An Observation of the Trapped Lee Waves Using Fengyun-4 Satellite Images and a U-Net Network

Abstract

Propagating in the lower troposphere, the trapped lee waves (TLWs) generate severe turbulence over a large area in the afternoon and thus have a significant impact on aviation safety and atmospheric environment. Due to the lack of high temporal and spatial resolution observations, the diurnal variation of the TLWs remains unknown. This study employed the U-Net deep learning model to identify 3,533 images with TLWs from over thirty thousand 500-m resolution Fengyun-4 satellite images at intervals of a few minutes during the cold season from 2020 to 2023. The results show that the wavelength increases in the afternoon (from 12:00 to 16:00 LST) with high low-level winds and low atmospheric stability while the amplitude and propagation area continually increase during the daytime (from 09:00 to 16:00 LST) as the boundary layer becomes turbulent. These findings explain environmental disasters related to severe turbulence in the afternoon over the mountainous region.