Wind Shear Events Research Articles

SPE-SHAP: Self-paced ensemble with Shapley additive explanation for the analysis of aviation turbulence triggered by wind shear events

Aviation safety is critically challenged by significant turbulence, often triggered by wind shear events, which jeopardize aircraft controllability and may cause structural damage to the fuselage. Current predictive models, primarily based on Pilot Reports (PIREPs), suffer from data imbalance, undermining their reliability and accuracy. This research addresses the need for a robust approach to accurately estimate and classify aviation turbulence events. We propose a novel framework, the Self-Paced Ensemble (SPE), integrated with Shapley Additive Explanation (SHAP) analysis. This framework utilizes tree-based classifiers such as Gradient Boosting Decision Trees (GBDT), Random Forests (RF), Extreme Gradient Boosting (XGBoost), and Extra Trees (ET) to enhance predictive accuracy and reliability. The SPE model, particularly when combined with XGBoost, demonstrated superior performance with a recall of 70.16 %, specificity of 74.75 %, G-Mean of 73.67 %, Matthews correlation coefficient (MCC) of 0.352, and area under the Receiver Operating Characteristic (ROC) curve of 0.726. Comparatively, the SPE model with RF achieved a recall of 70.05 %, specificity of 73.89 %, G-Mean of 72.57 %, MCC of 0.347, and AU-ROC of 0.721. Post-hoc Wilcoxon signed-rank tests further validated these results. SHAP analysis identified wind shear altitude, magnitude, and causes as the most influential factors affecting significant aviation turbulence prediction. The SPE-SHAP framework not only significantly improves the prediction of aviation turbulence but also enhances model transparency by elucidating key contributing factors. These insights can inform updates to aircraft operation protocols and enhance flight training, thereby increasing safety during wind shear events.

Wind Shear Operation-Based Competency Assessment Model for Civil Aviation Pilots

Wind Shear Operation-Based Competency Assessment Model for Civil Aviation Pilots

Assessment of Wind Shear Severity in Airport Runway Vicinity using Interpretable TabNet approach and Doppler LiDAR Data

ABSTRACT Severe wind shear events near airport runways pose serious safety risks and are a growing concern in civil aviation. Identification of severe wind shear risk factors may enhance aviation safety. Although rare, severe wind shear impacts safety by affecting the airspeed, lift, and maneuverability of aircraft. This study presents TabNet, a novel deep learning technique coupled with Bayesian optimization (BO) to predict wind shear severity in the runway vicinity using Doppler LiDAR data from Hong Kong International Airport. To address imbalanced wind shear data, it was first processed by resampling techniques and then used as input to TabNet. The analysis demonstrated that Bayesian-tuned TabNet (BO-TabNet) with SVM-SMOTE-processed data led to better performance compared to other strategies. The TabNet architecture employs the attention mechanism to enable model-specific interpretability. Analysis showed that the most important contributing factor was the summer season, followed by the wind shear encounter location (1 nautical miles from the runway at the departure end). Additionally, a more comprehensive model-agnostic LIME method was used to elucidate the model from a local perspective. By predicting severe wind shear and assessing contributing factors, aviation stakeholders can proactively manage and mitigate the associated risks, leading to safer and more efficient operations.

Interpretable ensemble imbalance learning strategies for the risk assessment of severe-low-level wind shear based on LiDAR and PIREPs.

The occurrence of severe low-level wind shear (S-LLWS) events in the vicinity of airport runways poses a significant threat to flight safety and exacerbates a burgeoning problem in civil aviation. Identifying the risk factors that contribute to occurrences of S-LLWS can facilitate the improvement of aviation safety. Despite the significant influence of S-LLWS on aviation safety, its occurrence is relatively infrequent in comparison to non-SLLWS incidents. In this study, we develop an S-LLWS risk prediction model through the utilization of ensemble imbalance learning (EIL) strategies, namely, BalanceCascade, EasyEnsemble, and RUSBoost. The data for this study were obtained from PIREPs and LiDAR at Hong Kong International Airport. The analysis revealed that the BalanceCascade strategy outperforms EasyEnsemble and RUSBoost in terms of prediction performance. Afterward, the SHapley Additive exPlanations (SHAP) interpretation tool was used in conjunction with the BalanceCascade model for the risk assessment of various factors. The four most influential risk factors, according to the SHAP interpretation tool, were hourly temperature, runway 25LD, runway 25LA, and RWY (encounter location of LLWS). S-LLWS was likely to happen at Runway 25LD and Runway 25LA in temperatures ranging from low to moderate. Similarly, a high proportion of S-LLWS events occurred near the runway threshold, and a relatively small proportion occurred away from it. The EIL strategies in conjunction with the SHAP interpretation tool may accurately predict the S-LLWS without the need for data augmentation in the data pre-processing phase.

Numerical Simulation of Flow over an Airport During A Low-level Wind Shear Event

In this paper we present the flow features of a severe low-level wind shear event near an airport simulated using a high-resolution numerical weather prediction model. The open-source model Weather Research and Forecasting (WRF) was used for the numerical simulations. Our initial analysis shows the ability of the model in capturing the possibility of the wind shear event.

Characterization of wind speed and directional shear at the AWAKEN field campaign site

The American wake experiment (AWAKEN) is taking place in northern Oklahoma, USA, close to the Atmospheric Radiation Measurement Southern Great Plains (ARM SGP) atmospheric observatory. The planning for the deployment of the instruments in this observational field campaign required an assessment of the wind characteristics of the site. This paper analyzes long-term data collected by instruments at the ARM SGP observatory to characterize the winds near the AWAKEN site. The analysis shows that this site experiences high wind shear and veer events with a large number of nocturnal low-level jets. A total of 7086 low-level jet wind profiles over 6 years are examined and found to be dominant from the south and southeast. Significant nocturnal wind veer is observed, which causes southerly wind near the surface to become westerly wind aloft. By identifying a strong relationship between atmospheric stability and wind shear, the wind shear at the site is predicted using the Monin–Obukhov similarity theory (MOST) and validated with the observational data collected by a scanning Doppler lidar. The results show that wind speed at a height of 91 m, a proxy hub height for wind turbines in this area, can be predicted from data collected at a height of 10 m with a bias of −0.35 and 0.65 m s−1 in unstable and stable atmospheric boundary layers, respectively. The bias of the predicted wind speed is mostly in the region of low wind speed, and wind speed above 5 m s−1 at a height of 91 m can be predicted with a bias of less than 0.2 m s−1, and the limitations of the MOST in predicting winds during the stably stratified boundary layer is well-observed.

Time-Series Prediction of Intense Wind Shear Using Machine Learning Algorithms: A Case Study of Hong Kong International Airport

Machine learning algorithms are applied to predict intense wind shear from the Doppler LiDAR data located at the Hong Kong International Airport. Forecasting intense wind shear in the vicinity of airport runways is vital in order to make intelligent management and timely flight operation decisions. To predict the time series of intense wind shear, Bayesian optimized machine learning models such as adaptive boosting, light gradient boosting machine, categorical boosting, extreme gradient boosting, random forest, and natural gradient boosting are developed in this study. The time-series prediction describes a model that predicts future values based on past values. Based on the testing set, the Bayesian optimized-Extreme Gradient Boosting (XGBoost) model outperformed the other models in terms of mean absolute error (1.764), mean squared error (5.611), root mean squared error (2.368), and R-Square (0.859). Afterwards, the XGBoost model is interpreted using the SHapley Additive exPlanations (SHAP) method. The XGBoost-based importance and SHAP method reveal that the month of the year and the encounter location of the most intense wind shear were the most influential features. August is more likely to have a high number of intense wind-shear events. The majority of the intense wind-shear events occurred on the runway and within one nautical mile of the departure end of the runway.

CORRELATION OF ATMOSPHERIC LABILITY INDEX TO VERTICAL WIND SHEAR AT I GUSTI NGURAH RAI AIRPORT

Wind shear is a condition that is detrimental to aircraft because it can cause the aircraft to experience lift, especially during take-off or landing, where wind shear can occur due to bad and unstable weather, so research on the correlation of the atmospheric lability index to wind shear is very important to prevent a plane crash. This research aims to determine the magnitude of the correlation of the atmospheric lability index to the vertical wind shear that occurs at I Gusti Ngurah Rai Airport, Bali. In this research, the Wind Profiler Radar (WPR) brand scintec LAP-3000 was used to obtain wind shear data in the form of wind direction and wind speed data then radiosonde to obtain atmospheric lability index data in the form of Lifted Index (LI), Total-Totals Index (TT), K-Index (KI), and Convective Available Potential Energy (CAPE) index values for each month in 2019-2020. Wind shear measurement was carried out at an altitude of 100-3000 m and data were taken on the difference in wind direction ≥ 60o and the difference in wind speed ≥ 10 knots. Meanwhile, measurement of the atmospheric lability index is carried out by a flying air balloon that has been equipped with a transmitter. All LI, TT, KI, and CAPE index value data and the number of wind shear events were analyzed with Pearson Product Moment correlation. The value of the correlation coefficient (r) between the LI, TT, KI, and CAPE indices was obtained for successive wind shear events of -0.786; 0,250; 0.738, and 0.713. These results show that the LI, KI, and CAPE index values can be used as a reference to predict wind shear events because they have a strong correlation, obtaining a correlation value between 0.60 to 0.79 compared to the TT index.

Prediction of Aircraft Go-Around during Wind Shear Using the Dynamic Ensemble Selection Framework and Pilot Reports

Pilots typically implement the go-around protocol to avoid landings that are hazardous due to wind shear, runway excursions, or unstable approaches. Despite its rarity, it is essential for safety. First, in this study, we present three Dynamic Ensemble Selection (DES) frameworks: Meta-Learning for Dynamic Ensemble Selection (META-DES), Dynamic Ensemble Selection Performance (DES-P), and K-Nearest Oracle Elimination (KNORAE), with homogeneous and heterogeneous pools of machine learning classifiers as base estimators for the prediction of aircraft go-around in wind shear (WS) events. When generating a prediction, the DES approach automatically selects the subset of machine learning classifiers which is most probable to perform well for each new test instance to be classified, thereby making it more effective and adaptable. In terms of Precision (86%), Recall (83%), and F1-Score (84%), the META-DES model employing a pool of Random Forest (RF) classifiers outperforms other models. Environmental and situational factors are subsequently assessed using SHapley Additive exPlanations (SHAP). The wind shear magnitude, corridor, time of day, and WS altitude had the greatest effect on SHAP estimation. When a strong tailwind was present at low altitude, runways 07R and 07C were highly susceptible to go-arounds. The proposed META-DES with a pool of RF classifiers and SHAP for predicting aircraft go-around in WS events may be of interest to researchers in the field of air traffic safety.

Hong Kong Airport Wind Shear Now-Casting System Development and Evaluation

A wind-shear now-casting system for Hong Kong International Airport is described. The system has been configured and run retrospectively for the period from January to April 2018 at 20-min intervals using two sets of meteorological inputs to quantify uncertainty. Outputs from the system include calculations of headwind for the northerly and southerly runways at the airport. Six metrics have been defined that attempt to identify areas of strong wind shear; these involve quantification of headwind and headwind gradient in both the horizontal and vertical directions. Receiver operating characteristic (ROC) curves have been generated for all metrics, runways, and meteorological conditions using pilot wind-shear reports to define wind-shear events. These curves allow for derivation of metric thresholds that could be used within an operational system to provide wind-shear alerts. The skill of the system for the 4-month period was quantified in terms of the probability of detection (POD, ideal value 1.0) of wind-shear events; system performance was better for the southerly runway, with the majority of POD values in the range 0.6 to 0.8.

Causes and impact of extreme Low Level Wind Shear (LLWS) event at Soekarno-Hatta International Airport

The impact of low-level wind shear (LLWS) is still a critical issue in aviation. These days, as the aviation industry grows, it must follow safety improvement efforts and transportation management, one of which is through the accurate publishing of LLWS early warning. Therefore, it takes a high understanding related to phenomena, especially at airports with busy schedules. This study examined the causes and impacts of pilot-reported LLWS on April 5, 2018, from 13:15 to 16:00 UTC at Soekarno-Hatta International Airport which is the busiest airport in Indonesia. The data used are Doppler weather radar, automated weather observing system (AWOS), as well as synoptic and upper air observations. Based on the result of this research, linear convective clouds created from deep convection due to very unstable atmospheric circumstances are the cause of LLWS. A sizeable line of convective cloud with a 50-55 dBZ reflectivity accompanied by an outflow microburst (MBA) was detected crashing the airport. Heavy rains and strong winds up to 15-20 m/s speeds from the line of convective cloud triggered the formation of destructive LLWS. Subsequently, there was traffic build-up at 5 flight holding points before landing. Furthermore, 7 flights (Airbus, Boeing, and CRJX) underwent go-around, and 3 flights (Airbus and Boeing) were diverted.

Effects of buildings on wind shear at the airport: Field measurement by coherent Doppler lidar

Wind shear has a negative impact on aircraft safety, particularly during takeoff and landing processes. The buildings inside or around the airport can induce wind shear by disturbing the airflow along the glide path. This paper investigates the building-induced wind shear along the glide path of aircraft by utilizing two coherent Doppler lidar systems deployed at Beijing Capital International Airport. A well-designed coordinated scanning strategy of lidar was used to measure the building-induced wind shear along the glide path and the wind field influenced by buildings. With a correlation coefficient of 0.97 for wind speed and 0.99 for wind direction, the wind field estimated by lidar was well validated by those from sonic anemometers. In addition, the lidar-observed wind shear events are verified by pilot reports. The results show that building wakes bring the velocity deficit and turbulence increase along the glide paths. It is susceptible to wind shear due to the rapidly changing wind speeds when the aircraft flies through these wake regions. The relative frequencies of wind shear events reached above 70 and 50 at 36L and 01, respectively, in the wind direction range of 270°∼300° with a wind speed greater than 12 m/s. The statistical analysis of lidar-observed wind shear reveals that the wind shear frequently occurs within the height of 40 ∼ 80 m and the distance of 763 ∼ 1526 m at both 36L and 01.

Historical analysis (2001–2019) of low‐level wind shear at the Hong Kong International Airport

AbstractThis paper analyses over 10,000 quality‐controlled pilot reports of low‐level wind shear (‘wind shear’) at the Hong Kong International Airport (HKIA; ICAO code: VHHH) between 2001 and 2019. HKIA is well known for its susceptibility to wind shear, which is a potential hazard for aircraft during landing and take‐off. Wind shear at HKIA exhibits strong seasonality with double peaks in the middle of the spring and summer months. There is a strong diurnal cycle in report numbers, peaking towards the afternoon, as modulated by air traffic at HKIA. On average, there are 115 days per year with wind shear reported, with possible power law distribution in the number of days with different number of reports. By comparing the background wind distribution during those minutes with pilot reports against the climatological distribution, wind shear is observed to be, on average, favoured under higher wind speeds (≥6 m s−1) and reduced under lower speeds with shifted frequency peak and heavier tail, although statistical behaviour can differ significantly across individual runway corridors. Preferred background wind directions for wind shear occurrence highlight the role of terrain influence upstream of HKIA. While positive shear (65.6%) is more often reported than negative shear (34.4%), both types of wind shear events appear to follow an exponential distribution in shear magnitude, with similar rates of decay. This is by far the largest statistical study of low‐level wind shear at an airport. Results can serve as an important local reference in addition to being a useful example for airports around the world.

Extreme wind shear events in US offshore wind energy areas and the role of induced stratification

Abstract. As the offshore wind industry emerges on the US East Coast, a comprehensive understanding of the wind resource – particularly extreme events – is vital to the industry's success. Such understanding has been hindered by a lack of publicly available wind profile observations in offshore wind energy areas. However, the New York State Energy Research and Development Authority recently funded the deployment of two floating lidars within two current lease areas off the coast of New Jersey. These floating lidars provide publicly available wind speed data from 20 to 200 m height with a 20 m vertical resolution. In this study, we leverage a year of these lidar data to quantify and characterize the frequent occurrence of high-wind-shear and low-level-jet events, both of which will have a considerable impact on turbine operation. In designing a detection algorithm for these events, we find that the typical, non-dimensional power-law-based wind shear exponent is insufficient to identify many of these extreme, high-wind-speed events. Rather, we find that the simple vertical gradient of wind speed better captures the events. Based on this detection method, we find that almost 100 independent events occur throughout the year with mean wind speed at 100 m height and wind speed gradient of 16 m s−1 and 0.05 s−1, respectively. The events have strong seasonal variability, with the highest number of events in summer and the lowest in winter. A detailed analysis reveals that these events are enabled by an induced stable stratification when warmer air from the south flows over the colder mid-Atlantic waters, leading to a positive air–sea temperature difference.

Assessing the risk of windshear occurrence at HKIA using rare‐event logistic regression

AbstractLow‐level windshear poses serious hazards to the flight safety of civil aircraft. The objective of the study was to assess the risk of windshear occurrence within 1 hr at runway corridors at Hong Kong International Airport (HKIA). Windshear report records from HKIA and weather data around HKIA were obtained from Hong Kong Observatory (HKO) between January 2016 and December 2018. The rare‐event logistic regression was employed to evaluate windshear risk considering the imbalanced nature of the data. Several explanatory variables were found to be significantly correlated with the risk of the hourly occurrence of the windshear. Results indicated that at HKIA, March, April, July and August are associated with a higher likelihood of windshear events, while November suffered a significantly lower windshear risk. Besides, winds from the east, southeast, south and southwest are more likely to cause windshear events at HKIA. However, the study found that it was hard to predict windshear occurrence precisely only with the hourly weather variables due to the transient and sporadic nature of windshear. Therefore, more refined data are needed to improve the prediction performance of regression models for windshear occurrence. The present study is the first to analyse the windshear occurrence at airports using the rare‐event logistic regression method. Findings from the study can provide some references and valuable information for related researchers to understand low‐level windshear.

Low-Level Wind Shear Characteristics and Lidar-Based Alerting at Lanzhou Zhongchuan International Airport, China

Lanzhou Zhongchuan International Airport [International Civil Aviation Organization (ICAO) code ZLLL] is located in a wind shear prone area in China, where most low-level wind shear events occur in dry weather conditions. We analyzed temporal distribution and synoptic circulation background for 18 dry wind shear events reported by pilots at ZLLL by using the NCEP final (FNL) operational global analysis data, and then proposed a lidar-based regio-nal divergence algorithm (RDA) to determine wind shear intensity and location. Low-level wind shear at ZLLL usually occurs in the afternoon and evening in dry conditions. Most wind shear events occur in an unstable atmosphere over ZLLL, with changes in wind speed or direction generally found at 700 hPa and 10-m height. Based on synoptic circulations at 700 hPa, wind shear events could be classified as strong northerly, convergence, southerly, and weak wind types. The proposed RDA successfully identified low-level wind shear except one southerly case, achieving 94% alerting rate compared with 82% for the operational system at ZLLL and 88% for the ramp detection algorithm (widely used in some operational alert systems) based on the same dataset. The RDA-unidentified southerly case occurred in a near neutral atmosphere, and wind speed change could not be captured by the Doppler lidar.

A comprehensive study of terrain‐disrupted airflow at Hong Kong International Airport – observations and numerical simulations

Complex terrain near the Hong Kong International Airport (HKIA) is likely to bring about low‐level windshear and turbulence to the aircraft, which could result in aviation hazards. In the past, some isolated case studies have been presented on the complicated airflow at the airport region in Hong Kong. This paper presents a more comprehensive discussion of the terrain‐induced airflow disturbances at the HKIA, including extensive mountain wake, transient low‐level jet stream, elongated mountain wake, reverse flow and mountain wave. The performance of a super high‐resolution numerical weather prediction model in the forecasting of these weather phenomena is also discussed. The model turns out to be capable of forecasting most of the terrain‐disrupted airflow, and it may be possible to obtain an earlier indication of the occurrence of low‐level windshear for airports with complex terrain. However, some mesoscale weather phenomena, such as low‐level jet stream, may not be captured by the models, and the associated windshear events would be difficult to forecast. This paper could hopefully be a useful reference to the aviation weather forecasters in the prediction of wind and windshear at airports with similar complex terrain in other parts of the world.

Airport low-level wind shear lidar observation at beijing capital international airport

Ocean University of China lidar team operated a pulse coherent Doppler lidar (PCDL) for the low level wind shear monitoring at the Beijing Capital International Airport (BCIA) in 2015. The experiment configuration, observation modes is presented. A case study shows that the low level wind shear events at the southern end of 18R/36L runway were mainly caused by the trees and buildings along the glide path under strong northwest wind conditions.

Wind shear over the Nice Côte d'Azur airport: case studies

Abstract. The Nice Côte d'Azur international airport is subject to horizontal low-level wind shears. Detecting and predicting these hazards is a major concern for aircraft security. A measurement campaign took place over the Nice airport in 2009 including 4 anemometers, 1 wind lidar and 1 wind profiler. Two wind shear events were observed during this measurement campaign. Numerical simulations were carried out with Meso-NH in a configuration compatible with near-real time applications to determine the ability of the numerical model to predict these events and to study the meteorological situations generating an horizontal wind shear. A comparison between numerical simulation and the observation dataset is conducted in this paper.

Aviation Model: A Fine-Scale Numerical Weather Prediction System for Aviation Applications at the Hong Kong International Airport

The Hong Kong Observatory (HKO) is planning to implement a fine-resolution Numerical Weather Prediction (NWP) model for supporting the aviation weather applications at the Hong Kong International Airport (HKIA). This new NWP model system, called Aviation Model (AVM), is configured at a horizontal grid spacing of 600 m and 200 m. It is based on the WRF-ARW (Advance Research WRF) model that can have sufficient computation efficiency in order to produce hourly updated forecasts up to 9 hours ahead on a future high performance computer system with theoretical peak performance of around 10 TFLOPS. AVM will be nested inside the operational mesoscale NWP model of HKO with horizontal resolution of 2 km. In this paper, initial numerical experiment results in forecast of windshear events due to seabreeze and terrain effect are discussed. The simulation of sea-breeze-related windshear is quite successful, and the headwind change observed from flight data could be reproduced in the model forecast. Some impacts of physical processes on generating the fine-scale wind circulation and development of significant convection are illustrated. The paper also discusses the limitations in the current model setup and proposes methods for the future development of AVM.