2014-15 winter outlook

[Cold Rain]

Haha, i love those clown maps! Not sure anyone takes them too seriously, but they are fun to look at and they occasionally are right, especially the further North you go.

 

Last year the NAM clown map had 10" of snow for virtually the entire state of NC like 3 days out (I think it was 3 days out...it was either 2 or 3).  We ended up getting a lot of sleet.

[JoMo]

Mack/folks,

 I've said this before in jest and I'll say it again, especially with winter fast approaching: I wish those Euro clown (snow) maps were banned! They're not even what the Euro is really producing where 850's are not at or below 0C through the entire event! Like actual clowns, they should be laughed at. Just a reminder: the Euro clowns show all precip. falling with sfc temp's of 32 or colder as snow! Yes, you heard that right, folks. They ignore the 850's! For example, this shows a whopping 14" of snow in north ATL. Now, did I ever think anything close to 14" (would almost certainly be heaviest on record there) was possible there? No,because 850's above 0C wouldn't let that happen. The Euro wasn't really predicting 14" of snow, regardless of what its own clown showed lmao. It was predicting 1.4" of liquid equivalent falling with temp.'s of 32 or colder. The vast majority of this was predicted with 850's above 0C...so mainly IP or ZR, which is what verified. Now granted, the 1.4" of liquid equiv. did turn out to be too high as the total there was under 1". But the ATL area still got a major storm out of this including the greatest IP by far since 1/1988 on the northside and a major ZR on the southside. There was no reason to be extremely disappointed in what actually verified unless one took this clown at face value with its 14" nonsense. The scary thing is that they were circulating on Facebook (thank goodness I'm not a member) amongst the wx illiterates. These ignorant folks have no clue about these clowns being "fake" in these situations!

 

BAN EURO CLOWNS ASAP! They belong in the circus, not on wx boards and certainly not on FB!

 

Well, you definitely feel strongly about this subject. I noticed last year that the maps were 'incorrect', specifically for the reason you mentioned. It just shows where precip falls when the surface temp is predicted to be 32 or below by the model. So, I use it as an indicator where the heaviest frozen precipitation might be. I can see your point about the general population though and most of them will take it as meaning that much snow will fall. I don't think they should be 'banned', but knowledgeable people should note the limitations and point out what the map actually means to the weather illiterates.  

[Shawn]

Through scripting, these vendors could make their Euro snowfall data more correct.

 

Basically, even if the raw output is "show snow on 32F temps", you can basically write a script in something such as GRADS like:

 

If surface temp + bl temps are <=32 and 850mb <=0c then

print snowfall

else

if surface temp <=32 and 850mb >= 1c then

print ice/mix/sleet etc etc

 

The only problem, is ratios would have to be programmed in also.. depending on temps... but a lot of maps are all assumed at 10:1 anyways.  Math can be done on the side to come up with better amounts.

[machetemoonlight]

Through scripting, these vendors could make their Euro snowfall data more correct.

 

Basically, even if the raw output is "show snow on 32F temps", you can basically write a script in something such as GRADS like:

 

If surface temp + bl temps are <=32 and 850mb <=0c then

print snowfall

else

if surface temp <=32 and 850mb >= 1c then

print ice/mix/sleet etc etc

 

The only problem, is ratios would have to be programmed in also.. depending on temps... but a lot of maps are all assumed at 10:1 anyways.  Math can be done on the side to come up with better amounts.

good post. I just figured it was self-explanatory, and I'm kinda dumb even.

[GaWx]

Through scripting, these vendors could make their Euro snowfall data more correct.

 

Basically, even if the raw output is "show snow on 32F temps", you can basically write a script in something such as GRADS like:

 

If surface temp + bl temps are <=32 and 850mb <=0c then

print snowfall

else

if surface temp <=32 and 850mb >= 1c then

print ice/mix/sleet etc etc

 

The only problem, is ratios would have to be programmed in also.. depending on temps... but a lot of maps are all assumed at 10:1 anyways.  Math can be done on the side to come up with better amounts.

Shawn ,

I recall you suggesting this last winter. What you say makes sense and doesn't seem difficult to implement.

[superjames1992]

Memories.......I'd rather forget! The big one, Feb of last year, a day or two before , this was being shown! Lol, it still hurts!

 

There was a run of the Weatherbell clown for the Canadian which depicted 28-30" of snow IMBY from the early March ice storm (which ended up being major IP/ZR and knocked out power for a week, but still).  That was fun.

 

Of course, the Euro clown prior to the "Big One" in February also depicted 16-24" a few times here.

[Wow]

There was a run of the Weatherbell clown for the Canadian which depicted 28-30" of snow IMBY from the early March ice storm (which ended up being major IP/ZR and knocked out power for a week, but still).  That was fun.

 

Of course, the Euro clown prior to the "Big One" in February also depicted 16-24" a few times here.

 

One of these days, they'll be right! 

[RaleighWx]

Ah winter isnt far away. I will post my winter outlook this year here. Probably sometime around Halloween.

[FallsLake]

Ah winter isnt far away. I will post my winter outlook this year here. Probably sometime around Halloween.

Allan, we're counting on you to roll the dice in our favor...

[RaleighWx]

Allan, we're counting on you to roll the dice in our favor...

 

You may be in luck. But I am still watching and researching.

[wxjunkie]

Last year the NAM clown map had 10" of snow for virtually the entire state of NC like 3 days out (I think it was 3 days out...it was either 2 or 3).  We ended up getting a lot of sleet.

Yeah, that sucked man. I hate it when you end up with that warm layer. I guess i was thinking more of 09-10 in Virginia. Those maps turned out to be fairly accurate. Not as familiar with down here yet, hope im not disappointed a lot, but seems like that could happen.

[Bob Chill]

I agree with you guys about wxbell snow maps. It's not just the euro either. It's all models because wxbell uses the same flawed math in all of them. However, they are a decent tool for precip totals in general. Pretty easy to move a decimal place. Other snow map output is more accurate but still flawed (stormvista, wunderground, etc).

OTOH, it's pretty easy to take model output and estimate your own snowfall. Marginal temp situations should always be scrutinized. Especially in changeover scenarios. Raw model output can be too quick to drop temps and also unreliable. If raw model output is showing 32 in my yard, its a huge flag for snow falling but not accumulating.

Matt (zywts) and I go through lengthy discussions about potential reality vs raw output with every threat at medium to short leads. Models are only math. GIGO can easily make it's way into any model when looking at snow accumulations. It's a fun exercise deciphering the fine details and forming opinions on ratios, changeovers, surface temps, 850's, and in between 850 and the surface. Warm noses are inherently difficult. It can be pretty displeasing when you have -2 850's and sub-freezing surface but +2 sitting somewhere in between.

Snow maps are a quick tool and deeper investigation is the next step. Soundings often uncover a can of worms when simple output like snow maps paint a different story. In a perfect setup, snow maps can even bust low because of ratios and the extremely complicated nature of nailing precip totals at even short leads.

GaWx and myself like to post euro ind member snow output because it gives a deeper understanding of consensus or too many outliers on the means. But even those use the same stupid equation and run at lower resolution so buying into that output verbatim comes with many caveats and is bad practice in general. The biggest takeaway with that output is getting an idea of how many members are coming up with solutions that are close enough to the possibility even with the flaws.

The short story of this tl;dr post is that snow maps are a good tool if you know how to interpret all the other important data. Snow maps posted on facebook by a weather moron with followers who have zero understanding as to how numerical models work causes all harm and no good.

[griteater]

Yeah, that sucked man. I hate it when you end up with that warm layer. I guess i was thinking more of 09-10 in Virginia. Those maps turned out to be fairly accurate. Not as familiar with down here yet, hope im not disappointed a lot, but seems like that could happen.

 

Go ahead and prepare your mind for it...now.  

[Cold Rain]

Thanks for stopping by, Bob.  Always enjoy reading your posts.

 

Can't wait for your cold and snowy winter forecast, Allan!  You'll be glad you put that out.

[BullCityWx]

Go ahead and prepare your mind for it...now.

Let's talk about the February 2014 storm. Even if it isn't the damn warm layer, you lose the moisture where you need it.

[Shawn]

Bob summed it up pretty well.  We know how to use the snow maps, and actually find them kind of funny sometimes.  In the end, we must always look at all the finer details and come up with our own snow totals.

 

Putting them out as is across Twitter and Facebook is a big mistake for the general public.  Down here in the Southern areas, it's much harder for them to comprehend; Not so much up North.

[DaculaWeather]

And guys... I'll add this.

 

Clown maps are no different than ANY computer model map, none of them can be taken literally. Every model is different, each has it's own strength and weaknesses. Models are only tools, not the final answer. You have to look at all models and all levels of the atmosphere to know what's going to happen. To do anything less, and then post it on the internet as fact, is just irresponsible, and gives all of us a bad name. 

 Just my .02$ :-)

[Cold Rain]

I'm not sure why it's so hard to create a snow map that lines up with what should really be snow.  Shawn's suggestion above makes sense.  Why doesn't it make sense for the ones creating the maps, and why is it so hard for them to implement a change, especially when they are so out of whack (talking the Euro ones here, specifically)?

[wxjunkie]

Go ahead and prepare your mind for it...now.  

LOL.. true.

 

I probably won't be the only one bitching that its not cold enough and there isnt enough snow. 

[Cold Rain]

LOL.. true.

 

I probably won't be the only one bitching that its not cold enough and there isnt enough snow. 

 

That's a dead lock stone cold straight up guarantee.  You'll be frequently joined by brick, jshetley, metal, brick, snowstorm, brick, jshetley, and brick...etc.

[wxjunkie]

That's a dead lock stone cold straight up guarantee.  You'll be frequently joined by brick, jshetley, metal, brick, snowstorm, brick, jshetley, and brick...etc.

HAHA  Awesome!

 

Looks like i will be in good company, but why arent you on the list?

[Cold Rain]

HAHA  Awesome!

 

Looks like i will be in good company, but why arent you on the list?

 

Haha really?

[wxjunkie]

Haha really?

I guess i missed it..lol

 

You guys ever check out some of the Winter Outlooks on youtube?

 

[GaWx]

I agree with you guys about wxbell snow maps. It's not just the euro either. It's all models because wxbell uses the same flawed math in all of them. However, they are a decent tool for precip totals in general. Pretty easy to move a decimal place. Other snow map output is more accurate but still flawed (stormvista, wunderground, etc).

OTOH, it's pretty easy to take model output and estimate your own snowfall. Marginal temp situations should always be scrutinized. Especially in changeover scenarios. Raw model output can be too quick to drop temps and also unreliable. If raw model output is showing 32 in my yard, its a huge flag for snow falling but not accumulating.

Matt (zywts) and I go through lengthy discussions about potential reality vs raw output with every threat at medium to short leads. Models are only math. GIGO can easily make it's way into any model when looking at snow accumulations. It's a fun exercise deciphering the fine details and forming opinions on ratios, changeovers, surface temps, 850's, and in between 850 and the surface. Warm noses are inherently difficult. It can be pretty displeasing when you have -2 850's and sub-freezing surface but +2 sitting somewhere in between.

Snow maps are a quick tool and deeper investigation is the next step. Soundings often uncover a can of worms when simple output like snow maps paint a different story. In a perfect setup, snow maps can even bust low because of ratios and the extremely complicated nature of nailing precip totals at even short leads.

GaWx and myself like to post euro ind member snow output because it gives a deeper understanding of consensus or too many outliers on the means. But even those use the same stupid equation and run at lower resolution so buying into that output verbatim comes with many caveats and is bad practice in general. The biggest takeaway with that output is getting an idea of how many members are coming up with solutions that are close enough to the possibility even with the flaws.

The short story of this tl;dr post is that snow maps are a good tool if you know how to interpret all the other important data. Snow maps posted on facebook by a weather moron with followers who have zero understanding as to how numerical models work causes all harm and no good.

 

Bob,

 Nice write-up. I do want to clarify/summarize my observations regarding the Euro and GFS snow maps. Based on both my vendor and Weatherbell, the Euro (op. and ensemble members) basically ignores temperatures other than the surface. (Shawn knows about the Wxbell Euro maps.) So, it is definitely not just Wxbell for the Euro. When I called my vendor, they said that the snow maps' formulas are tied to ECMWF, themselves, and that, therefore, their hands were tied. Also, my vendor's GFS snow maps definitely aren't based on the same erroneous calculations as the Euro's based on years of following them. The clearly won't generate snow accumulations when 850's are warmer than 0C like the Euro does if the surface is 32 or colder.

 A significant % of SE US wintry precip. falls in forms other than snow. So, this is often a pretty important issue here just as it was in Feb. 2014.

[Cold Rain]

I guess i missed it..lol

You guys ever check out some of the Winter Outlooks on youtube?

I watch them when I find them, even if they're for other areas. Even though they are, the forecaster still usually discuss their views for the country as a whole.

[superjames1992]

I really need a pure snowstorm.  No sleet.  No worries.  No freezing rain.  No lack of moisture.  Let's rock.  No more stress hoping the models didn't underpredict 850 mb temperatures by a tenth of a degree Celsius and all your pretty modeled snow turns to sleet.

 

The January 28th storm was just that except moisture was very much lacking...

[toxictwister00]

I really need a pure snowstorm.  No sleet.  No worries.  No freezing rain.  No lack of moisture.  Let's rock.  No more stress hoping the models didn't underpredict 850 mb temperatures by a tenth of a degree Celsius and all your pretty modeled snow turns to sleet.

 

The January 28th storm was just that except moisture was very much lacking...

 

True, I think the January 28th snowstorm is the only storm I can remember in a while that was purely all snow imby. If we could have taken the moisture from the February winter storm earlier this year and kept the overall setup with the cold air in the January event that would have been a phenomenal storm!

[Ellinwood]

Bob,

 Nice write-up. I do want to clarify/summarize my observations regarding the Euro and GFS snow maps. Based on both my vendor and Weatherbell, the Euro (op. and ensemble members) basically ignores temperatures other than the surface. (Shawn knows about the Wxbell Euro maps.) So, it is definitely not just Wxbell for the Euro. When I called my vendor, they said that the snow maps' formulas are tied to ECMWF, themselves, and that, therefore, their hands were tied. Also, my vendor's GFS snow maps definitely aren't based on the same erroneous calculations as the Euro's based on years of following them. The clearly won't generate snow accumulations when 850's are warmer than 0C like the Euro does if the surface is 32 or colder.

 A significant % of SE US wintry precip. falls in forms other than snow. So, this is often a pretty important issue here just as it was in Feb. 2014.

 

One of the biggest problems with the ECMWF is that its precip output is binomial, meaning that it's either snow (frozen) or rain (not frozen). The GFS output divides the frozen precip into snow, sleet, and freezing rain, which goes a long way in producing more logical snow maps. From there, most sites will take this snow QPF output and give it a 10:1 ratio and leave it at that. Some sites do use their own additional algorithms to take other variables into account that the models do not.

 

Of course, that's just one of many problems of snow output maps. Last season, I wrote a fairly comprehensive blog article highlighting the different ways snow maps can and often do fail: http://madusweather.com/2013/12/modelology-101-snow-maps/ I imagine those who are already pretty familiar with the snow maps know most/all of what I wrote about, but it's pretty useful for people who don't have much knowledge about snow forecasting.

 

Someone correct me on any of this if I am wrong... this is just what I have picked up from different people and sources.

[GaWx]

One of the biggest problems with the ECMWF is that its precip output is binomial, meaning that it's either snow (frozen) or rain (not frozen). The GFS output divides the frozen precip into snow, sleet, and freezing rain, which goes a long way in producing more logical snow maps. From there, most sites will take this snow QPF output and give it a 10:1 ratio and leave it at that. Some sites do use their own additional algorithms to take other variables into account that the models do not.

 

Of course, that's just one of many problems of snow output maps. Last season, I wrote a fairly comprehensive blog article highlighting the different ways snow maps can and often do fail: http://madusweather.com/2013/12/modelology-101-snow-maps/ I imagine those who are already pretty familiar with the snow maps know most/all of what I wrote about, but it's pretty useful for people who don't have much knowledge about snow forecasting.

 

Someone correct me on any of this if I am wrong... this is just what I have picked up from different people and sources.

 

Mark,

 Thanks for your reply. Regarding the Euro, this part of your Dec. 2013 writeup is most important imo:

 "The ECMWF has a particularly difficult time with this area, in which it highly favors snowfall when there is a layer of above freezing temperatures somewhere between the ground and the cloud tops. This often results in much greater snowfall in areas that may actually receive little to no snow!"

 

 This is consistent with what I was told. So, it generates either rain or snow as you said...so no ZR, graupel, or IP. It appears that as long as it is 32 or colder at the surface, all IP is treated as accumulating snow. Also, all ZR appears to be treated as accumulated snow. This is a big disappointment vs. the GFS and a big problem in the SE US since I think they get a higher % of total wintry precip. as these forms vs. many other parts of the country. Could it be they do that because perhaps there's relatively little ZR or IP in Europe??

[Ellinwood]

Mark,

 Thanks for your reply. Regarding the Euro, this part of your Dec. 2013 writeup is most important imo:

 "The ECMWF has a particularly difficult time with this area, in which it highly favors snowfall when there is a layer of above freezing temperatures somewhere between the ground and the cloud tops. This often results in much greater snowfall in areas that may actually receive little to no snow!"

 

 This is consistent with what I was told. So, it generates either rain or snow as you said...so no ZR or IP. It appears that as long as it is 32 or colder at the surface, all IP is treated as accumulating snow. Also, all ZR appears to be treated as accumulated snow. This is a big disappointment vs. the GFS and a big problem in the SE US since I think they get a higher % of total wintry precip. as these forms vs. many other parts of the country. Could it be they do that because perhaps there's relatively little ZR or IP in Europe??

 

Unfortunately, I do not know the reasoning behind why the ECMWF decides to output the frozen precip the way they do. It would be great to have some access to the developers of the ECMWF (if even just a fraction of what we have with NCEP).