Multi-Scale Wind Shear at a Plateau Airport: Insights from Lidar and Radiosonde Observations

Abstract

Low-level wind shear poses a significant hazard to aviation, especially at airports located on high plateaus and surrounded by complex terrain. In this study, we present a comprehensive analysis integrating Doppler Lidar and radiosonde measurements collected at the Xining Caojiapu Airport, situated on the northeastern Tibetan Plateau, during June 2022. The results indicate a remarkably high frequency of severe wind shear events (|Δv| ≥ 6 m/s), with an overall occurrence rate of 34% during the observation period. These events are predominantly confined to two distinct atmospheric layers: just above the surface and near the top of the convective boundary layer. The diurnal cycle of wind shear is closely associated with boundary-layer dynamics, exhibiting sharp increases after sunrise and pronounced peaks around midday, coinciding with enhanced turbulent mixing and surface heating. Case analyses further reveal that the most intense shear episodes occur at strong thermal inversions, where momentum decoupling produces thin, critical interfaces conducive to turbulence generation. In contrast, well-mixed convective conditions result in more distributed but persistent shear throughout the lower atmosphere. Diagnostic profiles of atmospheric stratification and dynamic instability, characterized by the Brunt–Väisälä frequency and Richardson number, elucidate the intricate interplay between thermal structure and vertical wind gradients. Collectively, these findings provide a robust quantitative basis for improving wind shear risk assessments and early warning systems at airports in mountainous regions, while offering new insights into the complex interactions between turbulence and atmospheric stratification.

1. Introduction

Low-level wind shear (LLWS), defined as abrupt vertical gradients in wind speed or direction near the surface, is among the most serious hazards to aircraft during takeoff and landing [1,2,3]. These risks are particularly acute at high-elevation airports situated in complex terrain, where interactions between thermal inversions, drainage flows, and orographic force create exceptionally hazardous boundary-layer dynamics [4,5,6]. Xining Caojiapu International Airport, located at approximately 2400 m on the northeastern edge of the Tibetan Plateau and surrounded by mountains exceeding 2500 m, exemplifies such challenging environments. The basin topography of this site promotes nocturnal drainage jets and persistent inversion layers, often leading to near-surface vertical shear and turbulence that jeopardize flight safety.

While LLWS has been extensively investigated at low-lying and coastal airports, its dynamics in plateau regions remain poorly understood. Ground-based Doppler Lidars deployed at airports such as Beijing and Lanzhou have effectively detected low-level shear events and demonstrated the instrument’s capability for continuous, high-resolution monitoring [7,8,9,10,11,12,13,14,15,16]. However, the complex interplay of stratification and dynamic force in plateau terrains remains largely unexplored, despite the demonstrated effectiveness of Lidar in diagnosing shear–stability relationships using metrics like the gradient Richardson number (Ri). Specifically, Ri values below 0.25 indicate shear-driven mixing, whereas values above 1 signify suppressed turbulence, even in the presence of strong shear.

In this study, we integrate ground-based Doppler Lidar and radiosonde observations collected at Xining in June 2022 to advance the quantitative understanding of LLWS in plateau environments. The Lidar system provided high-temporal-resolution wind profiles, while the radiosondes offered concurrent thermodynamic and kinematic measurements. Stratification (characterized by the Brunt–Väisälä frequency) and vertical shear were calculated using centered finite differences to identify turbulent regimes based on Ri. The main objectives are as follows: (i) to resolve the vertical and temporal structure of shear layers in complex terrain; (ii) to quantify the interplay between shear production and the buoyancy suppression of turbulence; and (iii) to assess the implications for operational LLWS detection and mitigation. By presenting detailed, Lidar-derived shear–stability profiles for a high-altitude, mountainous airport, this study addresses a significant gap in boundary-layer meteorology. Our findings contribute to the theoretical understanding of turbulence generation in plateau environments and have important implications for improving aviation safety at airports challenged by complex terrain.

2. Data and Methods

2.1. Study Region

Xining Caojiapu International Airport (36°31′N, 102°02′E; elevation 2400 m MSL) is located on the broad Huangshui River alluvial plain at the northeastern margin of the Tibetan Plateau. The surrounding terrain rises above 2500 m, forming a semi-enclosed basin. This distinctive topography channels nocturnal drainage flows and promotes pronounced thermal inversions under clear-sky conditions. The airport is equipped with parallel runways (oriented 070°/250°) and standardized approach corridors, which often intersect with localized low-level jets and terrain-induced wind surges, resulting in a highly dynamic environment for aviation operations. In light of these complexities, our analysis focuses on the spatiotemporal distribution of wind shear within the near-surface layer during early summer 2022.

2.2. Synoptic and Meteorological Background: June 2022

To contextualize the June dataset, we surveyed meteorological records for Xining and the surrounding region, and summarized them as follows:

➢

Temperature: June 2022 in Xining was characterized by significant diurnal temperature variation, with daily maxima generally ranging from 22 °C to 27 °C and minima between 9 °C and 13 °C. The mean monthly temperature was close to 17 °C, approximately 1 °C above the local climatological average, reflecting persistent sunny intervals and occasional warm air intrusions.

➢

Precipitation: Total precipitation for the month was moderate, with several convective events (notably on 7, 13, and 26 June), resulting in 40–50 mm of rainfall—below the long-term June norm. Most days were dry, supporting strong nocturnal radiative cooling and frequent formation of surface inversions.

➢

Humidity and Radiation: Mean relative humidity fluctuated between 40 and 60% during daylight hours, increasing sharply at night. Sunshine duration remained high (~270 h for the month), consistent with the region’s typical summer climate.

➢

Synoptic Setting: June marked the gradual onset of the East Asian summer monsoon’s influence over the northeastern Tibetan Plateau, with alternating periods of high-pressure control and intermittent weak trough passages. These synoptic variations contributed to both the convective precipitation events and the episodes of pronounced wind shear.

2.3. Instrumentation and Data Collection

As shown in Figure 1, a state-of-the-art LKW-03 coherent Doppler Lidar (CDL) system was deployed adjacent to the primary runway. The CDL operates at a wavelength of 1550 nm, offering range resolutions of 15 m and 30 m, a blind zone of 50 m, and radial wind-speed measurements from −60 to +60 m/s with a precision better than 0.1 m/s. Full-azimuth and wide-elevation coverage are achieved through a dual-axis gimbal mount, enabling dynamic scanning along both takeoff and landing glide paths. The Lidar’s rapid update interval (30 s) enables high-resolution temporal monitoring of evolving wind profiles and shear events. Detailed system parameters are provided in

Table 1. Parameters of the used LKW-03 CDL.

Technical Parameters	Technical Indicators
Wavelength (nm)	1550 ± 1
Distance resolution (m)	15/30
Blind-spot width (m)	50
Radial wind-speed measurement range (m/s)	±60
Radial wind-speed measurement accuracy (m/s)	≤0.1
Number of range gates	400
Azimuth (°)	0~360
Elevation (°)	−2~90
Angular resolution (°)	0.002
Pointing accuracy (°)	≤0.005
Maximum position update rate (Hz)	5

. Twice-daily radiosonde launches (at 00:00 and 12:00 UTC) supplemented the Lidar observations, providing thermodynamic and wind profiles up to 30 km. Radiosonde ascent rates averaged approximately 5 m/s, with GPS-derived winds collocated for validation.

Rigorous data quality control procedures were implemented to remove spurious outliers and low-signal returns. Only periods with continuous, high-quality Lidar and radiosonde data were included in the June 2022 analysis. Data points with a signal-to-noise ratio (SNR) below 5 dB, profiles with more than 10% missing gates, or physically implausible values were excluded. Shear events were classified by their magnitude, persistence, and association with synoptic and mesoscale forcing. For each Lidar scan, radial velocities were projected onto the approach and departure axes to estimate the headwind component. Data were aggregated in 10 m vertical bins and processed to extract key wind shear metrics (ΔV/Δz) using centered differences over 100 m layers.

2.4. Wind Shear and Stability Indices

In this study, wind shear thresholds and grading criteria were established in accordance with ICAO recommendations. The use of a 100 m layer thickness is consistent with the ICAO operational definition of low-altitude wind shear gradients, enabling classification into four grades [17,18,19,20]: I (<3 m/s), II (3–6 m/s), III (6–9 m/s), and IV (>9 m/s). Although ICAO-recommended wind shear thresholds (3, 6, and 9 m/s) were adopted to facilitate comparison with previous international studies, these criteria are also consistent with the operational safety requirements for plateau airports in China. This alignment facilitates the interpretation of local results within an internationally recognized risk framework.

We calculate the vertical wind shear ΔV/Δz as the absolute difference in headwind speed over a finite layer Δz = 100 m, centered at height z:

$${\displaystyle \frac{\Delta V}{\Delta z}}={\displaystyle \frac{\left|V_h(z+\frac{\Delta z}{2})-V_h(z-\frac{\Delta z}{2})\right|}{\Delta z}}$$

(1)

The gradient Richardson number is defined as follows:

$$Ri={\displaystyle \frac{g/\theta \cdot \partial \theta /\partial z}{{(\partial V/\partial z)}^2}}$$

(2)

where g is gravitational acceleration, θ is the potential temperature, and ∂V/∂z represents the vertical gradient of horizontal wind speed [21,22]. To enable a direct comparison between radiosonde and Lidar data, all profiles were interpolated onto a common 10 m vertical grid. Radiosonde temperature profiles were converted to potential temperature (θ) prior to interpolation. Vertical gradients of potential temperature (∂θ/∂z) and horizontal wind speed (∂V/∂z) were calculated using centered finite differences for interior grid points, while forward and backward differences were applied at the uppermost and lowermost levels, respectively. This collocated thermodynamic–kinematic analysis yields a dimensionless index of shear-to-buoyancy production. Layers with Ri < 0.25 are identified as shear-dominated and susceptible to Kelvin–Helmholtz instability (enhanced mixing), while Ri > 1 indicates strong, buoyancy-dominated capping inversions that suppress turbulence despite high shear. The vertical gradient of total horizontal wind speed was used in the Richardson number calculation, consistent with its standard definition in atmospheric dynamics.

3. Observations

3.1. Results of Lidar Detections in June 2022

Extreme wind shear values exceeding the 99th percentile or physically implausible thresholds (e.g., >60 m/s) were flagged and cross-validated with synoptic weather records. Outliers attributed to instrumental or processing artifacts were excluded from a subsequent analysis. The vertical distribution of severe wind shear (Grade IV, |Δv| ≥ 6 m/s) in June 2022 is illustrated in Figure 2. The highest incidence occurs in the near-surface layers, with approximately 45% at 0 m and 39% at 100 m above ground level (AGL), reflecting the combined influence of surface roughness near the runway and orographic channeling. The frequency decreases to about 28% at 300 m, then rises again to nearly 35% between 400 and 500 m, corresponding to the top of the mixed layer where buoyant thermals intersect with ambient wind gradients. Above 500 m, the occurrence of wind shear steadily declines to around 25% by 800 m, consistent with turbulence dissipation in the residual layer. This bimodal vertical profile highlights two critical shear bands: one immediately above the surface and another at the convective-layer inversion.

Figure 3 further elucidates the shear mechanisms by plotting shear magnitude (|Δv|) against vertical gradient (∂v/∂z), color-coded by height. Roughly 83% of all data points lie in the weak-shear regime (|Δv| < 3 m/s; ∂v/∂z < 1 m⁻¹), which is typical of mixed-layer turbulence. Level III events (3–6 m/s) span gradients up to ~2 m⁻¹, while Level IV events (>6 m/s) define a distinct branch at gradients of 1–6 m⁻¹, concentrated around 300–600 m AGL. A few outliers exceed ∂v/∂z > 6 m⁻¹, with these likely associated with sharp inversion ceilings or frontal boundaries. The scatter-plot distribution thus discriminates between boundary-layer, turbulence-driven shear and sharp inversion-layer shear.

The overall frequency of wind shear levels across all the Lidar profiles is summarized in Figure 4. ‘No risk’ events (|Δv| < 3 m/s) account for 52.1%, ‘Mild’ events (3–6 m/s) for 11.3%, ‘Moderate’ events for 0.6%, and ‘Serious’ (|Δv| ≥ 6 m/s) for 36.0%. Notably, over one-third of the observations exceed the IV-level threshold, indicating a substantial frequency of hazardous shear conditions. These probabilistic distributions provide a quantitative basis for weighting risk in wind shear alert algorithms and flight-management systems according to the time of day, altitude, and severity.

The diurnal variation in wind shear occurrence (|Δv| ≥ 3 m/s) is shown in Figure 5. Shear frequency remains low (22–25%) between 0000 and 0600 LST, reflecting a stable nocturnal boundary layer with minimal vertical mixing. After sunrise, surface warming destabilizes the near-surface layer; shear occurrence exceeds 25% by 0700 LST and rises above 40% by 1100 LST as the convective boundary layer deepens and turbulence intensifies. Shear peaks at approximately 50% between 1300 and 1500 LST, coinciding with maximum mixed-layer depth and intensified topographic channeling. By 2000 LST, reduced insolation and restored stratification lower shear frequency to below 30%, highlighting a critical hazard window for late-morning and early-afternoon flight operations. According to international operational guidelines (e.g., ICAO, FAA), a wind shear incidence rate exceeding 10–15% (for |Δv| ≥ 3 m/s) during critical flight phases is generally considered a threshold for heightened alert and operational risk. The much higher occurrence rates observed here—frequently exceeding 40–50% during midday—greatly surpass conventional warning levels and underscore the exceptional wind shear risk at Xining Airport.

Figure 6 presents the evolution of the daily maximum wind shear (max |Δv|) from 1 June to 1 July 2022. Between 1 and 12 June, the maximum values remain modest (3–7 m/s), indicative of boundary-layer processes. From 13 to 21 June, a synoptic-scale cut-off low and frontal passages elevate the maximum to approximately 23 m/s. On 25 June, a cold front accompanied by a convective band generated an extreme event (approximately 47 m/s), causing the highest value recorded during the month. Subsequently, values decrease to 9–20 m/s as large-scale forcing diminished. These temporal transitions underscore the shift from local convective dynamics to increasing meso- and synoptic-scale influences on shear intensity.

3.2. Causes of Wind Shear in June 2022

To analyze the causes of low-altitude wind shear, we selected three days with typical characteristics for further analysis. Three representative summer days—06, 18, and 25 June 2022—were chosen to span the full spectrum of low-level thermal stratification and its impact on wind shear:

➢

The day of 06 June experienced a persistently convective boundary layer with no significant inversion at dawn or dusk, illustrating conditions of well-mixed turbulent transport.

➢

The day of 18 June featured a pronounced low-level inversion at 08 LST that persisted into the morning but was eroded by daytime heating by 20 LST, exemplifying a transition from stable to convective regimes.

➢

The day of 25 June exhibited near-neutral lapse rates at 08 LST, evolving into a strong nocturnal inversion by 20 LST, demonstrating the classic diurnal development of a stable residual layer.

Specifically, low-level wind shear (0–0.8 km AGL) was quantified on three representative summer days (6, 18, and 25 June 2022) using twice-daily radiosonde launches at 08 and 20 LST. Temperature and wind-speed profiles (Figure 7 and Figure 8) were used to compute bulk shear:

$$S={\displaystyle \frac{U_{0.8km}-U_{0km}}{0.8km}}$$

(3)

Table 2. Summary of 0–0.8 km low-level bulk wind shear on selected case days.

Date	Time	Lapse Rate (K/km)	Wind Range (m/s)	Bulk Shear (m/s/km)	Dominant Stratification
06 Jun 2022	08 LST	+6.25 (unstable)	2.8 → 4.4	6.45	Convective unstable
	20 LST	+8.6 (residual mix)	3.9 → 4.7	2.06	Residual mixed layer
18 Jun 2022	08 LST	−1.0 (low-level)	1.9 → 2.6	4.02	Pronounced dawn inversion
	20 LST	+9.5 (re-mixed)	8.3 → 10.5	1.37	Daytime mixed convective layer
25 Jun 2022	08 LST	+2.5 (near-neutral)	0.3 → 2.3	2.31	Weakly unstable/neutral
	20 LST	−6.2 (strong)	8.0 → 4.0	6.15	Strong nocturnal inversion

summarizes the mean lapse rates, wind-speed ranges and bulk shear values on these dates.

06 June 2022:

➢

08 LST: The layer was convectively unstable, with the temperature falling from 8.5 °C at 0.1 km to 3.5 °C at 0.8 km (lapse ≈ 6.25 K/km). Winds increased from ≈2.8 m/s near the surface to ≈4.4 m s⁻¹ aloft, yielding the largest bulk shear of the series (6.45 m/s/km).

➢

20 LST: A residual mixed layer persisted (lapse ≈ 8.6 K/km), but winds were more uniform (3.9 → 4.7 m/s), reducing the bulk shear to 2.06 m/s/km.

18 June 2022:

➢

08 LST: A pronounced low-level inversion (ΔT ≈ +0.7 K between 0.1 and 0.4 km) inhibited vertical mixing. Above 0.4 km, temperature then increased to 16.0 °C at 0.8 km. Winds rose from ≈1.9 m/s at 0.1 km to ≈3.4 m/s at 0.8 km, giving a moderate shear of 4.02 m/s/km.

➢

20 LST: Daytime heating had re-mixed the boundary layer (lapse ≈ 9.5 K/km). Winds increased smoothly from ≈8.3 m/s near the surface to ≈10.5 m/s aloft, producing the weakest bulk shear of 1.37 m s/km.

25 June 2022:

➢

08 LST: The temperature decreased modestly from 17.5 °C at 0.1 km to 15.5 °C at 0.8 km (mean lapse ≈ 2.5 K/km), indicating a near-neutral to weakly unstable layer. Surface winds were light (<1.5 m/s), increasing gradually to ≈2.3 m/s at 0.8 km. Consequently, the bulk shear was only 2.31 m/s/km.

➢

20 LST: A strong nocturnal inversion developed (ΔT ≈ +5 K over 0–0.8 km), decoupling the surface from the residual layer. Winds at the surface strengthened to ≈8 m/s while aloft they remained at ≈4–5 m/s. This produced a bulk shear of 6.15 m/s/km and a pronounced shear maximum near the inversion base (0.6–0.7 km).

High-resolution radiosonde profiles obtained at 08:00 and 20:00 local standard time on 6, 18, and 25 June 2022 were used to construct continuous vertical sections of the Brunt–Väisälä frequency squared (N²) and the Richardson number (Ri) from the surface to 12 km above ground level (AGL), diagnosing the multi-scale coupling between thermal stratification and vertical wind shear. As illustrated in Figure 9, layers where Ri falls below 0.25—especially negative values—correspond to dynamically unstable regions, indicating a high likelihood of turbulent mixing and wind shear generation. In contrast, layers with Ri > 0.25 are dynamically stable and less susceptible to shear-induced turbulence. On 6 June, the 08:00 LST profile showed near-zero N² throughout the lower troposphere (0–2 km) with Ri < 0.25, indicating a fully mixed convective layer. In contrast, a pronounced peak in N² (~6 × 10⁻⁴ s⁻²) and Ri ≫ 1 at 10–11 km marked tropopause inversion. At 20:00 LST, elevated N² and Ri values again appeared at the residual mixed-layer top (∼1–2 km) and at the tropopause, while the bulk shear below remained at Ri values < 0.5. On 18 June, a strong dawn inversion between 0.1 and 0.4 km produced local N² maxima of 3–4 × 10⁻⁴ s⁻² and Ri values exceeding 10–100 at the inversion base, indicating a narrow shear-critical interface, while the layers immediately above and below remained well mixed (Ri < 0.25). By 20:00 LST, the low-level inversion had eroded (N² ≈ 0 and Ri < 0.25 below 2 km), confirming rapid convective re-mixing. On 25 June, the morning profile indicated transitional stratification, characterized by a weakly stable layer (N² ≈ 1–2 × 10⁻⁴ s⁻², Ri ≈ 0.2–0.7) spanning 0–1 km, with overlying small peaks indicating residual stratification. By nightfall, dual inversion peaks at 0.2–0.7 km (N² up to ~6 × 10⁻⁴ s⁻², Ri ≈ 10–40) and at the 10–11 km tropopause delineated two dominant momentum-decoupling interfaces.

To elucidate the dynamic response of the lower troposphere to diurnal thermal forcing, we analyze height-resolved vector wind shear profiles (0–0.8 km AGL) at sunrise (08 LST) and dusk (20 LST) on three representative summer days—6, 18, and 25 June 2022. This approach identifies the shear-critical interfaces associated with evolving inversion layers. As shown in Figure 10, at 08 LST on 6 June, the vector shear increased sharply from approximately 10 m/s/km at 0.8 km to a maximum of about 18 m/s/km at 0.1 km. This pattern reflects the presence of a residual nocturnal inversion base near the surface, which decouples low-level momentum and concentrates the vertical wind gradient. A secondary plateau of approximately 12–15 m/s/km between 0.2 and 0.4 km corresponds to the residual mixed-layer top or the base of the low-level jet, above which shear decreases as turbulent mixing weakens. By 20 LST on 6 June, the nocturnal inversion was fully developed (ΔT ≈ +5 K over 0–0.8 km), resulting in a shear minimum of approximately 2.1 m/s/km at 0.2 km, where momentum exchange is most suppressed, followed by a rapid increase to 8–9 m/s/km at 0.6–0.8 km. This inversion-base shear profile underscores how stable stratification inhibits near-surface mixing and shifts the primary shear zone upward.

At 08 LST on 18 June, a strong dawn inversion between 0.1 and 0.4 km generated shear values increasing from approximately 5 m/s/km at the base of the inversion to about 15 m/s/km at its peak. Immediately below (<0.1 km) and above (>0.4 km) this layer, shear decreases to 5–7 m/s/km, delineating a narrow, shear-critical interface. By 20 LST on 18 June, convective re-mixing largely erased the inversion (Ri < 0.25), yielding a smoother shear profile of 6–14 m/s/km that steadily decreases with height, characteristic of a well-mixed residual layer.

At 08 LST on 25 June, the low-level profile exhibits a moderate shear of 4 m/s/km at 0.1 km, increasing to 7–8 m/s/km near 0.2–0.4 km where a weak inversion persists, then decreasing to about 2 m/s/km by 0.8 km as the mixed layer develops. At 20 LST on 25 June, dual inversion peaks between 0.2 and 0.7 km produced shear maxima of 20–22 m/s/km, with a local minimum of about 12 m/s/km at 0.1 km and a gradual decrease to 16 m/s/km at 0.8 km. This double-peak structure highlights how successive stable layers can partition the lower troposphere into multiple momentum-decoupled interfaces.

These high-fidelity shear profiles demonstrate that stable inversion layers—whether nocturnal or residual—act as thin, high-Ri boundaries that concentrate vertical wind gradients, whereas well-mixed convective layers produce low-Ri boundaries and more homogeneous shear distributions. This multi-scale coupling between thermal stratification and momentum transport fundamentally regulates turbulence generation, pollutant dispersion, and aviation turbulence risk in the lower troposphere.

3.3. Synoptic Backgrounds of the Extreme Wind Shear Day

To comprehensively characterize wind shear at Xining Caojiapu Airport under varying meteorological conditions, detailed case studies were conducted for three representative days—6, 18, and 25 June 2022—using ERA5 reanalysis data, as illustrated in Figure 11. On 6 June, a synoptic analysis indicated relatively uniform mean sea level pressure patterns with weak pressure gradients across the study area. The 10 m wind field exhibited low wind speeds and limited directional variability, with minimal convective activity or precipitation. Under these stable conditions, the observed wind shear was weak, and vertical changes in both wind speed and direction were modest. This day served as a control, reflecting the background wind environment in the absence of significant synoptic forcing.

In contrast, 18 June was marked by a pronounced frontal system, indicated by closely spaced isobars and a well-defined pressure trough. ERA5 data revealed enhanced low-level wind speeds, abrupt wind direction shifts near the frontal zone, and organized bands of precipitation. The frontal passage was accompanied by strong pressure gradients, resulting in significant wind shear both near the surface and aloft. Lidar and radiosonde observations during this event confirmed pronounced vertical wind shear associated with frontal lifting and intense synoptic-scale disturbances.

On 25 June, the meteorological background was dominated by localized convective storms and mesoscale disturbances. The pressure field was less organized, but pockets of strong pressure gradients were present alongside vigorous convective activity, as evidenced by intense and spatially variable precipitation. The wind field exhibited pronounced local fluctuations, with abrupt changes in wind speed and direction likely driven by convective downdrafts and outflow boundaries. These conditions produced short-lived but potentially hazardous wind shear events near the airport, underscoring the impact of convective weather on wind field variability.

In summary, these case studies demonstrate that the most intense and hazardous wind shear episodes at the plateau airport are closely linked to frontal passages and strong convective events. These findings highlight the need for integrated monitoring of both synoptic-scale and mesoscale meteorological factors to accurately assess and mitigate wind shear risks at high-altitude airports.

4. Discussion

The integrated deployment of high-resolution Doppler Lidar and radiosonde observations at Xining Caojiapu Airport offers new insights into the vertical structure and temporal evolution of low-level wind shear under the complex topographic and climatic conditions of the northeastern Tibetan Plateau. The results reveal a remarkably high frequency and multi-modal distribution of severe wind shear (|Δv| ≥ 6 m/s) during June 2022, posing significant operational risks for aviation in this mountainous region.

4.1. Vertical and Temporal Characteristics of Wind Shear

The vertical distribution of Grade IV (|Δv| ≥ 6 m/s) wind shear exhibits a bimodal structure, with the highest incidence immediately above the surface (~45% at 0 m and ~39% at 100 m AGL), and a secondary maximum at the top of the mixed layer (400–500 m AGL). This pattern identifies two dynamically distinct zones of enhanced shear: the surface layer, shaped by runway-adjacent roughness and local orographic channeling, and the mixed-layer top, where buoyant thermals interact with pronounced ambient wind gradients. Above 500 m, the incidence of wind shear steadily decreases, consistent with turbulent decay in the residual layer. These findings are consistent with previous studies in complex terrain, which emphasize the influence of surface heterogeneity and boundary-layer depth on wind shear risk.

Diurnally, the frequency of significant wind shear (|Δv| ≥ 3 m/s) shows pronounced daytime enhancement. Shear frequency remains low overnight (22–25% between 0000 and 0600 LST), rises rapidly after sunrise due to surface heating and mixed-layer deepening, and peaks at midday (reaching up to 50% during 1300–1500 LST), coinciding with maximum boundary-layer turbulence and intensified topographic flows. These results corroborate classic boundary-layer theory and highlight the heightened risk period for aviation operations during late morning and early afternoon, especially under strong convective forcing.

4.2. Mechanisms of Wind Shear Generation

A detailed analysis of radiosonde-derived temperature and wind profiles on representative days (6, 18, and 25 June 2022) clarifies the physical mechanisms underlying low-level wind shear. The strongest shear events are closely associated with sharp thermal inversions—either nocturnal or residual—that act as momentum-decoupling layers. For example, at 20 LST on 25 June, a pronounced nocturnal inversion (ΔT ≈ +5 K over 0–0.8 km) produced a bulk shear of 6.15 m/s/km, with a distinct shear maximum near the inversion base. In contrast, well-mixed convective conditions (e.g., 08 LST on 6 June) resulted in large lapse rates and high bulk shear, driven by vigorous turbulent transport but characterized by more distributed vertical gradients.

To further examine the relationship between residual inversion duration and wind shear, we analyzed the temporal evolution of inversion strength and persistence after sunset. Our observations indicate that when a residual inversion persists for more than 2–3 h after sunset—typically until 22–23 LST—the likelihood of enhanced low-level wind shear increases substantially, with frequent occurrences of bulk shear exceeding 4–6 m/s/km. The longer the residual inversion persists into the night, the greater the tendency for momentum decoupling and the development of sharp vertical wind gradients, particularly near the base of the inversion. These results suggest that prolonged residual inversions are a critical factor in sustaining hazardous wind shear, highlighting the need for continuous monitoring of both thermal structure and wind profiles during evening and early nighttime hours at plateau airports.

The diagnostic profiles of N² and Ri further elucidate the coupling between stratification and wind shear. Stable inversion layers are characterized by maxima in N² and Ri values far exceeding the instability threshold (Ri ≫ 1), acting as thin, high-shear interfaces. In contrast, convective mixing leads to near-zero N² and subcritical Ri (<0.25) values, resulting in smoother and more homogeneous shear distributions. These observations are consistent with previous findings on the interplay between boundary-layer processes, inversion stability, and shear generation (e.g., Sun et al., 2012; Banta et al., 2006) [23,24].

4.3. Integrated Analysis and Operational Implications of Severe Wind Shear

In addition to diurnal and local boundary-layer processes, synoptic-scale weather patterns exert a substantial influence on wind shear intensity and structure at Xining Caojiapu Airport. For example, the event on 25 June 2022—marked by a passing cold front and an intense convective band—produced the highest observed shear of the month (~47 m/s), highlighting the potential for extreme hazards when large-scale forcing coincides with strong local inversion layers. Such cases underscore the necessity of continuous, high-resolution wind monitoring to capture both routine and extreme wind shear events in complex terrain.

From an operational perspective, the proportion of Lidar profiles exceeding severe-shear thresholds at Xining (over 36% at Grade IV, |Δv| ≥ 6 m/s) is substantially higher than that reported at many lowland or coastal airports, and frequently surpasses the standard alert levels defined by major wind shear warning systems worldwide (typically 5–8 m/s over short vertical layers). According to international (ICAO, FAA) and national (CAAC) aviation regulations, such shear magnitudes are classified as significant flight hazards, with the potential to trigger onboard wind shear alerts, prompt pilot advisories, or lead to deviations from planned flight paths during takeoff and landing. Previous investigations and operational reports from similar high-altitude airports have documented instances where comparable shear conditions resulted in turbulence encounters, missed approaches, or runway excursions, further reinforcing the practical risks identified in this study.

Although direct comparison with local wind shear alert records or pilot reports (PIREPs) could not be performed due to the confidential and internal nature of such operational data, the consistency between our quantitative wind shear analysis and the established aviation safety standards provides compelling circumstantial evidence for the real-world impact of these phenomena. This underscores the urgent need to revise wind shear alert criteria and flight management protocols at plateau airports to better reflect the elevated risk profile. In addition, our findings provide a robust foundation for developing probabilistic risk assessment algorithms that incorporate altitude, time of day, and meteorological context. Future work should prioritize building data-sharing partnerships with airport authorities and airline operators, enabling the integration of Lidar-based wind measurements with operational alerts, flight trajectory records, and pilot feedback. Such multi-source datasets will be essential for the comprehensive validation of wind shear risk assessments, the continuous improvement of early warning systems, and the advancement of flight safety in mountainous regions worldwide.

4.4. Limitations and Future Directions

Several limitations should be acknowledged. First, the study period was limited to June 2022 due to the availability of high-quality Lidar and radiosonde data during a targeted deployment. Consequently, our findings may not capture the full seasonal variability of wind shear mechanisms at this plateau airport. In future work, we plan to leverage ongoing and planned Lidar deployments to expand the observational dataset across additional months—particularly autumn, winter, and spring—enabling a systematic assessment of how wind shear characteristics and the associated risks evolve throughout the year. Second, although Lidar and radiosonde observations provide complementary information, increasing the frequency and continuity of radiosonde launches will help resolve rapid transitions, especially during frontal passages or convective outbreaks. Finally, we intend to incorporate numerical simulations and turbulence-resolving models in subsequent studies to generalize our findings and assess their applicability to other high-altitude airports with complex terrain.

This study demonstrates the utility of combining Doppler Lidar and radiosonde datasets to diagnose wind shear hazards in mountainous environments. The methodology and findings are broadly applicable to other high-elevation or topographically complex airports worldwide, where low-level wind shear presents a persistent challenge to flight safety. Moreover, the identified coupling between inversion layers, convective mixing, and momentum transport has broader relevance for air quality management and pollutant dispersion studies in such regions.

5. Conclusions

Integrated Lidar and radiosonde observations reveal that Xining Caojiapu Airport experiences frequent and structurally complex low-level wind shear, with pronounced maxima near the surface and at the top of the mixed layer. The evolution of wind shear is governed by a dynamic interplay among convective turbulence, thermal inversions, and larger-scale weather systems. In particular, inversion layers act as razor-thin, momentum-decoupling boundaries that concentrate vertical wind gradients and elevate turbulence risk. Quantitative analysis indicates that severe wind shear (|Δv| ≥ 6 m/s) occurred in 34% of all the Lidar profiles during June 2022. This high incidence underscores the need for tailored operational protocols and adaptive alert systems in high-altitude, mountainous environments. These findings not only advance understanding of boundary-layer processes over complex terrain but also provide a robust empirical basis for improving flight safety and environmental monitoring at similar sites worldwide.