Remember the concern over the QBO anomaly/disruption during 2016?
Quite a few papers were written on the topic
- Newman, P. A., et al. "The anomalous change in the QBO in 2015–2016." Geophysical Research Letters 43.16 (2016): 8791-8797.
Newman, P. A., et al. "The Anomalous Change in the QBO in 2015-16." AGU Fall Meeting Abstracts. 2016.
Randel, W. J., and M. Park. "Anomalous QBO Behavior in 2016 Observed in Tropical Stratospheric Temperatures and Ozone." AGU Fall Meeting Abstracts. 2016.
Dunkerton, Timothy J. "The quasi‐biennial oscillation of 2015–2016: Hiccup or death spiral?." Geophysical Research Letters 43.19 (2016).
Tweedy, O., et al. "Analysis of Trace Gases Response on the Anomalous Change in the QBO in 2015-2016." AGU Fall Meeting Abstracts. 2016.
Osprey, Scott M., et al. "An unexpected disruption of the atmospheric quasi-biennial oscillation." Science 353.6306 (2016): 1424-1427.
Note that the training region for the model is highlighted in YELLOW and is in the interval from 1978 to 1990. This was well in the past, yet it was able to pinpoint the sharp peak 27 years later.
The disruption in 2015-2016 shown with shaded black may have been a temporary forcing stimulus. You can see that it obviously flipped the polarity with respect to the model. This will provoke a transient response in the DiffEq solution, which will then eventually die off.
The bottom-line is that the climate scientists who pointed out the anomaly were correct in that it was indeed a disruption, but this wasn't necessarily because they understood why it occurred — but only that it didn't fit a past pattern. It was good observational science, and so the papers were appropriate for publishing. However, if you look at the QBO model against the data, you will see many similar temporary disruptions in the historical record. So it was definitely not some cataclysmic event as some had suggested. I think most scientists took a less hysterical view and simply pointed out the reversal in stratospheric winds was unusual.
I like to use this next figure as an example of how this may occur (found in the comment from last year). A local hurricane will temporarily impact the tidal displacement via a sea swell. You can see that in the middle of the trace below. On both sides of this spike, the tidal model is still in phase and so the stimulus is indeed transient while the underlying forcing remains invariant. For QBO, instead of a hurricane, the disruption could be caused by a SSW event. It also could be an unaccounted-for lunar forcing pulse not captured in the model. That's probably worth more research.
As the QBO is still on a 28 month alignment, that means that the external stimulus — as with ENSO, likely the lunar tidal force — is providing the boundary condition synchronization.