Some Observations of Waves at the Hong Kong International Airport and Their Numerical Simulations

Abstract

Because of terrain disruption of the airflow and interface between different airmasses, wave motions may be observed in the vicinity of the Hong Kong International Airport using plan position indicator scans and range height indicator scans with Doppler light detection and ranging systems. This paper documents three cases of wave motion that are not commonly observed near this airport and have never been described before in the literature for Hong Kong, including one mountain wave case and two cases of interfaces between airmasses. These waves may have impacts on aviation safety by leading to the occurrence of low-level windshear and turbulence. They are studied further using a high-resolution numerical weather prediction model. It is found that the model is capable of capturing some features of the waves, such as their occurrence mechanisms. However, some details of the waves are not successfully reproduced, such as the changes in the number of wave crests/troughs with time. Further study should also be conducted to reproduce the wavelengths of these waves.

1. Introduction

Wavy motions in the atmospheric boundary layer have been observed at the Hong Kong International Airport (HKIA), mostly using Doppler light detection and ranging (LIDAR) systems. They may arise from terrain disruption of the prevailing airflow and may be related to the occurrence of low-level windshear at HKIA. Mountain waves have also been observed in tropical cyclone situations. A review of the mountain waves at HKIA can be found in [1].

In the present paper, some novel waves are reported in the vicinity of HKIA and simulations are conducted as far as possible. The waves could be (1) associated with a mountain, but with the wavelength and the number of wave crests/troughs changing rather rapidly with time; (2) occurring in the interface between west to northwesterly winds near the surface and southeasterly winds above; and (3) arising from the undercutting of northerly winds by west to southwesterly winds near the surface. Such waves are very rare and have never been reported at HKIA before, indicating the novelty of the present study. The possibility of observing them recently at HKIA is related to the installation of long-range LIDARs with measurement ranges reaching around 15 km from the equipment, as well as the frequent updating of the range height indicator (RHI) scans of short-range LIDARs (which have measurement ranges in the order of 3 to 6 km). With the installation of more Doppler LIDARs at HKIA, new observations of waves never discovered before at the airport may continue to appear.

Atmospheric waves in the boundary layer may be related to terrain-disrupted airflow. Early work includes [2]. There are also other studies on mountain-disturbed airflow, such as in the Rocky Mountains [3] or the Alps [4]. More recently, such waves have been extensively studied using numerical weather prediction models. Some examples could be found in [5,6,7]. A comparative study on Doppler wind LIDAR measurements and aircraft observations on the mountain waves across the Scandinavian Alps is presented in [8]. Huang et al. [9] also performed observational studies of LIDAR observations in relation to mountain waves and investigated the spectral properties. Another major mechanism is the interface between the airmasses, such as the sea breeze front against the background wind. A review can be found in [10]. Some specific examples include [11,12]. In the vicinity of the Pearl River Estuary near Hong Kong, Liu et al. [13] also conducted an analysis of sea breeze fronts using LIDAR. The waves considered in the present paper include both mountain waves and sea breezes.

2. Meteorological Equipment at HKIA

The main tools used to observe the waves at HKIA are LIDARs. They include (1) long-range LIDAR, with a measurement range in the order of 10 to 16 km and range gate in the order of 100 m—this is used to perform plan position indicator (PPI) scans to be updated every couple of minutes, as well as glide path scans for windshear detection, and (2) short-range LIDAR, with a measurement range in the order of 3 to 6 km and a range gate in the order of 10 to 30 m, with RHI scans to be updated every minute or so.

There are three parallel runways at HKIA, and each runway is equipped with at least one long-range LIDAR. The main one used in the present study is located to the south of the north runway, which provides the best coverage at HKIA. A couple of short-range LIDARs are also considered in this paper: one at Siu Ho Wan (SHW) and another one located near the western end of the north runway (R3W).

The LIDARs use eye-safe laser beams with a wavelength of 1.5 microns. The aerosols in the air are tracked to retrieve the Doppler velocity. The LIDARs generally work well in non-rainy weather conditions. However, if there are liquid clouds, the laser beam will be greatly attenuated so that the measurement range would be significantly limited.

Apart from the Doppler velocity, LIDAR data can be used to calculate the turbulence intensity in terms of the eddy dissipation rate (EDR), which is an internationally recognized metric of turbulence for aviation applications. A review of LIDAR-based EDR calculations can be found in [14].

HKIA is also equipped with a network of ground-based anemometers, along the runways, on the outlying islands, and on the mountain tops. Over the seas, weather buoys are set up to measure the wind and other meteorological parameters as well.

3. Results of Numerical Modeling

The setup of the numerical model in the present study is the same as the one in [15]. The Regional Atmospheric Modeling System (RAMS) version 6.3 is used (https://vandenheever.atmos.colostate.edu/vdhpage/rams.php (accessed on 25 June 2025)). It is nested with European Centre for Medium-Range Weather Forecast (ECMWF) reanalysis data [16] with a spatial resolution of around 0.125 degrees. Five nests are created where the spatial resolution of the simulation is increased by a factor of five, namely 25 km, 5 km, 1 km, 200 m, and 40 m. The coarse domain has a timestep of 10 s, with a timestep ratio of 10. The first model level is about 30 m above the sea surface. There is then a stretching ratio of 1.08 (i.e., the ratio of the model level heights between the adjacent model levels). The simulation is essentially a large eddy simulation (LES) of terrain-disrupted airflow at HKIA.

Of particular importance is the choice of turbulence parameterization scheme. The Smagorinsky scheme [17] is employed for the first two nests and the Deardorff scheme [18] for the remaining three nests. The sensitivity of the simulation results to the choice of turbulence parameterization scheme will be considered in a follow-up study. The existing combination is found to produce reasonable results, e.g., as in [19]. The choice of other parameterization schemes, such as long-wave radiation and short-wave radiation, is the same as in [19]. The land use and vegetation data are obtained from the RAMS website.

Case 1: 20 February 2025

During the whole day of 20 February 2025, a ridge of high pressure persisted over the southeastern coast of China, bringing easterly winds to Hong Kong. The isobaric chart in the synoptic scale can be found in Figure 1a, and the surface observations in the region of HKIA can be found in Figure 1b. Easterly winds prevailed over the whole airport, and a mountain wake (circled in red) could be found to the west of Lantau Island, where westerly winds were recorded. On the mountain top, moderate south to southeasterly winds prevailed.

Some snapshots of a 6-degree PPI scan of the north-runway LIDAR can be found in Figure 2. From 03 UTC (Hong Kong time = UTC + 8 h) to 10 UTC, mountain waves could be found to the east of HKIA. Location-wise, this wave was more or less stationary. However, the number of wave crests/troughs and the wavelength changed with time. At first (Figure 2a), about four wave crests/troughs could be found, and the wavelength was about 2 km. However, as time progressed (Figure 2b–d), the wavelength became shorter, reaching about 1 km only. The number of crests/troughs increased to 5, as shown in Figure 2e, while some wave crests/troughs were missing (see Figure 2f) downstream of the mountain (area circled in yellow) due to cloud formation and the attenuation of the laser beam in liquid clouds.

When the wave crosses a runway corridor, there may be sudden change in the headwind, leading to low-level windshear encountered by the aircraft. One example is shown in Figure 3a, in which a departure to the east for the north runway (07LD runway corridor) has a headwind drop of around 14 knots due to passage of the wave over this corridor. The wave pattern is not only shown in the PPI scans of the long-range LIDAR. It also appears in the RHI scan of the short-range LIDAR at Siu Ho Wan (SHW), southeast of HKIA, scanning towards the northwest direction (Figure 3b). Two wave crests could be found in the RHI scan, and the wavelength was estimated to be around 2 to 2.5 km, which was consistent with the PPI scan results from the long-range LIDAR.

The changes in the number of wave crests/troughs and the wavelength may be related to changes in the stability of the atmospheric boundary layer as the vertical wind profile and the terrain do not change with time. An attempt was made to simulate the change in the wave pattern using RAMS. However, in the simulation results, just three wave crests/troughs could be analyzed (Figure 4a), and the pattern did not change with time. The simulation result could reproduce the change in headwind along the runway corridor 07LD (Figure 4b), with a headwind drop of around 13 knots. The results are comparable with the actual observations (Figure 3a). The RHI scan of the SHW LIDAR also shows two wave crests (Figure 4c), which is consistent with the actual observations (Figure 3b). The simulated wind field (Figure 4d) illustrated the domination of easterlies over the area east of HKIA (i.e., the area where waves were observed). Fluctuations in the wind speed could be observed within the easterlies, which explained the existence of the waves. This could be attributed to the disrupted mountain flow to the northeast or southeast or simply the wave features in relation to the difference in friction between sea and land (reclaimed airport area).

Case 2: 5 March 2025

On the synoptic scale, a northeastern monsoon prevailed over Southern China (Figure 5a). In the region of HKIA, an easterly wind prevailed over the majority of the airport, whereas northerly winds affected the northwestern corner of the airport, as well as the sea areas to the west (Figure 5b).

The convergence of the two airstreams is clearly visible in the PPI scan of the LIDAR (Figure 6a), although the wave structure is not observed in such conical scans. The RHI scan of a short-range LIDAR at the western end of the north runway (R3W) is considered in this study. It scans to the north–northwest at an RHI plane perpendicular to the runway orientation. From the RHI scan of the Doppler velocity (Figure 6b), four wave crests/troughs could be observed, at the interface between the inbound northerly flow and the outbound southeasterly flow. The cooler continental air undercut the warmer air above, giving rise to the interface between the two airmasses.

Through a simulation using RAMS, the mesoscale/microscale flow is marginally correct, with an inbound flow over the western and northwestern parts of HKIA (Figure 7a). However, the northerly winds over the Pearl River Estuary are not as extensive as in the actual observation (Figure 6a). Even so, the interface between the two airmasses could be marginally seen in the simulated RHI Doppler velocity image of the R3W short-range LIDAR (Figure 7b). Compared to the actual observations (Figure 6b), the height of the interface between the two airmasses is correct in the simulation result, although wave motions are only marginally seen.

Case 3: 20 March 2025

Synoptically, the coast of Southern China only has sparsely separated surface isobars, so that the winds are rather light (Figure 8a). From the surface observations at the airport (Figure 8b), a background wind of light northerlies prevails over the area to the north and to the west of Hong Kong. Over HKIA, there is the setting in of a sea breeze during the day and there is the prevalence of moderate south to southwesterly winds.

In both the 3.1-degree and 6-degree PPI scans of the north-runway LIDAR (Figure 9a,b, respectively), there are waves in the southwesterly flow at the low level. Altogether, there are four wave crests/troughs, and their separation is about 2 km. They cover the whole airport and the adjacent waters. In the EDR map based on the 3.1-degree PPI scan (Figure 10a), there are horizontal “bars” of moderate turbulence (colored green) against the background light turbulence (colored blue) over the eastern part of the airport, and their locations are consistent with the waves. From the RHI scan at the R3W short-range LIDAR (Figure 10b), we can see that there are waves at the interface between the cooler sea breeze (southwesterly winds, away from the LIDAR, in red) and the background light northerly above (towards the LIDAR, in blue). It is believed that the waves shown in the PPI scans of the long-range LIDAR are related to the interface of the two airmasses.

In order to confirm the above, a high-resolution numerical simulation was performed. The simulated 3.1-degree PPI scan of the long-range LIDAR can be found in Figure 11a. Three wave crests/troughs could be found near the surface. They are associated with west–southwest- to east–northeast-oriented bands of upward/downward motion, as shown in the vertical wind plot (Figure 11b). The simulated surface wind pattern can be found in Figure 11c. It can be seen that the wave may have been triggered at the surface convergence between the light background northerly and the southwesterly sea breeze, which is consistent with the surface observations (Figure 8b). At this convergence zone, significant upward motion (brown, around 3 m/s, circled in blue in Figure 11b) is also simulated. Moreover, this wave structure gradually decays at distances further downstream (more southwards) from this region of surface convergence. This echoes the study by Banakh and Smalikho [20], which shows that, when internal gravity waves appear, the amplitude of the fluctuations in the vertical wind velocity increases, sometimes by several orders of magnitude.

A simulated RHI scan of the R3W LIDAR is shown in Figure 12. Two wave crests can be found, and the pattern is generally consistent with the actual observations in Figure 10b. The height of the interface between the outbound and inbound velocities is also similar between the simulation and the actual observations.

4. Conclusions

The present paper describes three wave motion cases at HKIA as revealed by the PPI scans of the long-range LIDAR and RHI scans of two short-range LIDARs. The waves occurred as a result of terrain disruption or at the interface between two airmasses. Observations of such waves have only become possible recently due to the extended measurement range of the long-range LIDAR and the presence of a dense network of short-range LIDARs in the region of HKIA. Such waves may be hazardous for aviation safety, leading to low-level windshear and turbulence.

The wave motions were studied in detail using a high-resolution numerical weather prediction model. Some aspects of the wave motions could be successfully reproduced, such as mountain disruption and the occurrence of an interface among airmasses. However, not all features were well simulated, e.g., the number of wave crests/troughs could be different from the actual observations. The wavelength is another area that requires further investigation. Further studies are being conducted for a comparison with the theoretical results to determine whether the wavelengths can be successfully predicted.

The documentation of these wave cases could be useful for the timely observation and alerting of similar waves at other airports. This work may offer a useful reference for the measures to ensure aviation safety, particularly under low-level windshear and turbulence in association with these waves.