A networked station system for high-resolution wind nowcasting in air traffic operations: A data-augmented deep learning approach

• DOI: 10.1371/journal.pone.0316548
• Publication Date: 2025-01-14T18:35:59Z
• Authors: Décio Alves; Fábio Mendonça; Sheikh Shanawaz M...; Diogo Freitas; João Pestana; Dinarte Vieira; Marko Radeta; Fernando Morgado-...
• Keywords: Nowcasting

This study introduces a high-resolution wind nowcasting model designed for aviation applications at Madeira International Airport, location known its complex patterns. By using data from network of six meteorological stations and deep learning techniques, the produced is capable predicting speed direction up to 30-minute ahead with 1-minute temporal resolution. The optimized architecture demonstrated robust predictive performance across all forecast horizons. For most challenging task, forecasts, achieved Mean Absolute Error (MAE) 0.78 m/s MAE 33.06°. Furthermore, use Gaussian noise concatenation both input label training yielded consistent results. A case further validated model's efficacy, values below 0.43 between 33.93° 35.03° different approach shows that combining strategically deployed sensor networks machine techniques offers improvements in airports environments, possibly enhancing operational efficiency safety.