Prediction of a Pilot’s Invisible Foe: The Severe Low-Level Wind Shear

Abstract

Severe low-level wind shear (S-LLWS) in the vicinity of airport runways (25 knots or more) is a growing concern for the safety of civil aviation. By comprehending the causes of S-LLWS events, aviation safety can be enhanced. S-LLWS is a rare occurrence, but it is hazardous for approaching and departing aircraft. This study introduced the self-paced ensemble (SPE) framework and Shapley additive explanations (SHAP) interpretation system for the classification, prediction, and interpretation of LLWS severity. Doppler LiDAR- and PIREPs-based LLWS data from Hong Kong International Airport were obtained, trained, and evaluated to predict LLWS severity. The SPE framework was also compared to state-of-the-art tree-based models, including light gradient boosting machine, adaptive boosting, and classification and regression tree models. The SPE does not require prior data treatment; however, SMOTE-ENN was utilized to treat highly imbalanced LLWS training data for tree-based models. In terms of prediction performance, the SPE framework outperforms all tree-based models. Using SHAP analysis, the SPE was interpreted. It was determined that “runway 25LD”, “mean hourly temperature”, and “mean wind speed” were the most significant contributors to the occurrence of S-LLWS. The most optimistic projections for the occurrence of S-LLWS events at runway 25LD were during periods of low-to-moderate temperatures and relatively medium-to-high wind speeds. Similarly, the majority of S-LLWS events took place on the runway. Without the need for data augmentation during preprocessing, the SPE framework coupled with the SHAP interpretation system could be utilized effectively for the prediction and interpretation of LLWS severity. This study is an invaluable resource for aviation policymakers and air traffic safety analysts.

Intrinsic Explainable Self-Enforcing Networks Using the ICON-D2-Ensemble Prediction System for Runway Configurations

“…Despite the greatest efforts in recent years to improve weather forecasts, the uncertainties that still exist must be considered [2,3]. In addition, new challenges arise from climatic changes, extreme weather conditions, strong wind shear at low altitudes [4,5], lateral boundary perturbations [6], or general hazardous meteorological conditions [7] that require variation or extension of previous models [8][9][10].…”

“…A big challenge remains regarding the explainability of the results in the context of ATM [5,[25][26][27]. According to Degas et al [28], the problem for responsible end users in ATM is that the results are difficult to understand or are not transparent for safety-critical areas such as air traffic.…”