Wind vector and temperature information are important in civil aviation. Accurate wind and temperature information facilitates flight planning and ensures economic and safe flights. Accurate forecasting of future weather data is also useful for determining flight altitude and speed, planning take-off and landing routes, reducing flight delays and cancellations, and helping airports to provide more efficient services. This paper proposes a method for inverse reconstruction and prediction of meteorological information in a region based on aircraft monitoring data, using decoded Automatic Dependent Surveillance-Broadcast (ADS-B) and Mode-S Enhanced Surveillance (Mode-S EHS) data to solve for wind vectors and temperature. Meanwhile, the traditional Transformer model is also been improved to predict future weather information based on historical data exactly. The method presented in this paper has been compared and validated to improve the resolution and real-time performance of the weather inversion and to give effective forecasts of future weather, complementing the existing technology.
|