We introduce a transformer-based AI model for estimating Doppler velocity from incoherent scatter radar (ISR) spectra. Inspired by Vision Transformers, the model uses a standard transformer encoder adapted for radar data. Trained solely on simulated spectra, it performs well on real data from the Arecibo radar and significantly outperforms traditional least-squares fitting (LSF) methods. This approach is broadly applicable wherever spectral data can be parameterized.

We present a novel method to measure the amplitudes of traveling quasi-5-day oscillations (Q5DOs) in the middle atmosphere during sudden stratospheric warming events based on satellite observations. Simulations and observations demonstrate that the previously reported traveling Q5DOs might be contaminated by stationary planetary waves (SPWs). The new fitting method is developed by inhibiting the effect of a rapid and large change in SPWs.

Using radar observations and reanalysis data for 9 years, we demonstrate clearly for the first time that resonant interactions between tides and annual and semiannual oscillations do occur in the mesosphere and lower thermosphere. The resonant matching conditions of frequency and wavenumber are exactly satisfied for the interacting triad. At some altitudes, the secondary waves are stronger than the tides, thus in tidal studies, the secondary waves may be mistaken for the tides if no carefully.

El Niño has an important influence on climate systems. There are obviously negative water vapor anomalies from radiosonde observations in the tropical western Pacific during El Niño. The tropical Hadley, Walker, and monsoon circulation variations are revealed to play different roles in the observed water vapor anomaly in different types of El Niños. The Walker (monsoon) circulation anomaly made a major contribution in the 2015/16 (2009/10) strong eastern Pacific (central Pacific) El Niño event.