Accurate and real-time wind field data play a crucial role in ensuring the safety of civil aviation flights. In addressing the precise reconstruction of wind fields, this paper proposes a method based on aircraft monitoring data. The approach aims to utilize joint observations of Automatic Dependent Surveillance–Broadcast (ADS-B) and Mode-S Enhanced Surveillance (Mode-S EHS) to calculate wind observation values in the airspace. By integrating a Gaussian Process Regression model from machine learning and utilizing temporally and spatially discrete wind observation values for model training, the method achieves a complete reconstruction of the target airspace wind field. Experimental results demonstrate that the average absolute error of wind speed reconstructed is 2.72 meters per second, with a relative error of 8.21%, and the average absolute error of wind direction is 3.66 degrees. This validates the method's capability to rapidly and accurately reconstruct wind fields in real-time.