The curious case of QBO cycle 30

[xmetman]

Last month I wrote a simple application to download and graph monthly quasi-biennial oscillations [QBO] values from Physical Sciences Division [PSD]. I wrote an article about it, and that has more details about what it the QBO is and why it’s important. The values for September 2016 have just been released , and cycle thirty of the QBO (as far as my list of cycles is concerned) is still stuck in its westerly phase that started in June 2015. The problem with that is that the QBO should have flipped and be well into it’s E’ly phase by now. The problem started in April 2016 when the QBO index suddenly increased when it should have been decreasing and about to enter its E’ly phase, and that increase has continued till now September 2016, although it does look like the levels of increase has flattened out. Although the QBO has usually behaved itself since it was identified with a quite regular period of 28 or 29 months, this last cycle has gone a little haywire, and it’s now in its thirtieth month.

[Scott Whitehead]

Is it no longer a reliable indicator for long range forecasting? I believe it has only been used since 1950 which is no timescale in the grander scheme of things.

[Alastair B. McDonald]

There are two peaks  for each of the four phases you show. The last one is just a bit more extreme. As I wrote earlier could it not be that there are two cycle runing, e.g. one in the Northern Hemisphere and the other in the Southern Hemisphere.

[xmetman]

No, the QBO is purely an equatorial phenomena as far as I understand it. I looked at the length of the cycle in the previous article I wrote, an there has been nothing that resembles the double peak as in this one.

[Alastair B. McDonald]

Here is a picture that can look quite different depending on how you see it, either as a young woman or and old hag.



It is the same with your graph. picture. It can look like one cycle, which you have emphasised with the yellow graph, or it can be seen as two cycles slightly offset and overlapping. There was a double peak on the last cycle, but what about the  previous one?

[xmetman]

Alistair

I can see your point. Here are the last few cycles and there was a tendency in most of the westerly peaks to fall back slightly but not on the scale of cycle 30.

Bruce.

[Alastair B. McDonald]

Hi Bruce,

I have just found this: http://www.metoffice.gov.uk/research/news/2016/quasi-biennial-oscillation-disruption

It seems that I have been fooled by your charts only being relevant at one height (30 mb). The overall picture picture is slightly different, with a normally longer period of eastward winds, in this case breaking up into two phases.