226 Resolving chaos in Madden-Julian oscillation

Monday, 29 January 2024
Hall E (The Baltimore Convention Center)
Guosen Chen, Nanjing University of Information Science & Technology, Nanjing, Jiangsu, China

The Madden-Julian Oscillation (MJO), the key variability of the tropical atmosphere, modulates the Monsoons and El Niños, and triggers extreme rainfall, heat waves, and hurricanes around the globe. Understanding the MJO dynamics is critical for fulfilling the “seamless” prediction that bridges the weather and climate.

Owing to its low-frequency phase oscillation, the MJO is often thought to be less chaotic and more predictable than the mid-latitude synoptic disturbances. The prevailing MJO theories are basically linear, neglecting the nonlinear effects. However, using robust chaos detection technique, we demonstrate that the MJO amplitude evolution is deterministic chaotic, signifying the importance of the nonlinear dynamics.

Combining delay embedding and Koopman operator theory, we show that the evolution of MJO amplitude can be split into periodic part that has discrete spectra and aperiodic part that has continuous spectra. We find that the discrete spectra and the related periodic obits construct the key components of the MJO amplitude evolution, while the continuous spectra, representing the chaotic behavior, is produced by mixing of the discrete spectra through nonlinear interaction. By comparing the discrete spectra, we further reveal the coupling between the MJO amplitude evolution and background moisture.

The results shed light on how to resolve nonlinear dynamics in climate system and provide a practical way for detecting potential nonlinear coupling between different subsystems.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner