Snow Climo Classroom

[psuhoffman]

I have been seeing some of the same questions posted wrt patterns lately and been getting some PMs about it so I thought I would post some stuff I have been working on for a while wrt snow climo and patterns.  Others can add to this also.  I did a study of all warning level snowfall events at BWI going back to 1948 and I will post the results below but first I thought I would give some info.

One disclaimer:  most of this research and information is focused on warning criteria snowfalls.  We can fluke our way to a minor advisory level event in almost any pattern during mid winter.  It really muddies things if we include a lot of 1-3" type events.  But there are clear patterns if we exclude minor events and only look at what patterns are right for 5"+ events.  I am not dismissing minor snowfalls but its really hard to identify patterns for a 1-2" snowfall.  

Some common pattern driver indexes we reference

The colder phase of these indexes are -EPO (ridge), +PNA (Ridge), -AO (weaker TPV over the pole), -NAO (ridging).  

But colder doesn't always equal snow.  There is no one pattern for snow.  The trick is getting the jet right so that we get a storm into our "box".

The most common tracks to get a significant storm into that box are shown.  But there are a variety of pattern combinations to get the jet wavelengths right.  Changing one factor can change what we need from another.  But the key is getting the jet aligned for the track of storms to end up just to out south or southeast.  

For example...most of us know about the idea -NAO pattern...

But...a -EPO pattern can work too

or a Hudson Bay Ridge

 

I will go into specific patterns and what percentage of our warning criteria snowfalls came from those patterns in the next few posts

[psuhoffman]

Snowfall Study

5"+ events at BWI since 1948

88 Events studied

For my snowfall study I looked at every warning event (5"+) at BWI going back to 1948 (that is when the upper air data is available).  I know some of those events happened over a longer period than 12 hours and might not technically be a warning event but I thought 5" for the whole event was good enough.  I used BWI because I feel like it is a bit more indicative for the urban corridor than DCA.  If you live on a slab of concrete on an island in the Potomac south of DC...I apologize to you...and I feel bad for you, but oh well.  I also included 3 events that fell just shy (they were all 4.8 or 4.9" at BWI but were significant events for a large portion of our area.  That might be cheating but its my study and I can do what I want.  Perhaps one of these events didn't hit your exact location the same...but when a storm misses by 200 miles its the pattern...when it misses by 20 miles its bad luck.  So the patterns that caused these storms show potential, even if a particular event wasn't as good at your house as BWI.  I then looked at the upper level patterns.  For each storm it took a bit of analysis because sometimes I had to go back over the week leading up to determine what the predominant pattern or feature that lead to the storm was.  Sometimes the key to a storm is something that happened the week before, and if you just look at the day of the storm you can't tell why the pattern was right for a snowstorm at all.  So some of this is a bit subjective...the index might have shows something different on the day of the event...but if you look at the week leading up to it the pattern was in a different state.  Also some h5 features key to a snowfall here can be invisible on the indexes. A Hudson Bay ridge might now show up. An extremely west based NAO Block can be muted by a trough in the eastern NAO domain or a trough near the Azores. An east based epo can be hidden by a trough in the western epo domain.  I wanted to identify what features at h5 really indicate snow potential.  So with that out of the way...

There were 88 warning level events going back to 1948

Looking at each index on its own

The NAO was the most important single index for predicting snow

58 of the storms featured a -NAO

3 were neutral

27 were positive

 

The AO was second most predictive

54 featured a -AO

9 neutral

26 positive

 

The EPO was evenly split

42 -EPO

4 neutral

42 positive

 

The PNA surprisingly had more events in the negative phase (however most of our really HUGE snowstorms did features a +PNA ridge)

44 negative PNA

5 neutral

39 positive

This was the breakdown for the index combinations and snowstorms for each combo

                 the order for the chart is

                      EPO/PNA/AO/NAO

-	-	-	-	16
+	+	-	-	13
-	+	-	-	11
+	-	-	-	10
-	-	+	+	9
+	+	+	+	6
-	+	+	+	3

  

Some take aways from the chart...

the most common snowfall combo is a pattern driven by a -EPO ridge in conjunction with atlantic blocking.  

The next most common is a +PNA ridge with atlantic blocking

The 3rd most common is a combo EPO/PNA ridge with atlantic blocking.

So....atlantic blocking with ANY pacific help from either the PNA, EPO, or both...accounts for 40 of the 88 storms.

After that we are left with storms when only ONE factor was working in favor of it.  

For those....

10 were purely a function of High Lat blocking overcoming a crap pacific

9 were purely from an EPO ridge with no PNA or High Latitude Help (All of them were an east based EPO ridge)

6 were purely from a favorable PNA ridge and no other help

Only 3 came from a combo EPO/PNA ridge with no blocking.  So for everyone that throws around 1993....that was one of only 3 storms that hit with that look.  A huge EPO/PNA ridge with no blocking is just cold/dry and progressive.  Usually a quick cold shot then a warm up.  

One last note... by far the single biggest factor if we had to work with only one thing to get snow...is the Hudson Bay Ridge.  We had 16 storms where the only really positive feature working in our favor was a Hudson Bay Ridge.  Most of those storms don't show up on the chart because that feature often falls in between the typical index domains and in a lot of those storms all the other factors looked like crap.  Apparently a Hudson Block is a local enough phenomenon to overcome pattern issues in a lot of other areas.  It was by far the biggest way we got a snowstorm in an otherwise crap pattern.  

Also there were no cases with a -AO as the ONLY factor working in favor.  That is not because the AO isnt important its just really hard to have a -AO without there being an EPO or NAO ridge because there is some overlap of those domains.   It's also hard not to have EPO or NAO help...since the TPV being located near one usually opens the door to ridging near the other.  And if we did have lower heights all across both domains the AO would likely be very positive.  

I will show some of the common composites to these patterns in following posts

 

 

[psuhoffman]

EPO Ridge Snowfalls

I will start with the EPO dominant storms since we have been talking a lot about the EPO lately.

This explains an EPO ridge some

The EPO ridge shown there is a central (or even west based) EPO.  Those don't do us much good without some HL blocking.  The ridge shown there would likely lead to the storm track in red on the map.  IF...we had some HL blocking to suppress the flow and hold high pressure in then the track in purple is more likely.  Without any HL blocking...the EPO actually is almost equally able to bully the pattern into a snowy one as much as the NAO...9 events to 10 with only those factors in favor...but it has to be a specific type of EPO ridge.

Below is the loading pattern for each of the 9 EPO driven snowfall events absent atlantic help

You can see we need the ridge to be centered into northwest Canada NOT out over the PAC.  That configuration can create a pattern favorable for progressive waves that get under us.  We actually want the PNA negative for this to work.  A full latitude ridge with no atlantic help just is cold and dry.  The western trough with an EPO ridge over the top leads to a pattern like this...

 

 

If the EPO ridge is a west based EPO then we need at least SOME atlantic blocking help.  These all feature a -PNA also because a western EPO ridge almost always leads to a -PNA.

There were no examples of a west based EPO ridge without any blocking help.

Since we are starting a central PAC ridge in the face right now on guidance I will include those analogs.  The ridge north of Hawaii is really a WPO ridge more than a EPO although the two overlap some but a ridge centered there can only work with extreme NAO blocking.  There were no examples of a snowstorm with a ridge there without NAO help.

This is the composite from those snow events. 

[psuhoffman]

Hudson Bay Ridge

16 Storms

This was the single biggest factor able to equalize a crap pattern.  A lot of our "fluke" snows in otherwise bad patterns came because of this feature.  Basically a ridge near Hudson Bay is perfectly located to force a storm under us even without much else right.  Actually if other features are lined up right a hudson ridge would probably suppress a storm way to our south.  We would want the block much further north normally.  But in an otherwise crap pattern that feature has saved us often.  THis is the composite of those events.  

Seeing a ridge centered near Hudson Bay can indicate we have a shot...even if the pattern is otherwise crap

[psuhoffman]

Scand Ridge w/50/50 Pattern

10 storms

Another singular factor that showed up in many storms that seemed not to fit the profile of our typical snow events (10 storms) was a Scandinavian ridge that is encroaching into the eastern NAO domain.  That feature when combined with a 50/50 low can work out even in an otherwise flawed pattern.  

[psuhoffman]

The Money Look

HECS Storms

This is the common look for our big ones.  Almost all of our 20" plus events fit this mold.  The only exception is PDII.  That storm was pure lucky timing with a cold airmass and a 50/50 low.  But all our other HECS storms fit this pattern.  

Obviously the NAO block with 50/50 low is the prominent feature.  The PNA ridge also helps to get a system to dig into the east and cut off.  THe EPO is actually typically positive during these events.  It is hard to have extreme blocking on both sides.  

[psuhoffman]

I will do some more profiles tomorrow... others feel free to ask questions or add information to this thread.  Goodnight

[nj2va]

This is really great, thanks for doing it.  Maybe we can get a Hudson Bay ridge pop up magically to save the crap January pattern & gives us a fluke warning level storm.

[CAPE]

Many here have probably seen this CWG article by Wes from a few winters ago, but for those who have not, it is very good and explains/discusses the roles of the key indices(AO, NAO, EPO, ENSO, etc) and their impacts on cold and snow for the DC area, and includes insightful scatter diagrams.

https://www.washingtonpost.com/news/capital-weather-gang/wp/2015/11/12/what-the-different-weather-patterns-might-mean-for-snow-this-winter/

[George BM]

@psuhoffman I don't have anything to add but thank you so much for the time and effort put into this!

[Sparky]

The "Money Look" seems like a complicated setup from your diagrams, no wonder it is on the rarer side.  That aside, many if not most of our big dumpers tend to slow down as they are climbing up the eastern seaboard.  What is that attributed to?  What prevents them from hauling ass into that large elongated area of low pressure stretched across the northern atlantic?

[CAPE]

1 hour ago, Sparky said:

The "Money Look" seems like a complicated setup from your diagrams, no wonder it is on the rarer side.  That aside, many if not most of our big dumpers tend to slow down as they are climbing up the eastern seaboard.  What is that attributed to?  What prevents them from hauling ass into that large elongated area of low pressure stretched across the northern atlantic?

It's a little hard to get a sense of it looking at a composite, but a sustained -NAO (and often a quasi-stationary 50-50 low underneath) produces a true "block" in the flow- basically the opposite of a progressive pattern. Along with a favorable storm track and a more sustained eastern trough, it 'forces' a strengthening coastal low to move more slowly to the NE or ENE under the block.

[Eskimo Joe]

This is a phenomenal write up, and it's obvious you went through a lot to collate the information, thank you.  I have two follow up questions for BWI.

1.) Would it be possible to re-run the data for 10 inch or greater events?  

2.) Would it be possible to re-run the data for days when the high temperatures was at or below 32 degrees at BWI?  I'm curious to see if there is any strong correlation for cold weather.

[WinterWxLuvr]

Good stuff

[psuhoffman]

6 hours ago, Eskimo Joe said:

This is a phenomenal write up, and it's obvious you went through a lot to collate the information, thank you.  I have two follow up questions for BWI.

1.) Would it be possible to re-run the data for 10 inch or greater events?  

2.) Would it be possible to re-run the data for days when the high temperatures was at or below 32 degrees at BWI?  I'm curious to see if there is any strong correlation for cold weather.

1) that will be easy. I’ll work on that either tonight or tomorrow. The hard part was compiling the data set and examining the pattern for each one.  You can’t just look at the numerical indexes. Often an important feature is hidden. A Hudson Ridge often wont show up. Some west based NAO blocks don’t really show up well if the eastern NAO domain near Iceland isn’t impacted or there is a trough near the Azores.  An east based epo ridge can sometimes show as neutral if there is an Aleutian low encroaching on the western epo domain.  There could also have been blocking days before the event that by the time of the snow relaxed and the NAO flipped positive. Only looking at the day of the storm you would miss it.  Now that I have that data sorting it in different ways is easy. 

2.  I can easily do that but the day of a snow will obviously show a correlation to cold since it has to be below avg to snow. During the storm it’s very likely near or below freezing. A better study would be to look at temps the day before and after.  That would be harder. I would have to go back and look it up. I could give you the dataset and you could do it. But even then you will clearly see a correlation to cold. What you might not see is a high correlation to arctic cold. A lot of the events had temps near freezing. A lot were mixed events where 1.3 QPF lead to 6 or 8” of snow with a high of 34. That was a pretty common thing in the set. Pure snow warning events were rare. But that’s not shocking. The all snow zone on a lot of storms is relatively narrow. And most storms in the mid latitudes ride the boundary between warm/cold. We aren’t far enough north to often get a storm that’s riding the arctic boundary or a storm that’s late in its cycle and no longer pressing WAA north. You have to either have elevation or more latitude to get a lot of those. 

[Scarlet Pimpernel]

@psuhoffman, many thanks for compiling this!  Very informative and useful...appreciate your efforts and time!

[jaydreb]

Thanks Psu.  Is there a region of the country for which our “crap” pattern (i.e. the one we are facing) is actually a dream pattern?  Are there weenies out there in another forum cheering for a +EPO, +NAO, -PNA pattern?

[MDstorm]

Thanks for the great write-up.  Very informative and you obviously invested a lot of time in this endeavor.  

 

 

[CAPE]

1 hour ago, jaydreb said:

Thanks Psu.  Is there a region of the country for which our “crap” pattern (i.e. the one we are facing) is actually a dream pattern?  Are there weenies out there in another forum cheering for a +EPO, +NAO, -PNA pattern?

The further north you go and/or at higher elevations the less critical it is to have all the weather index "stars" aligned. The upper MW/GLs and NNE can still do okay with the advertised crapola pattern. If the trough does in fact lock in out west then obviously the mountain areas out there will do quite well.

[cbmclean]

21 hours ago, psuhoffman said:

THe EPO is actually typically positive during these events.  It is hard to have extreme blocking on both sides.

@PSUHoffman

I seem to remember you debating this with someone last year.  Seems like their argument was that it was physically impossible to get significant -EPO and significant -NAO at the same time.  You retorted that it was indeed possible and pointed to a specific time frame.  Can you remember any instances of this?  I think some may call it a "ridge bridge". 

[ravensrule]

Amazing write up PSU. This must have taken a ton of time. Thank you for investing so much time. 

[psuhoffman]

1 hour ago, cbmclean said:

@PSUHoffman

I seem to remember you debating this with someone last year.  Seems like their argument was that it was physically impossible to get significant -EPO and significant -NAO at the same time.  You retorted that it was indeed possible and pointed to a specific time frame.  Can you remember any instances of this?  I think some may call it a "ridge bridge". 

I remember the exact conversation you are talking about.  It is possible to get a ridge across the entirety of the EPO/AO/NAO domains...and yes that is commonly called a ridge bridge...and when it happens it is extremely good for our snow chances.  But it is very rare.  So the fact that there are not more snow events with both a -NAO and -EPO is simply because that combination is less common not that it isnt a good pattern.   But the reason our absolute biggest storms are usually associated with a +EPO is mostly because they are also associated with the most anomalous west based NAO blocks.  Typically a block that strong cut off over greenland makes it even harder to have a strong ridge in the EPO domain.  Frankly if we did it would probably be too much of a good thing and might flood the CONUS with so much cold it would just be a uber cold dry look.  Think January 1977

\

Brutal cold but not much snow.  

[psuhoffman]

1 hour ago, cbmclean said:

@PSUHoffman

I seem to remember you debating this with someone last year.  Seems like their argument was that it was physically impossible to get significant -EPO and significant -NAO at the same time.  You retorted that it was indeed possible and pointed to a specific time frame.  Can you remember any instances of this?  I think some may call it a "ridge bridge". 

These are some examples of warning events with a ridge bridge

[psuhoffman]

PNA RIDGE Snowfalls

+EPO/+PNA/+AO/+NAO snowfalls

6

I only got one more profile done today but figured I would add a PNA ridge pattern snow composite since that seems to be what we are working with for our only real shot at snowfall next week.  

These were the 6 storms where the only thing really working for us was an ideally placed PNA ridge.

This composite is centered for the day the storm was starting.  You can see the ideal placement for a PNA ridge there can sometimes overcome what is otherwise a pretty craptastic pattern there.  I did notice the ridge axis is slightly west of what is common for snowstorms when we have some blocking.  THat is likely because without any NAO blocking the pattern is more progressive and we want the PNA ridge slightly west of where we normally see it in most snow events.  It's not that far off...normally it would be closer to Boise then Tahoe but still its noticeable.  

I guess with the coming "threat" if we want to call it that...if we see a ridge looking like that out west at the perfect time it COULD be a legit threat.  

[psuhoffman]

15 hours ago, jaydreb said:

Thanks Psu.  Is there a region of the country for which our “crap” pattern (i.e. the one we are facing) is actually a dream pattern?  Are there weenies out there in another forum cheering for a +EPO, +NAO, -PNA pattern?

As CAPE said this is a great look for the west. Cold and snowy. Could be too cold for the northern ski resorts but places like Colorado when it’s sunny isn’t that cold. I skied Stramboat when it was -25 with the sun out and it actually felt ok.  During a storm with wind at -25 though...forget it. 

But Alta in Utah is already up near 200” on the year with 20-30” expected in the next 72 hours. Yes they get a lot of snow but that’s still a great start with their snowiest months yet to come.  I know Revelstoke was doing pretty good also last time I checked. 

I’m heading out there either the last weekend  in January or the first 2 weekends of Feb and I’m salivating already. With this pattern they should be found great with all backcountry chutes and bowls open. 

ETA: you have to get pretty far north in the east for that to be a good look. Depending on how much the boundary can press and where the TpV near Greenland is it CAN be a good pattern for the mountains in northern New England. Northern NY Vermont and Maine. But if the TPV isn’t far enough south this can be a rain pattern even up there.  I was up around Quebec City in the winter of 2008 (comp year to this coming pattern) and it rained all the way to where I was near Le Massif at 3000 feet elevation 40 miles north of Quebec City in mid winter.  That’s about as rare a feat for them as a warning snow is for us. 

[cbmclean]

On 12/31/2019 at 12:01 PM, psuhoffman said:

As CAPE said this is a great look for the west. Cold and snowy. Could be too cold for the northern ski resorts but places like Colorado when it’s sunny isn’t that cold. I skied Stramboat when it was -25 with the sun out and it actually felt ok.  During a storm with wind at -25 though...forget it. 

But Alta in Utah is already up near 200” on the year with 20-30” expected in the next 72 hours. Yes they get a lot of snow but that’s still a great start with their snowiest months yet to come.  I know Revelstoke was doing pretty good also last time I checked. 

I’m heading out there either the last weekend  in January or the first 2 weekends of Feb and I’m salivating already. With this pattern they should be found great with all backcountry chutes and bowls open. 

ETA: you have to get pretty far north in the east for that to be a good look. Depending on how much the boundary can press and where the TpV near Greenland is it CAN be a good pattern for the mountains in northern New England. Northern NY Vermont and Maine. But if the TPV isn’t far enough south this can be a rain pattern even up there.  I was up around Quebec City in the winter of 2008 (comp year to this coming pattern) and it rained all the way to where I was near Le Massif at 3000 feet elevation 40 miles north of Quebec City in mid winter.  That’s about as rare a feat for them as a warning snow is for us. 

Hey PSU.  If you are bored and are just looking for stuff to do, I would be really interested in seeing a similar dissertation about crappy patterns.  For example, my understanding is that the crap that started right before Christmas is a different sort of fail from the Pacific Doom Blob that we will soon be enjoying.  I think in one of your posts today you mentioned it as a north Pacific problem as opposed to a SW Pacific problem. 

[SpringfieldSkins]

As someone who mainly lurks and checks in when storms are on the horizon, this is some really really cool stuff you’ve posted.  Thanks!

[psuhoffman]

7 hours ago, cbmclean said:

Hey PSU.  If you are bored and are just looking for stuff to do, I would be really interested in seeing a similar dissertation about crappy patterns.  For example, my understanding is that the crap that started right before Christmas is a different sort of fail from the Pacific Doom Blob that we will soon be enjoying.  I think in one of your posts today you mentioned it as a north Pacific problem as opposed to a SW Pacific problem. 

Sure but I am not going to spend as much time studying our fail patterns as I did our winning ones.  I just don't hate myself that much.  But I know off the top of my head what years and patterns really suck balls and put together a quick profile of some of the worst ones.  

First of all... while we can score a fluke snowfall in many patterns...the truth is there are way more looks that just don't work out often then there are ones that do.  We are south of the mean northern stream jet in winter most of the time.  It takes an anomalous pattern to get snowfall.  The warm wet cold dry thing is real.  Often in winter the only time the cold boundary gets south of us is after a wave passes and the flow behind it presses the boundary south.  But without a favorable pattern the return flow ahead of the next wave is likely to press the boundary back to our north before it gets here.  Basically the average storm track is to our north.  But most patterns with some luck and some bootleg factor (like a Hudson ridge) working for us we can occasionally get snow.  But most patterns without luck we can easily go with no snow as well.  

Bad Luck

On top of that...sometimes we can just get unlucky in a perfect pattern.  A storm gets suppressed, the next one is just slightly too warm, another develops off the coast...next thing you know we wasted a 2 week perfect pattern.  The winter of 2007 would fall into that category.  Look at this H5 pattern for Feb

March was pretty good also for a good portion of it.  THe first half was pretty crappy and I will use December 2006 as an example of a crap look later but even with a wasted Dec into Jan if you told me that would be our look for Feb and Mar I would take my chances.  ANd while we did get some snow...we didn't cash in on the full potential and so the winter went down as a below avg snowfall year.  There were 2 blockbuster coastals that year...one in Feb and one in Mar and both were just SLIGHTLY too warm at the mid levels and the big snowfall ended up a little NW of us.  THere was one storm that had big potential but a PV lobe at the wrong time squashed it.  And a couple others just failed to reach potential and ended up minor snowfalls.  We mostly wasted a really good pattern.  That has happened other times as well.

On top of that, even more common, we waste a look that is OK, not great but definitely not awful and end up with a really bad result.  YOu can look at the H5 for a year like 1981 and say...that looks ok.  Not epic but you wouldn't think a horrible dud winter was coming, but it was.   Truth is snow here is anomalous and luck wrt timing and discreet features that fall under the pattern level have a LOT to do with our results.  Call it luck or chaos but pattern is only half the fight.  

So all that said I will focus on a couple patterns that just really really really suck and give us almost no chance for a significant snowfall, luck be damned.  

Some general observations

First of all, any pattern that features a +NAM state without either a perfectly placed east based EPO ridge...or a really favorable PNA ridge is a fail pattern.  We can score in a +NAM pattern but only with pacific help.  If the AO/NAO are both awful (like right now) and the pacific inst in a very favorable state...that is a total fail pattern.  That is because anything that pumps a SE ridge without resistance up top will not end well here.  

Additionally there is seasonal variance.  The NAO is way more important later in the season.   As the temperature profile tends to get colder across N. Amer later in the season and wavelengths start to shorten in Feb, and water temperatures cool along the coast, blocking on the atlantic side can influence and overcome the pacific pattern more.  Early in the season we really need pacific help to have a good chance.  

But bad looks can be very temporary or transient so what patterns tend to give us that and lock in for an extended period of time.  I assume that is what you mean because we can have a bad week even in the best winters...but what looks spell doom for a really long time and eat up a big chunk of our winter window.

AK Vortex Pattern

The first is an AK vortex pattern.  This one is especially a killer early in the season.  It can be overcome from Mid Jan onward but only with NAO blocking.  Early it tends to be a problem even with blocking.    Without blocking this look is a total fail pattern all winter long.  You pointed out earlier that 2012 was a good example.  December 2006 was also a perfect example.  The last 10 days also featured this problem.  

I have the h5 below...the issue with this pattern is the flow under the AK vortex floods the continent with warm pac air.  Without ideal blocking on the NAO side that mild air will flood across and not only will we be warm but there won't be any cold air anywhere near us.  This pattern can take a LONG time to recover from.  Even once the pattern improves we can waste a week just getting cold back onto our side of the hemisphere. 

Dec 2006

January 2012

Last week...

Pac Ridge +NAM pattern

The next one is the one we are about to go into according to all guidance.  How long it lasts is yet unknown...but this is the absolute worst pattern we can possibly ever get into in winter, both because of how bad it can be...numerous examples of completely snowless months with this pattern...but also because of how stubborn it can be.  Frankly this is the most common season destroyer of all the patterns.  This one accounts for the majority of our total fail winters.  

This is the composite of many of our total fail years that fit this description, the Jan/Feb of 1950, 1951, 1959, 1973, 1989, 1990, 2002, 2008, 2013.  All of these were single digit snowfall years at BWI except 1990 but all the snow fell that year BEFORE the pattern set in early January.  

look familiar?

This one is real simple...that pac ridge digs the trough out west...which pumps the ridge in the east.  Simple wave physics.  Without extreme blocking to offer resistance that will push the thermal boundary well to our north.  

The reason this pattern can be so stubborn is an anomalous central pac ridge is usually a primary effect not a secondary effect.  Meaning usually it is being driven by the tropical pattern in the central pacific and Maritime Continent regions.   It's not a result of something that caused something that caused something...and so on.  It is the direct result of a very major driver of the global pattern.  IF that driver doesn't change...that ridge can park there for months.  And that ridge loads a wave pattern over north america that sucks for us.  It can only be overcome with extremely -NAM state to suppress the SE ridge.  

Too much of a good thing

The last one I will cover that has accounted for some total crap years is the "too much of a good thing" pattern.  I almost didn;t include this one as its a weird anomaly.  This one is rare, usually only happening during a super nino.  And if you just glance at the h5 it doesn't look that bad.  And...it does increase the potential of a blockbuster storm.  We had this pattern a few times and sometimes we get lucky and time up just enough cold to get a HECS.  But if we dont....we can go long stretches simply too mild to get snow.  The best examples are 1983 and 1998.  Yea we had the HECS in 1983 but imagine if that one storm had not hit.  The rest of the prime part of that winter was a total wasteland.  Basically like 1998.

The issue is hard to see just from h5 but its kinda there.  The super nino.  That pac trough is actually a bit much...if it was a little less anomalous and near the Aleutians  not so expansive it would be perfect, and that is the typical moderate nino look.  But in this case that extreme trough in the pac is flooding the CONUS with mile air.  THe trough over the gulf is the juiced up STJ but the flow under the pac trough combined with the ridging across the US/Can border has flooded the US with mild air and cut off any cross polar flow to press cold into the pattern.  The result is just a mild pattern with a good storm track that yields cold rain.  It's not a mild look.  Just not cold enough to snow.  Pretty miserable really.  This one suck but its not that common so I wouldn't worry too much.  Plus it gives us the best chance of all the fail patterns at a fluke storm, and a BIG one at that.  But since 1998 was one of our worst winters I figured I would throw in an explanation of why.  

There are other fail patterns...pretty much any look with a SE ridge and no blocking... but those first 2 are the main ones that can last a long time and wreck months or even a whole season.   If anyone wants to add more great but I am depressed enough now so have a good night.  Hope this was what you were looking for.  

[cbmclean]

12 hours ago, psuhoffman said:

Sure but I am not going to spend as much time studying our fail patterns as I did our winning ones.  I just don't hate myself that much.  But I know off the top of my head what years and patterns really suck balls and put together a quick profile of some of the worst ones.  

First of all... while we can score a fluke snowfall in many patterns...the truth is there are way more looks that just don't work out often then there are ones that do.  We are south of the mean northern stream jet in winter most of the time.  It takes an anomalous pattern to get snowfall.  The warm wet cold dry thing is real.  Often in winter the only time the cold boundary gets south of us is after a wave passes and the flow behind it presses the boundary south.  But without a favorable pattern the return flow ahead of the next wave is likely to press the boundary back to our north before it gets here.  Basically the average storm track is to our north.  But most patterns with some luck and some bootleg factor (like a Hudson ridge) working for us we can occasionally get snow.  But most patterns without luck we can easily go with no snow as well.  

Bad Luck

On top of that...sometimes we can just get unlucky in a perfect pattern.  A storm gets suppressed, the next one is just slightly too warm, another develops off the coast...next thing you know we wasted a 2 week perfect pattern.  The winter of 2007 would fall into that category.  Look at this H5 pattern for Feb

March was pretty good also for a good portion of it.  THe first half was pretty crappy and I will use December 2006 as an example of a crap look later but even with a wasted Dec into Jan if you told me that would be our look for Feb and Mar I would take my chances.  ANd while we did get some snow...we didn't cash in on the full potential and so the winter went down as a below avg snowfall year.  There were 2 blockbuster coastals that year...one in Feb and one in Mar and both were just SLIGHTLY too warm at the mid levels and the big snowfall ended up a little NW of us.  THere was one storm that had big potential but a PV lobe at the wrong time squashed it.  And a couple others just failed to reach potential and ended up minor snowfalls.  We mostly wasted a really good pattern.  That has happened other times as well.

On top of that, even more common, we waste a look that is OK, not great but definitely not awful and end up with a really bad result.  YOu can look at the H5 for a year like 1981 and say...that looks ok.  Not epic but you wouldn't think a horrible dud winter was coming, but it was.   Truth is snow here is anomalous and luck wrt timing and discreet features that fall under the pattern level have a LOT to do with our results.  Call it luck or chaos but pattern is only half the fight.  

So all that said I will focus on a couple patterns that just really really really suck and give us almost no chance for a significant snowfall, luck be damned.  

Some general observations

First of all, any pattern that features a +NAM state without either a perfectly placed east based EPO ridge...or a really favorable PNA ridge is a fail pattern.  We can score in a +NAM pattern but only with pacific help.  If the AO/NAO are both awful (like right now) and the pacific inst in a very favorable state...that is a total fail pattern.  That is because anything that pumps a SE ridge without resistance up top will not end well here.  

Additionally there is seasonal variance.  The NAO is way more important later in the season.   As the temperature profile tends to get colder across N. Amer later in the season and wavelengths start to shorten in Feb, and water temperatures cool along the coast, blocking on the atlantic side can influence and overcome the pacific pattern more.  Early in the season we really need pacific help to have a good chance.  

But bad looks can be very temporary or transient so what patterns tend to give us that and lock in for an extended period of time.  I assume that is what you mean because we can have a bad week even in the best winters...but what looks spell doom for a really long time and eat up a big chunk of our winter window.

AK Vortex Pattern

The first is an AK vortex pattern.  This one is especially a killer early in the season.  It can be overcome from Mid Jan onward but only with NAO blocking.  Early it tends to be a problem even with blocking.    Without blocking this look is a total fail pattern all winter long.  You pointed out earlier that 2012 was a good example.  December 2006 was also a perfect example.  The last 10 days also featured this problem.  

I have the h5 below...the issue with this pattern is the flow under the AK vortex floods the continent with warm pac air.  Without ideal blocking on the NAO side that mild air will flood across and not only will we be warm but there won't be any cold air anywhere near us.  This pattern can take a LONG time to recover from.  Even once the pattern improves we can waste a week just getting cold back onto our side of the hemisphere. 

Dec 2006

January 2012

Last week...

Pac Ridge +NAM pattern

The next one is the one we are about to go into according to all guidance.  How long it lasts is yet unknown...but this is the absolute worst pattern we can possibly ever get into in winter, both because of how bad it can be...numerous examples of completely snowless months with this pattern...but also because of how stubborn it can be.  Frankly this is the most common season destroyer of all the patterns.  This one accounts for the majority of our total fail winters.  

This is the composite of many of our total fail years that fit this description, the Jan/Feb of 1950, 1951, 1959, 1973, 1989, 1990, 2002, 2008, 2013.  All of these were single digit snowfall years at BWI except 1990 but all the snow fell that year BEFORE the pattern set in early January.  

look familiar?

This one is real simple...that pac ridge digs the trough out west...which pumps the ridge in the east.  Simple wave physics.  Without extreme blocking to offer resistance that will push the thermal boundary well to our north.  

The reason this pattern can be so stubborn is an anomalous central pac ridge is usually a primary effect not a secondary effect.  Meaning usually it is being driven by the tropical pattern in the central pacific and Maritime Continent regions.   It's not a result of something that caused something that caused something...and so on.  It is the direct result of a very major driver of the global pattern.  IF that driver doesn't change...that ridge can park there for months.  And that ridge loads a wave pattern over north america that sucks for us.  It can only be overcome with extremely -NAM state to suppress the SE ridge.  

Too much of a good thing

The last one I will cover that has accounted for some total crap years is the "too much of a good thing" pattern.  I almost didn;t include this one as its a weird anomaly.  This one is rare, usually only happening during a super nino.  And if you just glance at the h5 it doesn't look that bad.  And...it does increase the potential of a blockbuster storm.  We had this pattern a few times and sometimes we get lucky and time up just enough cold to get a HECS.  But if we dont....we can go long stretches simply too mild to get snow.  The best examples are 1983 and 1998.  Yea we had the HECS in 1983 but imagine if that one storm had not hit.  The rest of the prime part of that winter was a total wasteland.  Basically like 1998.

The issue is hard to see just from h5 but its kinda there.  The super nino.  That pac trough is actually a bit much...if it was a little less anomalous and near the Aleutians  not so expansive it would be perfect, and that is the typical moderate nino look.  But in this case that extreme trough in the pac is flooding the CONUS with mile air.  THe trough over the gulf is the juiced up STJ but the flow under the pac trough combined with the ridging across the US/Can border has flooded the US with mild air and cut off any cross polar flow to press cold into the pattern.  The result is just a mild pattern with a good storm track that yields cold rain.  It's not a mild look.  Just not cold enough to snow.  Pretty miserable really.  This one suck but its not that common so I wouldn't worry too much.  Plus it gives us the best chance of all the fail patterns at a fluke storm, and a BIG one at that.  But since 1998 was one of our worst winters I figured I would throw in an explanation of why.  

There are other fail patterns...pretty much any look with a SE ridge and no blocking... but those first 2 are the main ones that can last a long time and wreck months or even a whole season.   If anyone wants to add more great but I am depressed enough now so have a good night.  Hope this was what you were looking for.  

Very informative, thanks.

[Ji]

This should be pinned. Thanks man