Analog96 Winter Forecast

2014-15 WINTER FORECAST
The calendar says October, and the amount of days left in the month are getting checked off one by one, and that means two things: a) winter is on the way, and b ) some people (like me), have to make a winter forecast. So, let’s start first with a review of last winter and last winter’s forecast. Last winter was COLD. I think that goes without saying. Although other winters had more extreme cold outbreaks, last winter’s continuous cold will be remembered. It was also very snowy. Although as we got later into winter, the storm track shifted further south and left our area high and dry, albeit cold. My winter forecast on temperatures was a resounding success, as I called for very cold temperatures across the Eastern United States. My precipitation forecast of near normal also did well, but it snowed more than I expected. Of course, forecasting near record snowfall that far out is not prudent, and snowfall is by and large the craziest variable to forecast. Consider this: You could have a completely snowless winter, and get one large snowstorm and be almost at your seasonal normal snowfall in New York City. Also, what if one or two of the storms last year missed or had more rain than snow? All these wild cards make seasonal snowfall forecasting a huge wild card. But all in all, I was very happy with my winter forecast last year.
Now, let’s start to look at this year’s forecast. First off, I am not nearly as confident one way or the other as to this year’s forecast. My reasoning is that last year’s upper air patterns were fairly steady all fall (of course, they had oscillations, that is the definition of “transition season”), but they were more steady, by a lot, than they are this year.
One of the most important pieces of the puzzle that I examine when making a winter forecast is the upper air pattern. The pattern between 850 mb and 300 mb can often foretell the weather months in advance. For those of you who don’t know, 850 mb and 300 mb are pressure surfaces. Sea level pressure is usually around 1010-1015 mb, but can obviously vary depending on whether or not there is high or low pressure present. Pressure lowers as altitude rises, so 850 mb is generally around 5,000 feet, 500 mb is generally around 18,000 feet, and 300 mb is generally around 30,000 feet. The 850 mb level becomes very important, along with the surface level, in the winter, for determining precipitation type. The 500 MB level is possibly the most important for determining a weather pattern going forward, because at that height, you are above the frictional effects land can cause on modeling (unless you’re at Mount McKinley)! The 200-300 mb level is also very important, because that is where we typically find the jet stream.
A term you’ll typically see me throw around in this winter forecast is “analog years”. An analog year is a year which had similar patterns to this one, and thus, is being used for one of the years of this winter forecast. Keep in mind, an analog year does not mean that every day of the analog year was the same as every day of the coming winter is going to be. If forecasting were that easy, we could just run the analog winter’s historical weather and forecast the same conditions every day of the coming winter. It isn’t that easy, but the purpose of analog years is to show the general conditions over the course of a season and expect that this season would be similar. If there was consensus that most of the analog years had a big snowstorm in February, for example, one could argue that there was a heightened chance of a big snow in February this year. But getting any more specific than that is just crazy!
All that said, these years had the closest upper air pattern to this fall: 1953-54, 2008-09 2005-06, 1979-80, and one year had the best match of all: 2003-04. The next thing I did was check the upper air analog years against the ENSO state of the analog fall vs. the ENSO state this fall. ENSO is the El Nino Southern Oscillation, what you typically hear of El Nino and La Nina. We are currently in an ENSO-neutral state, as we have been for the last two years. However, there are indications that ENSO is trending toward positive, or El Nino, albeit weak. Since the trends are weak and slow, if a year was neutral or weak negative, it can be included on the list, but the list should lean toward El Ninos, and really should not include any La Ninas, or any very strong El Ninos. After I checked this, I eliminated 1953-54, 2008-09, and 2005-06 from the list. Although 2008-09 was a generally neutral winter, it was more negative than what we are going to see this winter. 2003-04 and 1979-80 saw very similar ENSO profiles, so they made the “next cut”.
Speaking of the aforementioned ENSO profiles, when I went and checked our current ENSO state vs. historical ENSO profiles, the following years matched well, in addition to the two that I mentioned, included: 1962-63, 1967-68, 2001-02, 2006-07, and 2012-13. Now, even though these years may have had similar ENSO profiles, we have to check to see if the upper air profiles of those years matched this fall. If the upper air profiles don’t match at all, the ENSO profile could be exactly the same, and the analog is trash. Of the years listed, the only one that had a matching upper air profile at all was the fall of 1967, so 1967-68 was added to the analog composite. The other years did not have a matching upper air profile, and were tossed.
The final index to check is the QBO index, or the Quasi-biennial index. It is important to note that there are very few years we can add to the composite from the QBO index, because very few match this fall’s, and the two best matches were La Ninas. The closest match was actually 1974-75, but that was a strong La Nina, and I can’t use a strong La Nina in a coming weak El Nino winter. 2003-04 was a decent match, and that was already a good analog for this winter. The only other analog for the QBO was 1970-71, which was also a La Nina, but a fairly weak one, so I added it as a weak analog for this winter.
The last thing I assess is the tropical season. There is disagreement in the meteorological community as to whether or not the past tropical season has any connection to the upcoming winter. I personally believe it has to, because tropical seasons are influenced by the same factors that influence winters. A very close tropical season to this winter was the tropical season of 2009. 2009-10 also had a very similar upper air setup, so that was added to this year’s list.

To sum it up, the strongest analog by far and away for this winter is 2003-04. Another fairly strong analog is 1979-80. Other decent analogs that were included for this forecast are 1967-68, 1970-71, and 2009-10. So what does all that mess look like in maps? Glad you asked!



As you can see, most of the country is cold, except the Northern Rockies and the West Coast. The strongest cold anomalies are centered over the Eastern and Southeastern parts of the country.

Next, we’ll look at precipitation. Precipitation maps are a good way to figure out the overall storm track in a winter.



From this map, the best inference I can make of a storm track is mostly Miller A in origin. A Miller A is a Gulf of Mexico, then up the coast storm track. The secondary storm track is probably from the Pacific Northwest, through the Northern Plains with some Miller B (redevelopment) off the mid-Atlantic Coast.

It should also be noted that in the winter, there are typically two branches of the jet stream: The Northern Branch, which typically delivers the cold air to the United States, and the Southern Branch, which typically transports moisture from the Gulf of Mexico. Storms can track along either branch of the jet stream. Storms that track along the Northern Stream are generally drier, but can sometimes deliver more moisture if they can develop a secondary low near the East Coast and tap into Atlantic moisture. But historically, our biggest snowstorms are when the two jet streams can come together and produce phasing- bringing the cold air, the moisture, and lots of energy for storms to work with. Last year, we relied mostly on Northern Stream systems. This is why although we had Decea lot of snow and frequent storminess, we really didn’t get any “huge” storms. If a weak El Nino develops, which appears is underway and is forecast to happen, then it would be likely that the Southern branch of the jet stream would be more active this winter, thus increasing our chances of a “big one” drastically.
Finally, I am going to experiment with something I never have done before. I am going to run the analog set monthly, to try to give a month-by-month forecast for this winter, from December through March. I have never done this before, so at least for me, it should be considered experimental. But I personally wanted to try it and see if it’s worth the effort. So, here goes:

December: Temperatures: Near Normal
December Precipitation: Above Normal
December Snowfall: Slightly above Normal
January Temperatures: Much Below Normal (Possible big-time Arctic outbreak)
January Precipitation: Below Normal
January Snowfall: Slightly above Normal (mostly from small events)
February Temperatures: Much Below Normal (Possible SECOND major Arctic outbreak)
February Precipitation: Normal (could be a good event in there somewhere)
February Snowfall: Above normal (good chance at a long run of snowcover in Jan-Feb).
March Temperatures: Slightly Above Normal
March Precipitation: Above Normal
March Snowfall: Above Normal (winter could end with a big one in early March).

So, there you have it, the winter forecast for the winter of 2014-15 is hot off the presses! Let’s hope this verifies!

PETRIDIS

Ok for some reason my maps didn't post... let me try that again.

Can someone help me?
Why do my maps not post?
They are saved as pictures.

Greg,

I like the Winter outlook and thanks for posting - take this in a frickin heartbeat!!

Call up the file under host an image and make sure to resize it. That should do it.

Mugs

Idk, I have tried a few times to host it, and the whole process goes through, then the post remains blank.

Great write-up Greg! It will be interesting to see how your month-to-month forecasting goes!

Greg,

I would host the images to a third party site like tinypic.com or postimage.org. For some reason uploading from your desktop is not showing.

Analog first thanks for your write up...really good stuff.  regarding your images try this if you havent already.  
" />

First: click on the icon labeled 1.  Next left click browse or choose image.  Find the image you have saved on your hard drive click on it then click open, Once loaded into the que click host it.  (Note: you should consider saving the images you want to post to a file on your desk top so you can find it easy when browsing for it during this step.  If you are simply trying to post it directly from a website the way Im describing is not the way to do it)  

After clicking host it something like this should load with the URL next to each one.  

Thumbnail :   blah blahblah
Image :   blahahblahblahh
Image url :  bl ahblahbla hba

I next left click anyone of the URL.  (I usually choose the middle or bottom one.  I don't know if it makes a difference)  This should highlight the entire URL.  Once the entire URL is highlighted Right click it and click copy, then click upload.  They all should then disappear.

Next, using the image above again, click on the icon labeled two and simply paste the prev copied and uploaded URL into the space labeled "URL" and click insert.  That should do it.  You can click preview first before you send it for all of us to see to make sure it works.  

See how easy it is I hope we have at least one of these this winter..in case you don't know..its a Frankzilla storm.  It will only come out when over 2 feet fall.

 " />

HAHAHAHA, oh Scott. You made a funny

I knew your like that Frank. Lol