Introduction
Terminal Control Areas (TMAs) serve as the critical transition zones between high-altitude cruise and airport approach/departure phases. As global air traffic volume intensifies, these areas have become significant operational bottlenecks. Traditional methods of analyzing terminal airspace often rely on 2D flow patterns or entire trajectory paths, which can obscure the underlying reasons for air traffic controller (ATC) interventions.
This research proposes a shift in perspective: instead of analyzing continuous paths, the study focuses on "characteristic points"—specific locations where pilots are instructed to change heading, altitude, or speed. By clustering these intervention points, the authors provide a data-driven map of operational hotspots and controller workload.
Methodology
The proposed analytical framework follows a multi-stage pipeline:
- Trajectory Reconstruction: Raw radar data is filtered and converted into a 3D Cartesian coordinate system (UTM).
- Characteristic Point Extraction: The system identifies three primary types of intervention:
- Heading Transitions: Turns exceeding 5 degrees.
- Altitude Transitions: Start and end points of level flight segments during climbs or descents, indicating a controller held an aircraft at an altitude to resolve a conflict.
- Speed Adjustments: Velocity changes greater than 20 km/h.
- Clustering and Optimization: These 3D coordinates are normalized and grouped using the K-means algorithm. To determine the most effective number of clusters, the researchers employed the "Elbow Method" based on the Sum of Squared Errors (SSE).
As shown in the SSE analysis, K=7 was identified as the optimal number of clusters for the dataset, balancing model complexity with descriptive power.
Dataset Analysis
The study utilized a comprehensive dataset from the Northwest Air Traffic Management Bureau of China, covering a full day of operations.
The raw data comprised 403 flight tracks and over 545,000 individual trajectory points. By filtering for flights below 4,500 meters and extracting only characteristic points, the researchers reduced the data to 3,622 critical intervention points. This significant data compression allows for efficient analysis without losing the operational context of the airspace.
Key Insights and Results
The clustering analysis revealed several critical aspects of terminal area operations:
- Workload Hotspots: Clusters 1 and 4 were identified as "Airport Control" zones. Cluster 1, despite its smaller spatial footprint, exhibited a much higher speed adjustment ratio, indicating a zone of intense controller activity and complex traffic management.
- Spatial Correlation: A strong negative correlation (-0.94) was found between turning points and altitude adjustments. In areas where horizontal maneuvering is restricted, controllers rely more heavily on vertical separation (leveling off aircraft) to maintain safety.
- Airspace Asymmetry: The analysis highlighted a significant imbalance, with the southern sector of the terminal area experiencing more frequent conflicts and interventions than the northern sector, suggesting a need for route redesign.
Conclusion
By focusing on points of intervention rather than entire trajectories, this method provides a proxy for air traffic controller workload and airspace complexity. The ability to visualize these "clouds" of intervention allows air traffic managers to identify bottlenecks and optimize flight procedures based on actual operational behavior rather than theoretical models. Future work will aim to incorporate aircraft types and weather conditions to further refine these performance evaluations.
Summary updated