And we begin

[GaWx]

Thanks, Dave. The recent pattern of more blocking in October across Greenland and the Arctic since 2002

seems to favor this teleconnection pattern leading to greater Eurasian snow cover.

comp.png

Bluewave,

Just for the sake of being the devil's advocate (since I still largely believe in Cohen's idea) and taking your point one step further, how much of the correlation of DJF -AO with October Eurasian SAI/SCE could instead be mainly just a simple connection with a strongly -AO October?

[Isotherm]

Bluewave,

Just for the sake of being the devil's advocate (since I still largely believe in Cohen's idea) and taking your point one step further, how much of the correlation of DJF -AO with October Eurasian SAI/SCE could instead be mainly just a simple connection with a strongly -AO October?

 

 

I don't have the SAI/SCE statistics in front of me, but thinking about it meteorologically, it wouldn't be surprising that a -AO October is strongly correlated to high SAI/SCE and thus an increased likelihood for the ensuing DJF -AO. A predominance of higher the normal heights near the pole tends to promote more amplified troughs for 30-60N latitude, which in turn would be liable to cause significant snow cover increase. Alternatively, moderate to strongly +AO Octobers would feature a tight jet circumnavigating the north pole and thus little in the way of snow cover advance south of 60N latitude. The composite posted above by bluewave would certainly be more likely to occur in a -AO October than a +AO one. I'd wager to say there were still some good SAI/SCE years in a +AO regime, as nothing in meteorology correlates perfectly as you know, but the majority of high SAI/SCE years probably featured a -AO to some extent, with the greater magnitude -AO episodes being more efficient in expanding the southern periphery of that snow cover. Therefore I don't think it's shocking that modestly negative -AO Octobers like 1968, 1976, 2003, 2009, etc., (and now this year) all shared the common denominator of a high-magnitude AO plummet during the month and rapid snow increase. Once you lay down a ton of snow cover via a strong -AO pulse, particularly if it occurs in the middle to latter part of October, there's not going to be enough time for much of it to melt by October 31st.

 

So overall, the science seems to support rooting for -AO October if you're hoping for a better opportunity for high SAI/SCE and thus and ensuing winter negative AO. I'm sure there will be exceptions however.

[griteater]

All very logical Isotherm.  Also, you want to see the the high latitude blocking focused in the eastern hemisphere, stretched out across northern Eurasia, which forces the storm track south in Eurasia as you mentioned.  Posted here is a composite of the high snow advance years (SAI).

 

[hosj III]

All very logical Isotherm.  Also, you want to see the the high latitude blocking focused in the eastern hemisphere, stretched out across northern Eurasia, which forces the storm track south in Eurasia as you mentioned.  Posted here is a composite of the high snow advance years (SAI).

 

If you look at the larger sample of all SAI years, the most important feature is the ridge in north central siberia.

[The_Global_Warmer]

 

 

Apparently, there is.

 

 

Next question is there a correlation to increased moisture availability from the Arctic Ocean during October.

 

The 2014 graphic is likely skewed from the last 10 days of October not being averaged in.  Never the less it;s quite a huge change. 

 

 

I would think it would be practical during the first couple weeks of October when you have large vortex's moving across the Northern Russian coastline drawing a long wind fetch over that open water around it's backside down into Russia that it would be a more efficient precip producer versus that wind crossing mostly much drier and colder ice.

 

 

But since it's already so cold up there is might not be much of a difference. If it is we could be in store for more large October snows over Eurasia.

 

 

[DaculaWeather]

It is cold. 

http://www.daculaweather.com/4_world_extremes.php?extremes=North

 

 

 

[chionomaniac]

If you look at the larger sample of all SAI years, the most important feature is the ridge in north central siberia.

And that fits in well with this study, which suggests that decreased storm tracks across the Barents Sea/ Taymyr Peninsular and Laptev Sea during October will have a higher correlation with a negative winter AO. It has been termed the Taymyr Circulation anomaly or TCA. It is another variant as far as I can see on Cohen's work - the storm tracks will be further south depositing the October snow in the critical SAI south of 60ºN area.

 

http://onlinelibrary.wiley.com/doi/10.1002/joc.3968/abstract

 

This year (so far) there is a weak to moderate positive 500hPa anomaly in the region though I note that the last few days of the month may counter this.

 

[hosj III]

And that fits in well with this study, which suggests that decreased storm tracks across the Barents Sea/ Taymyr Peninsular and Laptev Sea during October will have a higher correlation with a negative winter AO. It has been termed the Taymyr Circulation anomaly or TCA. It is another variant as far as I can see on Cohen's work - the storm tracks will be further south depositing the October snow in the critical SAI south of 60ºN area.

 

http://onlinelibrary.wiley.com/doi/10.1002/joc.3968/abstract

 

This year (so far) there is a weak to moderate positive 500hPa anomaly in the region though I note that the last few days of the month may counter this.

 

compday.Jk6UpninKv.gif

From the OPI thread, here is a composite of -OPI and +OPI Octobers. The ridging vs troughing in the area you mentioned appears to be the main distinction between the two, although the Aleutian low/+NAO are also correlated negativly with the OPI and subsequent DJF AO.

[GaWx]

 I'm still anxiously awaiting the next Natice/Rutgers updated maps and hope there is something by 10/31!

 

 In the meantime, although there has very likely been impressive SCE gains over the last few days and models suggest some more rather impressive snow in some new areas for the next two days (mainly over E Asia as W Asia and E Europe look about done for new areas from what I see), the last of the sig. bare areas to get sig. new snowcover looks to be ~tomorrow per model consensus. I'd really love to see the SCE for 10/27 since I'm guessing it would be near the max of the month. So, I see little new sig. SC aditions 10/28-31 per models. That would mean that 10/28-31 will be subject to net drops due to net melting of non-thick snowcovered areas (mainly over sub 55N areas) with some warming projected in many areas. So, my educated guess is that the peak for this month will be ~10/27-8 with a net drop thereafter into 10/31. However, I doubt that any net drop in SCE for 10/28-31 will be really large. My guess is that any net drop wouldn't exceed 1 msk for full Eurasian SCE. It may even turn out to be only ~0.5 msk or even lower. We'll see. Therefore, I still see the high potential for the full October 2014 SAI to be one of the highest on record and, therefore, to be quite promising for a strong -AO this winter. I'm still expecting to see cfbaggett's next update to make quite a statement assuming that updated data would finally let the poor guy do his thing.

 

Edit: Upon further review, the consensus has what appears to be new snowcover over a portion of northern China ~10/30. It would appear to cover as much as ~~0.5 msk. So, if that occurs, that would probably cancel out at least a good portion of the losses in other areas for that period.

[Absolute Humidity]

Curious if there is a way to extrapolate the current msk from these nasdis noaa imagery with the Rutgers data before the outage? At any rate, looks like some nice gains below 50 yesterday.

[weatherdude]

Seems like the difference between reanalysis data and the Rutger's SCE might be in the Rutgers 42% coverage rule..?? (if this has been addressed on here somewhere, please accept my apologies):

 

From Rutgers:

 

Daily Products
Daily snow charts were generated by degrading 1024 x 1024 NOAA Interactive Multisensor Snow and Ice Mapping System (IMS) charts to 128 x 128 resolution. Each 128 x 128 grid cell is considered snow covered if 42 percent or more of the IMS cells within it are snow covered. Using charts produced on Mondays, this value most closely approximates the weekly snow product based on mean SCE area comparisons. An 89 x 89 cell subset matching the National Meteorological Center Limited-Area Fine Mesh grid is then charted, representing the northern hemisphere.

[Mike.Ventrice]

Hey Mike,

I got the same number of snow-covered points and coverage ratio using the GFS data. In order to get a percentage coverage, I think you need to calculate the ratio of the area of all the cells associated with a snow-covered point and the total area of all the cells. It looks as if you are using NCL which has a function 'area_global_rectilinear_grid(-lat,lon,False) ' in the 'shea_util.ncl' script. For some weird reason, I had to put a minus sign in front of the lattitude array to get a positive area.

FYI, It's also more efficient to use the 'mask' function to find the number of snow-covered points than to loop through all the lat/lon points.

snow_cover_gfsw1.png

There does seem to be some disparity between the model data and the satellite obs, for whatever reason.

Thank you for doing this and pointing out some NCL functions to make the process easier! Much appreciated to hear that we came to the same answer using different methodologies. The post from Weather Dude may be an explanation of the discrepancies in our analysis verses Rutgers data.

[millwx]

Nice plot from the Canadian Cryospheric Information Network.  No surprises, but nice to see a plot:

 

[weatherdude]

Nice plot from the Canadian Cryospheric Information Network.  No surprises, but nice to see a plot:

 

BEAUTY!  Thanks.

[griteater]

Here's the latest plot from the NESDIS site.  16.07 on Oct 26 is down from the 16.70 reading on Oct 23

 

[GaWx]

Here's the latest plot from the NESDIS site. 16.07 on Oct 26 is down from the 16.70 reading on Oct 23

Until I see Rutgers update, I won't feel we have a true apples to apples comparison to old Rutgers numbers. The 16-17 msk range for Eurasian SCE would be very disappointing if it were a Rutgers number since I've been targeting 19+. However, the good news is that it appears that the Rutgers numbers use a different scale that would yield a good bit higher number than the 16-17 range. Regardless, the net drop of 630,000 sk on the Nesdis graph from 10/23 to 10/26 is, itself, a bit disappointing as that would imply close to a 750,000 drop on Rutgers. Hopefully, the SCE doesn't crap out from now til 10/31. I am admittedly a little concerned as per previous posts as I don't see a whole lot of new snowcovered areas between now and 10/31 outside of some today over E Asia and some ~10/30 in a portion of N China. I've been expecting a net drop around now for a few days. The amount of the drop is the question.

[das]

Here's the latest plot from the NESDIS site.  16.07 on Oct 26 is down from the 16.70 reading on Oct 23

 

 

Yup, a fair amount of melting on the western flank yesterday.

 

[GaWx]

Watch the SCE/SAI this last week of October. While it has been looking about as good as one could want, some rather significant losses apparently occurred between 10/23 and 10/26 per Nesdis data. Whereas losses over recent days would not be a shocker due to reduced new snows, I honestly didn't expect that much of a drop that soon. I had been thinking the peak would be near 10/27-8 followed by a fairly modest drop 10/28-31. Hopefully, we're not going to end up with a full crapout for the 10/23-31 period as that could potentially turn things from fantastic to mediocre in a hurry. Regardless, the Rutgers data is what I've been using. It has been out since the last update of 10/19! Until that is restored, I won't feel too confident about where Eurasian SCE is in relation to what I had been seeing as well as past years.

Now, Millwx may not care so much about 10/23-31.

[The_Global_Warmer]

Wouldn't it be a bit to late now for the edges to really factor that much?

 

Never the less there will probably be some melting today in parts of Southern Russia where WAA and rain is taking place but it looks like a cold front sweeps thru the area over the next couple days.

 

 

 

 

 

The Western front on the other hand is being bombarded with pretty decent winds and waa.

 

 

 

 

 

 

 

 

Looking at the snow depth chart Mike Ventrice posted you would think there will be a pretty big drop this week.

 

 

But again I can't see how that could really effect the atmospheric processes already in play.

[GaWx]

Wouldn't it be a bit to late now for the edges to really factor that much?

 

Looking at the snow depth chart Mike Ventrice posted you would think there will be a pretty big drop this week.

 

 

But again I can't see how that could really effect the atmospheric processes already in play.

 

Friv,

 Nope, it isn't too late. Although unlikely, you could lose as much as 2-3 msk within ~8 days per history over Eurasia, especially when you get way more than normal SCE south of 60N since it would be more vulnerable to meltback in October. It can be sort of a situation where the SCE got too far ahead of itself in a way. See 1978 and 1976 for examples. In 1976, the meltback was mainly in the first half of November. So, it didn't affect the Oct. SAI/SCE much. If an 8 day period of late Oct. were to be dominated by crappy conditions (little new snow/warming/rain falling over SC), a great SCE/SAI can certainly be reduced to a more mediocre SCE/SAI. Regarding SAI: keep in mind that 2014 started off with a handicap of sorts, a much higher than average 9/30 SCE. For 2014 to get a great SAI, we can't afford a crappy 10/24-31.

The SAI and SCE look at October as a whole, not just 10/1-23. If one is going to use those indices to forecast the DJF AO, then one would best not fudge the indices just because late Oct. sucked imo if one wants to have the highest confidence in using this data.

 Also, how do you know it couldn't really affect the atmospheric processes already in play? If the last 8 days of October don't matter, should the SAI be changed to be based on only, say the first 23 days?

 

Edit: I realize MillWx wouldn't care as much about this period. However, that's beside the point.

[ORH_wxman]

Friv,

 Nope, it isn't too late. Although unlikely, you could lose as much as 2-3 msk within ~8 days per history over Eurasia, especially when you get way more than normal SCE south of 60N since it would be more vulnerable to meltback in October. It can be sort of a situation where the SCE got too far ahead of itself in a way. See 1978 and 1976 for examples. In 1976, the meltback was mainly in the first half of November. So, it didn't affect the Oct. SAI/SCE much. If an 8 day period of late Oct. were to be dominated by crappy conditions (little new snow/warming/rain falling over SC), a great SCE/SAI can certainly be reduced to a more mediocre SCE/SAI. Regarding SAI: keep in mind that 2014 started off with a handicap of sorts, a much higher than average 9/30 SCE. For 2014 to get a great SAI, we can't afford a crappy 10/24-31.

The SAI and SCE look at October as a whole, not just 10/1-23. If one is going to use those indices to forecast the DJF AO, then one would best not fudge the indices just because late Oct. sucked imo if one wants to have the highest confidence in using this data.

 Also, how do you know it couldn't really affect the atmospheric processes already in play? If the last 8 days of October don't matter, should the SAI be changed to be based on only, say the first 23 days?

 

Edit: I realize MillWx wouldn't care as much about this period. However, that's beside the point.

 

 

I would think 2014 is going to have a very high SCE even if we have a lackluster final week. It started high and then went even higher.

 

SAI is more of a wildcard if the final week is bad enough.

[GaWx]

I would think 2014 is going to have a very high SCE even if we have a lackluster final week. It started high and then went even higher.

 

SAI is more of a wildcard if the final week is bad enough.

 

ORH,

 Agreed. The SAI is more at risk to end up mediocre, which would be enough to at least lower my confidence in a strong -AO this DJF IF that happens. From the stats I've analyzed, it appears to me that the SAI has had a stronger correlation with the DJF -AO than has the SCE. So, ending up with a not so good SAI would not be what I'd want to see to say the least. However, I'm not saying that would mean we couldn't still get a strong -AO in DJF. If one were to give the OPI and SCE a decent amount of weight, things could still look pretty good I think, especially if we actually get a Nino. Also, I'll be looking at the November AO.

 Regardless, I'm still hopeful that the SAI will end up being excellent! This lack of Natice/Rutgers data is lowering confidence.

[AlaskaETC]

ORH,

 Agreed. The SAI is more at risk to end up mediocre, which would be enough to at least lower my confidence in a strong -AO this DJF IF that happens. From the stats I've analyzed, it appears to me that the SAI has had a stronger correlation with the DJF -AO than has the SCE. So, ending up with a not so good SAI would not be what I'd want to see to say the least. However, I'm not saying that would mean we couldn't still get a strong -AO in DJF. If one were to give the OPI and SCE a decent amount of weight, things could still look pretty good I think, especially if we actually get a Nino. Also, I'll be looking at the November AO.

 Regardless, I'm still hopeful that the SAI will end up being excellent! This lack of Natice/Rutgers data is lowering confidence.

 

Well, it's not like they don't want to be the bearer of bad news...

[GaWx]

Well, it's not like they don't want to be the bearer of bad news...

 

Lol. I didn't mean it that way, of course. I mean it is lowering confidence in what exactly has been and is happening from day to day SCEwise from my perspective. Rutgers is a fantastic database that I was following very closely and it also allows apples to apples comparsons to past years. Just seeing random maps/graphs posted here from different sources isn't as good from my perspective. Don't get me wrong. I very much appreciate those posting maps/graphs here, especially while there are no Natice updates. They're more valuable than ever right now!! However, they use different scales, they aren't being posted every day, and it is much more difficult to compare day to day and current to past years for me.

[The_Global_Warmer]

The indices would definitely fall some with the last week.

 

But that doesn't mean there is more than a negligible response from the atmosphere.

 

Insolation between 50-60N is already really low on par with early February levels.

 

 

I wouldn't worry to much about it even if snow cover extent below 60N drops a lot the next 5 days.

 

Unless there is some lingering into November effect that still matters.

[AlaskaETC]

Lol. I didn't mean it that way, of course. I mean it is lowering confidence in what exactly has been and is happening from day to day SCEwise from my perspective. Rutgers is a fantastic database that I was following very closely and it also allows apples to apples comparsons to past years. Just seeing random maps posted here from different sources isn't as good from my perspective. Don't get me wrong. I very much appreciate those posting maps here, especially while there are no Natice updates. They're more valuable than ever right now!! However, they use different scales, they aren't being posted every day, and it is much more difficult to compare day to day and current to past years for me.

 

I'm actually feeling pretty confident...but my knowledge of this subject is minimal at best. However, the snowfall, at least compared to last year is quite promising. Even if we do swing and miss to close out October, I would say we would have a pretty good set up. Just my guess though...

[GaWx]

The indices would definitely fall some with the last week.

 

But that doesn't mean there is more than a negligible response from the atmosphere.

 

Insolation between 50-60N is already really low on par with early February levels.

 

 

I wouldn't worry to much about it even if snow cover extent below 60N drops a lot the next 5 days.

 

Unless there is some lingering into November effect that still matters.

 

Ok, Friv. I'm no longer worried. Forget this last week of Oct. Are you happy now?

 

PS. I'm just trying to be objective.

 

PPS: I can't believe Friv, the Global Warmer, is suddenty more optimistic than me for this winter. That sounds backward.

 

PPPS: I see Mike is in the house. If I went literally by his graphs for SAI and SCE, I'd really be pessimistic. However, I'm not. No offense, Mike.

[millwx]

 

 

Edit: I realize MillWx wouldn't care as much about this period. However, that's beside the point.

 

To be clear... I don't think it's irrelevant.  The Week 43 stats (which ends tomorrow) maintain a pretty decent (albeit noticeably lower) correlation.  So, you really want to keep the snow throughout Oct.  Late snow melt is definitely NOT welcome, even by my measure.  BUT, what I will tell you is that it does matter LESS.  Note that Week 41 also has a high-but-slightly-lower correlation.  I think what that tells me is that Week 42 is the centerpoint on timing... that is, the process by which this aids North American winters (albedo -> strat warming -> trop cooling) takes a long time.  And, obviously, the amount of time it takes is imperfect and varies... but Week 42 probably catches the peak.  But I think Week 41 and 43 maintaining a high correlation points not only to the imperfect timing, but I think it also indicates that a week of significant coverage isn't enough... in order to have the desired effect the snow cover needs to be in place for a solid couple weeks (at least).  So, you want Week 41 and/or Week 43 staying high with Week 42.

 

Point being, you're right, my focus is on Week 42, but I do think this week's snow cover is also of relevance.  FYI, the correlation drops another good notch on Week 44 and gets ugly by Week 45.  As we've previously discussed, the weeks seem to be aligned a little early this year.  So, even though Week 43 ends tomorrow, it would be fair to argue that everything up to the end of October matters.  That said, if most of our loss this week comes after tomorrow (though it seems like it may come on/before), I'll be pretty dubious of its impact.  Though the entire rest of this week is relevant, that relevance is on a gradual fade already.  But, as I said... not totally irrelevant.  It's not going to kill us... the job is essentially done... but, in a perfect world, we still do NOT want snow loss this week.  By mid-Nov it IS completely irrelevant (and damned nearly so by early Nov).

[Mike.Ventrice]

Here is my latest analysis. There was a drop of snow cover during the October 21-23 time frame, but it seems there has been a rebound since.

 

 

[GaWx]

Millwx,

 In your week 42 analyses, do you literally look at week 42 without adjustments for differing calendars each year?