And we begin

[billgwx]

Thanks! This is fantastic since I haven't been able to locate a graph like this online.

[griteater]

Any reason why not all years are represented on your graph?

 

The SAI using daily snow data only goes back to Oct 1998.  I think what baggett was trying to show was an example of a year with the slowest snow growth (Oct 2007), and an example of a year with the fastest snow growth (Oct 2009)...then he kept recent years in there (2012-2013) as he started producing the charts.

[cfbaggett]

The SAI using daily snow data only goes back to Oct 1998.  I think what baggett was trying to show was an example of a year with the slowest snow growth (Oct 2007), and an example of a year with the fastest snow growth (Oct 2009)...then he kept recent years in there (2012-2013) as he started producing the charts.

Exactly right.

[weathafella]

Well the first third of October sure is snowing spectacular vs other years increase despite a high launching pad. To put it another way, it's night and day vs a year ago even though we have roughly the same starting point.

[Sunny and Warm]

Exactly right.

Thank you for the graph.  Any chance you could add a 'climatology' line to simulate what the last 16 years average looks like.

[DTWXRISK]

GIVEN how  badly your  forecast least autumn  sucked and sucked and sucked and sucked...

You gave me a rash of   crap  and never  admitted   you were  way off and  your argument was worthless

Once the niño gets rolling it's likely that ridge will get the boot into Canada/W. Half of the conus.

This winter will almost certainly need strong NWATL/GIS/Hudson Bay blocking.

When I was looking at analogs the other day for ONI DJF above 0.8 and 1.2 there was tremendous blocking with -NAO.

With the NEPAC under a vortex in almost everyone.

But heights were still well below normal for the Southern half of the Conus. With the above normal temps only reaching the upper Midwest and lakes. With Russua and the Russian side of the arctic below normal.

Right mow I personally give the ONI a an 80% chance of averaging somewhere between 0.8c and 1.2c during DJF.

The key is the -NAO. Likely West based.

I give an 75% chance of a +epp this winter.

But above normal snowfall from N. Texas thru Southern Kansas, Dallas to the NE, Notth of Little Rock to St. Lous, from South Central IL to around Laf, from Western Kentucky to Indy, South Central Ohio , West Virginia, Southern Pennsyvania, to Phiily, DC, Baltimore and the North half of Virgina.

I see well above normal over the Mid South or NE Miss, Northern Alabama, Northern Georgia, and the Southern Mtns.

Another band of potential well above normal from Northerm Jersey, Eastern New York, NYC, Hartford, Boston, SE Maine,

I think there will be a huge gradient to below normal snow NW of there in WNY, Vermont, And NH.


I think the upper Midwest is normal or below but does well with clippers, the lakes normal to to possibly below in LEK zones because of the lakes cooling off to fast.

[DTWXRISK]

FOOKING  FABULOUS   

I thought I would provide a mid-week update...

 

[millwx]

If the GFS 8-Day snowfall forecast is correct (I'd link it here, but if you're not subscribed to the model suite you won't see it), the Eurasian snow cover for Week 42 will be epic.  Based on current coverage, plus the GFS forecast (excluding snow under only a few centimeters as I think that's meaningless and won't get into the "official" snow cover amount), the Week 42 Eurasian snow cover would be almost identical to 1978 (the second highest on record).  Of course, a major caveat in that... that's based on the GFS.  The Euro isn't TOO dissimilar, but it definitely has somewhat less snowfall over the next week.  (FYI, I point this out now because Week 42 Eurasian snow cover has the highest correlation to U.S. winter temperatures; Week 41 and 43 are close behind - in that order - but Week 42 is the one with the best signal.)  At this point, I'm going to be a little skeptical, since that's such extreme coverage, and it depends on the GFS being right (yikes!); but it's certainly going to be interesting week for tracking the snow extent.

[DTWXRISK]

MILLIWX
 

  There are some   FREE   sources   that one can use  for snow cover/ fall in Siberia  ...if  one knows  where to look

DT  knows 

 

 

If the GFS 8-Day snowfall forecast is correct (I'd link it here, but if you're not subscribed to the model suite you won't see it), the Eurasian snow cover for Week 42 will be epic.  Based on current coverage, plus the GFS forecast (excluding snow under only a few centimeters as I think that's meaningless and won't get into the "official" snow cover amount), the Week 42 Eurasian snow cover would be almost identical to 1978 (the second highest on record).  Of course, a major caveat in that... that's based on the GFS.  The Euro isn't TOO dissimilar, but it definitely has somewhat less snowfall over the next week.  (FYI, I point this out now because Week 42 Eurasian snow cover has the highest correlation to U.S. winter temperatures; Week 41 and 43 are close behind - in that order - but Week 42 is the one with the best signal.)  At this point, I'm going to be a little skeptical, since that's such extreme coverage, and it depends on the GFS being right (yikes!); but it's certainly going to be interesting week for tracking the snow extent.

 

[NJwinter23]

I can't say I'm in love with the 1-2 week pattern being modeled at all. Cold bottled up north of 60 in Asia and a reversal of our north pacific  slp signal...might be a bland end to the month in the sai department

[InstantWeatherMaps]

If the GFS 8-Day snowfall forecast is correct (I'd link it here, but if you're not subscribed to the model suite you won't see it), the Eurasian snow cover for Week 42 will be epic.  Based on current coverage, plus the GFS forecast (excluding snow under only a few centimeters as I think that's meaningless and won't get into the "official" snow cover amount), the Week 42 Eurasian snow cover would be almost identical to 1978 (the second highest on record).  Of course, a major caveat in that... that's based on the GFS.  The Euro isn't TOO dissimilar, but it definitely has somewhat less snowfall over the next week.  (FYI, I point this out now because Week 42 Eurasian snow cover has the highest correlation to U.S. winter temperatures; Week 41 and 43 are close behind - in that order - but Week 42 is the one with the best signal.)  At this point, I'm going to be a little skeptical, since that's such extreme coverage, and it depends on the GFS being right (yikes!); but it's certainly going to be interesting week for tracking the snow extent.

I might be a little biased here, being the owner of the site in question, but this map is from a free resource and clearly shows a huge amount of snow cover in Eurasia (assuming you meant the 00Z GFS since that's the one with the largest snow cover... the 06Z and 12Z are nothing to laugh at either though... also 60N is the second circle from the inside):

Source

[das]

I thought I would provide a mid-week update...

 

Thanks in advance for doing these again this year.  Much appreciated!

das

[millwx]

MILLIWX

 

  There are some   FREE   sources   that one can use  for snow cover/ fall in Siberia  ...if  one knows  where to look

DT  knows 

 

 

If you run through the historical maps, snow cover down into extreme northern Kazhakstan and to the Mongolian border by Week 42 is a pretty rare event (as projected by the GFS).  1978 is an excellent match.  Again, to be clear, I'm skeptical.  But it will be an interesting week ahead watching how the EA snow extent evolves.

[NJwinter23]

 

Great analysis.....

 

Just something I was contemplating when following the models today...and would love to hear educated opinions on it... was trying to compare the end of the month pattern up against high SAI years when making that comment. I might be worried for nothing...and even if it isnt a good look, maybe we've done more than enough damage up to that point with the fast start we've had.

 

Here are high SAI years in the final half of Oct

 

2009

 

2002

 

2012

 

11-15 day ensembles seems to keep the cold locked over existing snow cover and may not support further advancement from where we are south of 60N at this point? I dont know if that +WPO look is detrimental or not, but it wasnt present on those select high SAI years

 

[Mike.Ventrice]

Current SAI. Values plotted are snow extent in Eurasia south of 60N. Like last year, we are starting with above normal snow extent. Let's see where things go over the next 21 days.

Hey guys,

 

What is the source of this data? Thanks in advance

[jburns]

Weird that my comments are being deleted, but DT's personal attacks are untouched.

Your deleted comments consist of an emoticon and one sentence trying to explain why you used it.  Drop it and get back on topic.

[GaWx]

If the GFS 8-Day snowfall forecast is correct (I'd link it here, but if you're not subscribed to the model suite you won't see it), the Eurasian snow cover for Week 42 will be epic.  Based on current coverage, plus the GFS forecast (excluding snow under only a few centimeters as I think that's meaningless and won't get into the "official" snow cover amount), the Week 42 Eurasian snow cover would be almost identical to 1978 (the second highest on record).  Of course, a major caveat in that... that's based on the GFS.  The Euro isn't TOO dissimilar, but it definitely has somewhat less snowfall over the next week.  (FYI, I point this out now because Week 42 Eurasian snow cover has the highest correlation to U.S. winter temperatures; Week 41 and 43 are close behind - in that order - but Week 42 is the one with the best signal.)  At this point, I'm going to be a little skeptical, since that's such extreme coverage, and it depends on the GFS being right (yikes!); but it's certainly going to be interesting week for tracking the snow extent.

 

 The 0Z GFS fwiw is suggesting ~~18.5 msk as of 10/31/14 (~week 43.5), which would be ~4.25 higher than normal, ~3 higher than 2013, and would be near 2012, 2009, 1973, and 1969. Only 1976 and 1972 would be clearly higher. Regarding Oct. increase, it would be ~13, which would be ~2 higher than normal.  Only 2012, 09, 06, 03, 76, 70, & 68 would be higher while there'd be ~ a tie with 11 & 84. So, out of 45 years, 2014 would be ~tied for 8th.

 

 On a related note, I checked 2012, 09, 06, 03, 76, 70, 68, 11, & 84 to see how many of them had a strong -AO in DJF. 2012, 09, 03, 84, 76, and 68 (2/3 of them) all had strong to very strong DJF -AO's with them being 14th, 1st, 17th, 11th, 2nd, and 3rd most -AO, respectively. 1970 only had a modest -AO (~30th). 2011/2006 (12th/7th most +AO's of 64) were big fails. So, these 9 averaged about 21st most -AO DJF out of 64. With the median being between 32nd and 33rd, I'd call the correlation between Oct. Eurasian snowcover increase and DJF -AO a moderate correlation though the top three -AO's of the 64 all were at/near the top for Oct. snow increases, which alone is quite impressive.

[millwx]

 The 0Z GFS fwiw is suggesting ~~18.5 msk as of 10/31/14 (~week 43.5), which would be ~4.25 higher than normal, ~3 higher than 2013, and would be near 2012, 2009, 1973, and 1969. Only 1976 and 1972 would be clearly higher. Regarding Oct. increase, it would be ~13, which would be ~2 higher than normal.  Only 2012, 09, 06, 03, 76, 70, & 68 would be higher while there'd be ~ a tie with 11 & 84. So, out of 45 years, 2014 would be ~tied for 8th.

 

 On a related note, I checked 2012, 09, 06, 03, 76, 70, 68, 11, & 84 to see how many of them had a strong -AO in DJF. 2012, 09, 03, 84, 76, and 68 (2/3 of them) all had strong to very strong DJF -AO's with them being 14th, 1st, 17th, 11th, 2nd, and 3rd most -AO, respectively. 1970 only had a modest -AO (~30th). 2011/2006 (12th/7th most +AO's of 64) were big fails. So, these 9 averaged about 21st most -AO DJF out of 64. With the median being between 32nd and 33rd, I'd call the correlation between Oct. Eurasian snowcover increase and DJF -AO a moderate correlation though the top three -AO's of the 64 all were at/near the top for Oct. snow increases, which alone is quite impressive.

 

Interesting.  FWIW, however, despite being closer to winter, Week 43.5 is less important than Week 42.  I'm not going to get into the SAI versus raw extent debate again, as it's "muddy" (oddly, the SAI does a better job on the AO link, but a worse job on the actual weather, as it picks up on several, uncommon -AO warm winters that the raw extent does not pick up... but that could just be a sample issue, so who knows... but it's why - for now - I just prefer extent).  But speaking just from the extent perspective and the actual weather... Week 42 is the #1 correlator to U.S. winter temperatures.  Why would the correlation fade later in the month?  I'm not 100% sure I know the answer to that, but have multiple theories (#1: the physics behind why the snow cover matters is such that it takes time for the impact on U.S. weather to be realized... Week 42 may just be the most optimal point in time for it to matter; #2: extent becomes large enough, climatologically, by late month that the anomalies are more around the edges and less critical).  I know for the SAI Week 43.5 (end of Oct) is what matters.  But for raw extent (which holds a better relation to U.S. weather), Week 42 is the peak.

 

Week 42 began two days ago.  So, this week is the most important.  And, FYI, it's not in the documentation on the Rutgers site, but just looking at past data it doesn't look like the "Week" is a weekly average; it looks to me like just a snapshot of the final day of the week.  So, the data for 10/21 (to be reported on 10/22) is the key datapoint.  Obviously, it shouldn't change wildly day to day, so all the days right around that should be revealing too.  But, just from a specific, technical standpoint... Week 42, which I believe is the 10/21 daily snapshot, is the peak of the correlation to U.S. winter weather.

[GaWx]

Interesting. FWIW, however, despite being closer to winter, Week 43.5 is less important than Week 42. I'm not going to get into the SAI versus raw extent debate again, as it's "muddy" (oddly, the SAI does a better job on the AO link, but a worse job on the actual weather, as it picks up on several, uncommon -AO warm winters that the raw extent does not pick up... but that could just be a sample issue, so who knows... but it's why - for now - I just prefer extent). But speaking just from the extent perspective and the actual weather... Week 42 is the #1 correlator to U.S. winter temperatures. Why would the correlation fade later in the month? I'm not 100% sure I know the answer to that, but have multiple theories (#1: the physics behind why the snow cover matters is such that it takes time for the impact on U.S. weather to be realized... Week 42 may just be the most optimal point in time for it to matter; #2: extent becomes large enough, climatologically, by late month that the anomalies are more around the edges and less critical). I know for the SAI Week 43.5 (end of Oct) is what matters. But for raw extent (which holds a better relation to U.S. weather), Week 42 is the peak.

Week 42 began two days ago. So, this week is the most important. And, FYI, it's not in the documentation on the Rutgers site, but just looking at past data it doesn't look like the "Week" is a weekly average; it looks to me like just a snapshot of the final day of the week. So, the data for 10/21 (to be reported on 10/22) is the key datapoint. Obviously, it shouldn't change wildly day to day, so all the days right around that should be revealing too. But, just from a specific, technical standpoint... Week 42, which I believe is the 10/21 daily snapshot, is the peak of the correlation to U.S. winter weather.

Millwx,

Interesting. I may take a closer look at 42. Also, I just noticed that the three strongest -AO winters since 1950-1 were all Nino's and all had at/near the top SAI as per my earlier post. So, assuming we end up, say, in the top 10 SAI's, the chance at a solid -AO would seemingly be better than the overall moderate correlation would suggest. Otoh, 06 was a big fail though it wasn't as high a SAI as 09, 76, and 68 as a group.

[griteater]

Euro ensemble brings back some higher heights across northern Eurasia in the 7-10 day timeframe, which would aid storms and snow advance to the south.  Seems like this Oct each time the models try to rid that area of above normal heights, it's mostly a fail.  The higher heights have won the battle more times than not.

[DTWXRISK]

 FOLKS

 

Taking a look at the new European model from early on Friday morning for eastern and northern Asia we see more promising looking weather patterns and interesting developments .  All the data shows a massive system at 500 mb will develop a severe negative tilt and 72 hours which will develop in turn a enormous system over central and eastern Siberia.
 

This map shows the negative tilt that 96 hours and the impressive 988Mb  Low from the operational European model. 

 

 

Once the system passes up into far eastern Siberia and close to the Bering sea ...the next very powerful and strong piece of energy of 500mb from sweeping of Central Russia.  We can see that here at DAY 5: 

 

The massive LOW is still over eastern Siberia but there  is another piece of energy racing eastward from Central Russia.    

 

At  day 7  the  0z Friday European  ensemble shows a very strong  shortwave or close 500 low and Central Russia.  At day 9 the  euro ensemble tracks the system to the southeast very close to nor the Manchuria China. ( see red   X).  Again if the European  ENSEMBLE is correct this would be a significant expansion of snow cover fairly far to the south.

[CoastalWx]

Agree DT.

[NJwinter23]

Glad to see my end of month concerns don't appear warranted at all.

 

And to further back up the good news...fantastic look in the mid-strat right through the end of the latest runs.

 

 

 

[GaWx]

 The 12Z euro/18z gfs look really good to me regarding predicted Eurasian snowfall below 60N.

[weathafella]

I'm hoping cfbaggett gives a graphical update soon. It looks great with some good gains again today.

[ORH_wxman]

I'm hoping cfbaggett gives a graphical update soon. It looks great with some good gains again today.

 

 

Today's gains were huge...might have been the biggest yet S of 60.

[griteater]

Today's gains were huge...might have been the biggest yet S of 60.

 

Comparison between last year and this year on Oct 17 is no comparison

 

[GaWx]

Big news although no surprise: Mantua Sep. PDO out & is +1.08, which is at ~91 percentile for +. I'm now raising chance for Mantua +PDO DJF from 80% the last few months to 95%. Why? There were 17 Sep. PDO's of +0.70+. ALL of them ended up with a +PDO (+0.11+) for DJF! Keep in mind that exactly 50% of the 114 DJF's were +/-...i.e., no bias. Also, I'm now going with a 2/3 chance that the DJF PDO will be +0.50+. Why? Out of those 17 Sep.'s, 12 had a +0.50 DJF. Also, 2/3 of the nine Nino's within these 17 had a +0.50+ DJF PDO.

[GaWx]

 The 10/17/14 (at week 41.5) Eurasian SCE is massive. I'm roughly estimating it is near 14.5 msk, which is near climo for 11/1! See the two maps below that show how I estimated it. I believe that 2014 is quite possibly running behind only 1976 and 1966 in SCE as of 10/17. So, with two weeks to go, the SAI is already ~9 vs. climo of ~6.5! The complete Oct. SAI climo is 11. The odds are quite high that that will at least be met. The largest SAI for the entire Oct. was 1976's ~21. 2nd was 2009's ~16. 3rd was 2012's ~15.5. 4th was 1970's ~14.5. 5th was 2003's ~14.25. 6th was 1968's ~14.

 

Map for 10/17/14:

 

Now, compare this map with the 11/1 climo map (concentrate on 50% climo, which is near where the blue-green and brighter green join) and you'll see they're very close:

[millwx]

 The 10/17/14 (at week 41.5) Eurasian SCE is massive. I'm roughly estimating it is near 14.5 msk, which is near climo for 11/1! See the two maps below that show how I estimated it. I believe that 2014 is quite possibly running behind only 1976 and 1966 in SCE as of 10/17. So, with two weeks to go, the SAI is already ~9 vs. climo of ~6.5! The complete Oct. SAI climo is 11. The odds are quite high that that will at least be met. The largest SAI for the entire Oct. was 1976's ~21. 2nd was 2009's ~16. 3rd was 2012's ~15.5. 4th was 1970's ~14.5. 5th was 2003's ~14.25. 6th was 1968's ~14.

 

Map for 10/17/14:

SCE101714.png

 

Now, compare this map with the 11/1 climo map (concentrate on 50% climo, which is near where the blue-green and brighter green join) and you'll see they're very close:

SCENov01Climo.png

 

Impressive.  As you know, I look at raw coverage for Week 42.  Looking at it that way, it's 1976 and 1978 (not 1966) that top the list.  The coverage now (we are midway through Week 42) is on par with 1978.  Sorry, because of the image handling for the weeklies on the Rutgers site, I can't post 1978 here.  But compare today:

 

 

...to the image at this link:

 

http://climate.rutgers.edu/snowcover/chart_vis.php?ui_year=1978&ui_week=42&ui_set=0

 

In the details, looks like this year still trails 1978 a little - we definitely have less around the edges right now.  But the weeklies, to the best of my knowledge, are just a daily snapshot at the END of the week. So, if we get any further expansion over the next four days, this could easily match or even surpass 1978.  Of course, at this time of year, that's a big "if"... we could, potentially, see melt around the edges over the next several days.  We'll just have to wait and see.  But clearly, this is some pretty epic snow cover we've got right now.  Of course, high snow cover is still no guarantee (hello 2006!), but it sure doesn't hurt!