Login
Username:
Password:
You are NOT Logged in.
Chat about all aspects of snowsports, backcountry, climbing and mountaineering.
Goto Thread: PreviousNext
Goto: Forum ListMessage ListNew TopicSearchLog In
alan


Posts: 8932
Joined: Nov 1994
Last Visited: 09:59
23rd May 2013
What's this?What's this?What's this?
Teleconnections, Sunspots and Scottish Snow
Date Posted: 13.32hrs on Wed 12 Oct 11
This explanation of the QBO on netweather got me thinking about these things and solar activity, esp with the media fuss about the Met Office's latest piece of research to appear in Nature Geoscience.

chionomaniac wrote:

The quasi-biennial oscillation (QBO) is a tropical stratospheric wind that travels easterly then westerly over a period for the full cycle of just over 2 years. You are right in suggesting that the westerly winds are known as positive and the easterly winds are negative. The pulse of westerly or easterly winds tend to commence in the upper regions of the tropical stratosphere and push downwards towards the troposphere. This can leave us in a state where the upper winds in the tropical stratosphere are westerly and the lower regions are easterly or vice versa depending on what stage of the QBO we are in. Sometimes these winds can push through and affect the polar stratosphere and other times they do not. The general reading for the state of the QBO mean winds is normally taken at the 30 hPA level.

Generally, it is thought that the stratospheric vortex polar is increased in a westerly QBO regime and decreased in an easterly regime. However the exact timing and positioning of these winds descending the stratosphere is going to have a big influence on polar regions.

Then we also need to take into consideration the Brewer Dobson circulation (BDC). This is the mechanism that transfers ozone from the tropical stratosphere to the polar stratosphere. Labitzke has shown that this mechanism is influenced by the state of the QBO and the amount of solar flux present. An easterly QBO in solar minimum conditions increases the BDC as does a westerly QBO in solar maximum conditions.

Increased polar ozone will lead to a warmer polar stratosphere and a reduced thermal gradient between the polar and tropical stratosphere. This in turn will lead to a reduced polar vortex and a more disrupted tropospheric jet stream.

Then we need to add in to the mix the ENSO state and how this interacts with the state of the QBO and tropical stratosphere and we begin to realise how complex all this is and that to just say a negative QBO is favourable for winter is just a little bit too simplistic!


The table below uses January, February, March to define winter in this case as the core months of the Scottish Snowsports Season. Abbreviations used:

QBO: Quasi Biennial Oscillation
AO: Arctic Oscillation
MEI: Multivariate ENSO Index
ENSO: El Nino Southern Oscillation

The season quality index is 0 to 5, 3 being average, 5 is the exceptional snow seasons like 2010, 2001, 1994 and 1979. 0 is those hall of shame for being right up there as the collection of years making up the worst seasons of all time.

This is of course rather subjective, thoughts on the season rankings welcome. Review of CairnGorm seasons in the Scottish Ski Club centenary journal contributed to season rankings for 1973 to 1992, along with other evidence and personal experience.

JFM	JFM	Season	JFM	JFM
Season	AO	QBO	Quality MEI	Sunspot Count
	                
1973	0.9 	-0.31 	1	 1.35 	48
1974   -0.3 	-1.10 	1	-1.83 	25
1975	0.6 	-15.04 	2	-0.67 	10
1976	0.8 	 9.28 	1	-1.42 	11
1977   -1.8 	-15.65 	4	 0.29 	16
1978   -1.0 	 6.21 	4	 0.87 	74
1979   -1.2 	 2.29 	5	 0.32 	148
1980   -1.5 	-6.22 	4	 0.65 	147
1981   -0.7 	 7.58 	2	 0.00 	130
1982	0.4 	-14.63 	0	-0.12 	143
1983   -0.3 	 11.47 	4	 2.88 	67
1984   -0.6 	-11.66 	4	-0.25 	75
1985   -1.2 	 3.38 	4	-0.63 	17
1986   -0.5 	 10.28 	5	-0.16 	13	Hale Winter
1987   -1.5 	-11.84 	5	 1.38 	9	Hale Cycle
1988   -0.3 	 6.61 	4	 0.75 	59	
1989	2.6 	-2.69 	0	-1.16 	152	
1990	2.5 	-4.83 	0	 0.56 	150	
1991   -0.2 	 8.72 	4	 0.33 	148	
1992	0.9 	-15.04 	0	 1.87 	139	
1993	1.5 	 10.60 	3	 0.88 	73	
1994	0.2 	-9.59 	5	 0.23 	42	
1995	0.6 	 8.39 	3	 1.01 	28	
1996   -0.8 	-7.54 	4	-0.50 	8	
1997	0.8 	 1.05 	2	-0.45 	8	
1998   -0.8 	-2.91 	0	 2.68 	42	
1999   -0.3 	 5.84 	4	-1.12 	66	
2000	0.6 	 4.85 	3	-1.17 	117	
2001   -1.1 	-15.74 	5	-0.60 	97	
2002	1.2 	 7.32 	2	-0.14 	114	
2003	0.2 	-2.04 	1	 0.98 	78	
2004   -1.0 	 1.07 	2	 0.18 	50	
2005   -0.8 	-0.66 	2	 0.71 	34	
2006   -0.6 	-10.15 	3	-0.49 	19	
2007	0.6 	 2.09 	2	 0.54 	11	
2008	0.8 	-4.98 	4	-1.33 	4	
2009	0.1 	 11.49 	3	-0.71 	2	Hale Cycle
2010   -2.4 	-17.56 	5	 1.35 	11	Hale Winter?
2011	0.4 	 9.89 	4	-1.58 	34


Scatter charts of teleconnection values against season index were, well scattered. The QBO one though showed up an interesting pattern:

 


Good seasons tend to have a well established QBO phase, near neutral/transitional condition of the QBO seems to favour poorer seasons (-5 to +5), though 1979 goes against this, indeed 1979 seems to be a total freak, proof that however various cycles might stack the odds, anything can still happen.

The MEI scatter chart is such a shambles it would seem to imply that ENSO state alone is not significant, yet we seem to have periods where El Nino is good for winter and periods where it's bad - how it interconnects with other factors?



Edited 1 times. Last edit at 13.38hrs Wed 12 Oct 11 by alan.

Attachments: qbo.png (30kB)  
Hustler


Posts: 676
Joined: Oct 2006
Last Visited: 11:58
22nd May 2013
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 16.31hrs on Wed 12 Oct 11
Aye, very good.
But when is it going to start snowing again....?

timstirling


Posts: 484
Joined: Feb 2003
Last Visited: 05:58
23rd May 2013
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 22.35hrs on Wed 12 Oct 11
Seems like there should be a stronger connection with the AO. Most good winters have a negative (or near neutral) AO, and most negative AO have a good winter. Vice versa for positive AO.

moffatross


Posts: 1319
Joined: Mar 2006
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 22.56hrs on Wed 12 Oct 11
^

Aren't the AO's getting confused with the NAO's ? In a negative NAO phase, there is higher than normal pressure over Greenland and lower than normal pressure over the Azores if that makes sense. It follows that a winter with a negative NAO (North Atlantic Oscillation) would be cooler than a positive NAO because air will more often tend to be flowing from the east or north.

I think that the 'Arctic Oscillation' value on the other hand is related to the spin and position of a polar vortex or bits of it, i.e. it's much more difficult to pin down what, if anything a notional -'ve or +'ve value means to anything 'weathery' because a polar vortex is kind of self containing and doesn't really share itself with the generally westerly flow in either of our hemispheres.

Oops, edited for muppetry. smiling smiley






Edited 3 times. Last edit at 23.05hrs Wed 12 Oct 11 by moffatross.

timstirling


Posts: 484
Joined: Feb 2003
Last Visited: 05:58
23rd May 2013
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 23.05hrs on Wed 12 Oct 11
yeah, look like Alan posted NAO values not AO, so the correlation makes sense.

alan


Posts: 8932
Joined: Nov 1994
Last Visited: 09:59
23rd May 2013
What's this?What's this?What's this?
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 23.31hrs on Wed 12 Oct 11
When the AO is positive pressure is lower over the polar regions, when the AO is negative pressure is higher, so a negative AO is associated with northern blocking. The values posted in the table above are AO 3 monthly averages not NAO figures.

moffatross


Posts: 1319
Joined: Mar 2006
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 00.55hrs on Thu 13 Oct 11
^
Never really thought about that but your explanation kind of makes sense now as another 'index' of stuff that's happening that affects winter synoptics. Are folk thinking of the AO as a symptom of or as a cause ?

timstirling


Posts: 484
Joined: Feb 2003
Last Visited: 05:58
23rd May 2013
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 09.54hrs on Thu 13 Oct 11
I apologies, then he AO looks quiet good as a predictor.

Looking at god winters (4-5), 13 have a negative AO while 3 have a positive (including one at only +0.2, 1994).

Looking at bad winters (0-1), only 2 have a negative AO, while 7 have a positive.

In the middle, winters rated 2 are pretty even negative and positive. Winters rated 3 have 1 negative AO, 1 just positive (0.1) and 2 positive, so a little mixed. One must note that although 3 is claimed to be average, there are actually very few winters rated at 3 (winters with a rating of 3 are the the equal least likely category) and there are more 5 winters than 3 winters. So this I would put down to noise.

The 2 most positive AO winters had a score of 0, while the 7 most negative AO winters had ratings 4 to 5.

A nice result is that when the AO is negative there is very low chances of a poor winter (0-1) with only 2 occurrences,but when it is moderately positive there are still chances for an average to good winter (3-5) with 6 occurrences


The AO and the winter ratings have a correlation coefficient of r= -0.6, i.e. a negative AO correlates with a higher winter rating, which should explain about 36% of the winter ratings.
With r =-0.6, we can get a student's t value of t=-4.623 , df = 38, and a highly significant p value p <.0001
The caveat being that categorical ratings tends to produce skewed statistics but this should be statistically sound to a certain level.





Edited 2 times. Last edit at 10.02hrs Thu 13 Oct 11 by timstirling.

timstirling


Posts: 484
Joined: Feb 2003
Last Visited: 05:58
23rd May 2013
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 10.20hrs on Thu 13 Oct 11
The QBO has a linear correlation of r=0.04.
The MEI Enso has a linear correlation of r=-0.05

The sunspot count has a linear correlation of r=-0.26, i.e lower sunspot counts tend to give better winters, which makes sense.

I am tying to think of ways of combing the AO and sunspot predictors, all I have so far is a function AO*SQRT(SUNSPOT) gives a correlation of r=-0.61.

The thing is, AO correlates with the sunspot number with r =0.26 (more sunspots tend to give positive AO winters). SO I wonder if simply the sunspots counts have an affect on the AO which affects the winter ratings, so incorporating sunspot information into the AO doesn't increase forecast reliability very much.

Anyone know of a reliable AO forecast systems???

alan


Posts: 8932
Joined: Nov 1994
Last Visited: 09:59
23rd May 2013
What's this?What's this?What's this?
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 11.41hrs on Thu 13 Oct 11
Short to medium term AO forecasts tend to be more reliable in the 7 to 10day window than surface level synoptics, but the lack or regular periodicity of the AO makes forecasting it beyond difficult. Unlike the QBO which has a fairly regular oscillation, this leads to the question Ross asked above frequently coming up, is the AO a cause or a symptom!

I wonder from what I posted above if the quasi 22year hale cycle is more important than the 11 year solar cycle in terms of effect of solar maximum? Late 70s to start of the 80s a strong solar max brought mostly good seasons. Less said about end of the 80s/early 90s the better... so it sort of looks like good max - bad max - good max.... but clearly the approaching cycle 24 maximum looks very low by modern standards.

Purely from the visual spread on the scatter chart, all of the worst seasons had a QBO value below 0, extremely Easterly QBO seems to increase the likelyhood of seasons tending to one extreme or the other.

In 1982 and 1992 an extreme easterly (negative) QBO value at or near the peak of strong solar cycles give appalling seasons (the two points on the 0 line about -15 on the chart), which fits with what was quoted by Chionomaniac from netweather above.


daveski


Posts: 1315
Joined: Jan 2008
Last Visited: 07:19
23rd May 2013
What's this?What's this?What's this?
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 12.03hrs on Thu 13 Oct 11
So is this season going to be good, bad or somewhere in between

timstirling


Posts: 484
Joined: Feb 2003
Last Visited: 05:58
23rd May 2013
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 12.46hrs on Thu 13 Oct 11
I agree that from scatter plot negative QBO seems to suggest more extremes, neutrial QBO worse winters, positive QBO mostly average to good. But there isn't so many data points in much of this, coincidentally there is relatively few data points for the average and terrible seasons to make strong assertions.

I think part of the complexity in using such predictors is that you don't always get positive outcomes from potentially good synoptics and winter pattersn. E.g. a Strongly negative NAO winter may have far less zonal weather, but a cold dry winter can be useless. Similarly, I believe 1994 was zonal, but of a the good kind with temperatures just low enough for mountain snow. So although something like the AO or QBO may have good correlations small changes in the resulting synoptic can radically change the final outcome.


As for ENSO, we are told that La Nina tends to make colder early winter conditions with milder later winters, the opposite for El Nikon. Last winter and 1998 seems to support this. Maybe it would be enlightening to look at a Dec-Jan and Feb-March winter rating separately for ENSO conditions.

alan


Posts: 8932
Joined: Nov 1994
Last Visited: 09:59
23rd May 2013
What's this?What's this?What's this?
Re: Teleconnections, Sunspots and Scottish Snow
Date Posted: 00.25hrs on Sat 11 Aug 12
JFM 2012 registered a QBO avg value of -16.0, having made the transition 3-4 months later in 2011 than was the case in 2009. Though April and May delivered fantastic conditions to CairnGorm Mountain, given we saw the first time with no lift served riding in Scotland in March since commercial snowsports begun, it's probably fair to rate the JFM period of the 2012 season as 0 on the graph above. In which case there is a third point on the 0 line for a strongly negative QBO, while some of the best seasons of all time including 2010 also have a cluster around a strongly negative QBO.

Obviously the graph above doesn't consider the time lag from the QBO becoming negative till the core of the snowsports season, the 7 month lag to the freezer suddenly been switched on in late Dec 2009 and again in April 2012 is curious! Anyone fancy a stap at a long range forecast? tongue sticking out smiley

Your Name: 
Your Email: 
Subject: