurements. The records obtained from these sources will afford an exceptional opportunity to shed new light on the wind and turbulence structures of hazardous windstorms. The proposed project aims to comprehensively investigate the wind and turbulence structures of non-synoptic hazardous windstorms, including TCs, thunderstorms, and tornadoes, and to establish models to describe the wind and turbulence profiles of these storms in the ABL. This study will comprise four closely related tasks: (1) Observational study of wind and turbulence structures of tropical cyclones; (2) Machine learning-based detection and wind characteristics investigation of thunderstorms; (3) Remote sensing-based identification and 3D wind field analysis of tornadoes; (4) Comparison and artificial intelligence-aided modelling of wind and turbulence profiles in TCs, thunderstorms, and tornadoes. To address task (1), a database will be established based on a huge variety of observations from HKO and SZMGT. The mean wind and turbulence structures of TCs, e.g., boundary layer height scales, low-level jets, turbulence intensity profiles, and power spectral density will be comprehensively investigated. To complete task (2), a machine learning-based thunderstorm detection and classification system will be developed. Spatiotemporal evolution of nonstationary wind and turbulence characteristics in thunderstorms will be analyzed. To handle task (3), tornado wind fields will be reconstructed based on remote-sensing records from Doppler weather radars and 3D scanning LiDARs. The 3D distributions of wind speeds and turbulence intensity in the tornado vortices will be thoroughly studied. To tackle task (4), wind structures of synoptic winds and non-synoptic windstorms will be systematically compared. Wind and turbulence profile models for TCs, thunderstorms, and tornadoes will be proposed with the aid of artificial intelligence techniques. The proposed project will enhance our knowledge of wind and turbulence structures of TCs, thunderstorms, and tornadoes, offer useful wind and turbulence profile models for wind-resistant design of structures in Hong Kong and other regions prone to these windstorms, and provide valuable information for improvement of wind and structural design codes and standards.