NBC Action Communities

Blogs
Welcome to NBC Action Communities Sign in | Join | Help
in
Back to NBCActionNews.com Home Blogs

NBC Action Weather Blog

Battle of the Computer Models

Watch NBC Action News in HD for Kansas City's most accurate forecast!

Good morning bloggers,

Today we are going to compare the ECMWF to the GFS as these models have been consistent in the past few runs in rather diverging solutions in the 6 to 10 day forecast period.  There was a discussion about the LRC in yesterdays blog, and it was quite educational and somewhat entertaining.  I barely had enough time to read all of the comments and I wish I had much more time to address some of the questions. I will find the time in about two weeks as we finish up our winter forecast, so please be patient. If you do not believe in the LRC then I say you do not believe that what happened in October into early November will return this winter.  Because, the basics of my theory are that a pattern sets up, cycles, and repeats. So, later this fall and through the winter if the pattern repeats and can be shown to be related to what happened earlier in the fall then it is strong evidence that the LRC does exist.  Do you agree?  I know, here we go again.  I will try to find some more time to answer any questions you may have, but I do have a full slate today.

O.K., let's begin with the battle between the beloved ECMWF (European) model and the GFS model.  The ECMWF has been consistent in its last three runs to NOT dig a trough into the west during the next ten days.  The GFS and its ensembles have been consistent in the other direction. Let's take a look at these models:

The above 500 mb forecast map is the 216 hour ECMWF model.  As I said the European model has been consistent with a solution of not digging anything into the western states. This is very different from what the GFS has been doing during the past few days.  Take a look at the GFS valid around the same time as the above ECMWF:

So, the battle is on.  Which model has the right idea?  We are still analyzing the developing weather pattern for the season. This is the time of the year with the lowest amount of confidence in the pattern, but as the days and weeks go by my confidence grows substantially. We are still learning more about what this pattern will mean for the fall and winter months, not to mention the spring and summer.  Let's see what happens during the next ten days first.

A huge warm-up is in the forecast for the weekend.  It may surge to near 80º Saturday. Our current forecast is for 74º Friday and 76º Saturday.  Wow! I am looking forward to this nice weather. 

Have a great day! Be sure to watch our weather forecasts on NBC Action News as we will be working on some special graphics to describe these changes!

Gary

Published Wednesday, November 04, 2009 7:30 AM by glezak

Comments

 

Andrew_Stafford said:

Map# 1= shows a zonal flow for next friday? or that for this friday?


Map# 2 = Shows at buckling flow  with a vigorous storms showing up for next friday. Which could equal rain.  

Gary are those the right dates on the maps?

If so the ECMF and GFS have very two different solutions for next friday,
November 4, 2009 7:39 AM
 

twinkiekid said:

What did you expect, it is friday the 13th.  I hope that the European model is correct.  I am planning a scouting event on Saturday and would much prefer a bright sunny day and not the wet cold one that the GFS is implying.  
November 4, 2009 7:52 AM
 

MCIRamp said:

I hope the flow stays zonal all winter long and never stops. I will enjoy that part of the cycle to say the least!
November 4, 2009 8:29 AM
 

RDub said:

Hi Gary, Scott and others....as a follow-up to yesterday, I do have one suggestion for organzing info that you guys have already written up. When a new blog entry is done on lrcweather.com, you could give it subject tags, and set up a way to click on a tag and go to all blogs with that subject.

An example of a blog that does this is www.fivethirtyeight.com...on the right they have a blog archive linked by dates, and then below that is a list of "labels" which can be followed to find posts on certain subjects. You could also do that for this blog....

-------------

Good suggestion Rdub. We will look into this.

Gary

November 4, 2009 8:57 AM
 

hippygoth said:

Personally I don't trust any forecasts more than 3 days out as wild as it's been the last 5 weeks or so... :)

Talking of wild how are we looking from an official number on rainfall this year, we must be hitting close to last years total.

As for LRC, well there are pattens in everything in nature. In that sense I believe LRC is quite viable, now if someone could do me a write up in layman terms, I might have an easier task of really understanding it, beyond the pattern sets up and will repeat within a 35-75 day period. :)

H.

-----------

We will provide more details on this pattern in a way that you can understand as we issue our winter forecast in two weeks.

Gary

November 4, 2009 9:07 AM
 

ChicagoExPat said:

Thanks Gary. It would be nice to see a detailed analysis for the bonafide geeks (and I mean that in the nicest way) as well as a Cliff's Notes version for the rest of us!
November 4, 2009 10:15 AM
 

WINTERSTINKS said:

I am very excited about the possability of a long warm and tranguil pattern fo the first half of novemeber. This should prelcude a fine first half of jan 2010 if the lrc sets up on a 60 day recurinng cycle. If this patterns holds  for the whole month then jan 2010 well be warm, i pray so . sunny and 70 always  

------------

Remember, the flow is much stronger in January and this same pattern could block up and it may prevent it from being warm.  It will be interesting to watch unfold.

Gary

November 4, 2009 10:19 AM
 

Snow Day said:

It's really interesting watching both these models. Both of them seem very consistent with doing their own thing. We also need to look for a building ridge in the Pacific and this could in turn force another trough into the central and western part of the country.

I just dont understand with the technology these days, why cant these models be more accurate.

------------------

No matter what the technology we are still predicting the future. It is amazing how accurate these models are at times.  They are more accurate than they have ever been.  But, when you are talking about 5 or 10 days out they have many compiling errors that send them off in differing directions.

It will probably end up being a balance between the two solutions.

Gary

November 4, 2009 10:36 AM
 

supercell said:

Here's my layman's explanation of the LRC:  

1)  Think of waves.  Ocean waves come in sets.  Surfers have learned how to read the sets so they will know when to expect a good wave.  Sometimes the sets vary, but once the pattern is recognized, an extremely accurate prediction of future sets can be made.  There's no debate ocean waves come in sets.  What we debate is why?  Because we can see the waves and study their movement with our own eyes, we firmly accept the conclusion they come in patterns and not some random variation.  The bigger question is Why?  Much like the LRC.

2) Think of waves again.  This time it's a human wave.  Imagine Arrowhead Stadium with fans doing the wave (how long has that been?).  Sometimes the fans are really into it and the wave can easily be seen making its way around the stadium.  Sometimes the fans are not as energetic and the wave will be strong to start and fade in the middle.  However, we still recognize the wave even if only a few people are doing it because we recognize the characteristics of the wave.  With just a few characteristics of the wave in place, it's possible to accurately predict what will happen with the wave.  Will it die out?  Will it make it around? Why? Because it's a pattern we recognize.  The strength of the wave makes it easier or harder to predict what will happen.  Just like the LRC.

My point?  Patterns are easy to recognize if you look for them.  Not many things are random.  In no way do I believe weather is random.  There is a pattern.  The prediction of what will happen next comes from observation and recognizing what you see.  That's what Gary and his team are doing now.  Observing and looking for characteristics so they will be able to accurately tell us what will happen over the next set.  Will this year's LRC be a long set or short set?  Will we have big waves or small waves?  Will it be 35 days or 70 days between sets?  Just like waves.  Weather has a pattern and someone is working to figure it out.

To me, the bigger question isn't whether or not there's an LRC.  My question is Why?  If someone can figure out what causes the pattern, then long-term forecasting will take a quantum leap.  If you know why, then a forecaster could predict a pattern BEFORE it forms.  As it stands now, we wait for the pattern to form and then predict what it will do.  Imagine taking the next step...

--------------

Wow!  This really is a good way of looking at so many things, especially the LRC and weather patterns.  Great job, and I may post it tomorrow in the blog if you don't mind.

Gary


November 4, 2009 11:41 AM
 

kcwxguy said:

Rdub - working on it.
November 4, 2009 11:45 AM
 

kcwxguy said:

supercell - That is a VERY provocative and intriguing analogy.  Very thought provoking in many ways.  I can see some parallels but see some things I need to think about.

In doing just a quick google search on wave sets, I quickly found this article explaining wavesets a bit more indepth, but still quite readable.

I will let you read it and if you are thinking about the LRC as you read it, there are some points that seem quite familiar.

Not sure if this is any type of answer to any of the LRC questions, but certainly there are many facets that will get the cognitive juices flowing...

http://www.weather.com/blog/weather/8_15920.html
November 4, 2009 12:02 PM
 

supercell said:

kcwxguy- Thanks for sending the link!  Good read by Dr. Lyons.  I did think about the LRC, especially the point that rarely do ocean wave sets form identical.  Each wave set is unique.  Sound familiar?  Good explanation of why they form as well.  It seems consistent with other articles.  Now we need to get to the same level of knowledge with the LRC...
November 4, 2009 12:33 PM
 

Snow Day said:

The ECMWF has now came in line with the GFS;

http://www.ecmwf.int/products/forecasts/d/getchart/catalog/products/forecasts/medium/deterministic/msl_uv850_z500!Geopotential%20500%20hPa!240!North%20America!pop!od!oper!public_plots!2009110412!!chart.gif

What do you think gary?
November 4, 2009 1:12 PM
 

Bizarro Jerry said:

I think Gary just exemplified one of the primary struggles that many of us have with his theory:  He stated that this warm period we are in right now "will return" because it all cycles, but that next time through, instead of zonal flow bringing warmer-than-average temperatures, the flow could block up and essentially do the inverse - prevent warm air from flowing in.

If that's what it could do the next time around, how is there any use to the repeating cycle theory?  What exactly is truly cycling if the zonal flow might be blocked up and not zonal the next time around?  

I think this goes to what some others were asking before:  what exactly is cycling if it isn't the features of the pattern (warm/cold, zonal/progressive, etc...)?

--------------

Bizarro,

But it is the features of the pattern!  The next time through the cycle, this part of the pattern probably won't block up.  But, it will likely be stronger and the jet stream farther south.  So, envisioning what it will do is just one of the challenges we face when making our long range predictions.  When you make your statement "What exactly is truly cycling if the zonal flow might be blocked up and not zonal the next time around?" is a valid question. And, I know what I said.  As you analyze this part of the weather pattern, that is now occuring, it will be quite difficult to block up and be more amplified. This would most likely be the warmer part of each cycle we are about to go through.  But, with this said, the flow will be much stronger in December through February than it is now and there may be a snow pack over parts of the plains. Today's northwest wind may be a much colder one during the winter, and the two waves of clouds and sprinkles we had Monday and Tuesday could have been bands of snow. 

My point is that we strongly beleive it is cycling, and it is regular, and it will repeat. It won't be perfect but we should be able to make some accurate predictions once we "know" the pattern. I am still studying and experiencing, and observing.  I will use a particular year in an example in tomorrow's blog.

Gary

November 4, 2009 1:15 PM
 

supercell said:

Gary,

Sure, post away.  It's the way my mind makes sense of the LRC so maybe it will help others.
November 4, 2009 1:51 PM
 

MTongate said:

can anyone tell me if it will be a colder next week, taking vacation to bow hunt big whitetails and need it to be 50's or lower?
November 4, 2009 2:04 PM
 

ChicagoExPat said:

Supercell, that was an excellent observation about wave sets, and a fantastic metaphor to describe Gary's theory. I agree about the incidences of patterns and randomness -- life has a definite ebb and flow to it. Recognizing and defining patterns take a lot of skill.
November 4, 2009 2:19 PM
 

kcwxguy said:

Rdub - I have finished your request for tags.  Enjoy.

Bizarro - I think it is accurate to say the pattern will return, but its effect at the surface may have different results.  When thinking synoptically, even a flux of a hundred miles makes a huge difference at the surface, while in the broader view may not be much of a difference in the overall pattern reflection.

Also, it is the flow before and after that goes into the pattern/cycle, not just one regime or one area's results.

I understand your point, however.
November 4, 2009 2:32 PM
 

Billy said:

Decent web-briefing talking about El Nino and the area on the Weather service website.  Seems like a good explaination of what climate maps really mean.

http://www.crh.noaa.gov/eax/briefing/

November 4, 2009 5:58 PM
 

juba said:

Lol Gary! I bet Jeremy would do the same thing as Jack if he didn't here your forecast until you put it on air. When was the last time it was 80 degrees, it will be a great day! The trees that aren't all the way dormant yet sure be confused, too bad they can't turn green as soon as the chlorophiyl leaves the leafs. Have a good evening!

Skylar
November 4, 2009 6:29 PM
 

Turd Fergenson said:

Gary. What is your email. I have to ask you a question..

-------------

I did get your email, so you figured it out.

Gary

November 4, 2009 6:41 PM
 

juba said:

Gis email is lezak@nbcactionnewsl.com
November 4, 2009 7:34 PM
 

Turd Fergenson said:

I have it and emailed him
November 4, 2009 7:43 PM
 

Nick Rau said:

Wow, that northeast Pac. storm is a bruiser, strong and big as it and the energy associated with it slams in to north America its effect on us will be the ol' blow torch, yep warm(almost hot possibly) dry air being whipped by gusty winds... must be open burning season in St. Joe, oh and it is;) It may mean that I will have to use some fans again on Sat./ Sun as the house may get a bit toasty, but hey 30's for high's the first week of Oct. and now possibly 80's for highs the first week of Nov. The "averages" are hiding under the covers right now;)

---------------

Nick,

I am going to bed, but I just posted some maps to look at.  Did the ball land in the right spot or not?

Gary

November 5, 2009 12:19 AM
Anonymous comments are disabled

This Blog

Post Calendar

<November 2009>
SuMoTuWeThFrSa
25262728293031
1234567
891011121314
15161718192021
22232425262728
293012345

Syndication

Inergize Digital Media This site powered by Inergize Digital Media. The views expressed here do not necessarily reflect the views of this station.