Wednesday, December 17, 2014

The Different Faces of Central Pacific El Ninos

A few weeks back, I had an email conversation with Professor Jin-Yi Yu of University of California Irvine.  I ran across his website and 2009 El Nino paper on Central Pacific El Ninos and was looking for some additional information on what the difference is between a Central Based El Nino and a "Modoki" El Nino according to his work. If you remember, the current Central Pacific El Nino was and still is a significant variable in our winter weather outlook. This is what he had to say after I emailed him the sea surface temperature analog composite for January asking him if this was a Modoki pattern we've heard so much about in the media:

"...this SST anomaly pattern is a Central Pacific El Nino pattern,
not a Modoki El Nino pattern. The Modoki pattern requires the warming in
the central Pacific to be flanked by cooling in the eastern Pacific and
western pacific
This graphic illustrates the differences between both El Ninos. El Nino Modoki on top, Eastern El Nino (typical El Nino) on bottom. Notice the cooler water flanking the relative warmth in the middle of the Pacific near the dateline.

Dr. Yu's 2012 paper summarizes each of the 21 El Nino events since the early 1950s using three El Nino Indices. The El Nino Modoki Index (EMI Method), the basic Nino method using the Nino3/4 sea surface temperature readings and the EP-CP Index derived by Dr. Yu and his colleagues published in his 2012 paper. I highlighted the seven consensus Central Pacific El Nino events across all three indices.

7 Consensus Central Pacific El Ninos (The paper above describes what thresholds are used in identifying EP or CP El Ninos.)
While the list above doesn't identify specific Modoki El Ninos among the Central Pacific El Ninos, checking the average sea surface temperatures starting in September and overlapping each month through February/March, "Modoki type" temperature signatures (cold pockets flanking warmth near dateline) can be seen.  Below are the seven consensus Central Pacific El Ninos in chronological order. Each evolved differently. Each one has central warmth (2009-10 was the warmest). Only 1958-59, 1963-64 and 1977-78 seem to have the strongest "Modoki" signature (cool pockets east and west) by visual inspection.

One other note, notice the warmth along the west coast of North America in 1958-59, 1963-64, 77-78 and partially in 2009-10. Professor Yu shows in this paper that this warmth is connected to the development of the Central Pacific El Nino. The coastal warmth is associated with the North Pacific Oscillation teleconnection.

MODOKI signature present
Weaker MODOKI Signature. Not as strong as 1958-59

Weaker MODOKI Signature
Strong Modoki Signature
Weak MODOKI Signature
Weak MODOKI Signature

Weak MODOKI Signature
What can we say about these seven events?  The central warmth seems to mature in January. The cool pools east and west were firmly established by October in the first four events (1950s through the late 1970s).

What about our current event?

The central warmth us weaker than 2009-10. The coastal warmth is off the charts due to the North Pacific Oscillation index rising to the highest levels ever for November going back 60 years! The cold pools east and west are showing some signs of developing.  As of mid December, the Modoki signature comparing all three sea surface temperature regions to the other seven events is weak.

The next 6 weeks will be critical in determining whether this Central Pacific El Nino will continue to mature into a formidable driver of the early 2015 winter pattern. Based on past analogs, it should.

Tuesday, December 09, 2014

Don't Be Fooled. Lots of Winter Left!

Winter arrived very early this year across much of the US. Per NOAA, the average contiguous U.S. temperature was 39.3°F, which is 2.4°F below the 20th century average. This ranked as the 16th coldest November in the 1895-2014 record, and was the coldest November since 2000.

US snowcover for the month was the most since 1966, a 48% increase from last (2013) November's snowcover.
November temperatures in northern Ohio were the 8th coldest on record at 37.1 degrees with a stretch of 9 days below 40. November snowfall was the most since 1996. We are ahead of last year  when we count days with at least ONE INCH of snow cover at 12 (only 9 as of Dec 9th last year).
Many people falsely believed that the arctic cold would continue into December. Look at the top 16 coldest Novembers (16th is this year) in the chart below. Then look at the Decembers that followed. The temperature ranks are no where as cold relative to normal. The Decembers that followed had an average temperature rank of....48th!  Only 2 cracked the top ten coldest.

Here is a comparison of November vs December through the 9th.  December temperatures are not as cold relative to normal. The coldest air has lifted back into central Canada.

So what does this mean looking into our winter future?

All of the factors that went into our winter weather outlook are still present and valid. Nothing has changed.

1) Central Based El Nino has developed or is developing depending on which index you use. This central based Pacific El Nino favors colder than normal temperatures across the eastern US. Note the warm pool along the equator and the smaller cool pools flanking it. The cool pools are indications of a Modoki along with warm pool extending along the west coast of North America.
Are their similar years to what we see above?  Yes.  Below is a blend of those similar years. The scale is a bit different but the warm and cool pools are in similar locations and intensities. They're not exact matches but they are close.

2) Warmer than average water off of the west coast along with the cool pool north of Hawaii strongly favor ridging in the western US, troughiness in the eastern US. The Pacific North American Index (or PNA Index) measures this tendency. In most Decembers where the PNA is more than +1, the temperatures across the eastern US are usually below normal.

 The forecast continues to favor a POSITIVE PNA into mid December.
3) The wild cards are always how the Arctic and North Atlantic will behave.  The Arctic Oscillation and North Atlantic Oscillations are indices that measure changes in the pressure patterns over each location. The AO is more of a player earlier in the winter, the NAO during the second half. If the indices are negative, the pressure patterns favor a more variable jet stream with a tendency for more cold air outbreaks.

4) The QBO (another wild card), read about it here, are stratospheric winds over the tropics which alternate direction easterly to westerly every 2 years. These easterly winds have been shown to disturb the Polar Vortex over the arctic making it more unstable. The propensity for cold outbreaks is higher across North America but not guaranteed. We monitor warming in the stratosphere over the arctic. Sudden warming can be a sign of a weakening of the jet stream and colder outbreaks. A good description of these warming events is on this site. Check out this site for more on Sudden Stratospheric Warming events and their potential effects on our winter weather.

Each one of these variables should not be overstated. One factor doesn't make-it or break-it. Our outlook is a blend of the combined influences that each one of these variables bring to the table. All told, our Winter Weather Outlook from early November still stands


Tuesday, December 02, 2014

Signs of Colder Air Returning Before Christmas

You live by your long range outlook sword and die by it.

This past summer and fall, we hinted that the variables present across the globe were conducive for a colder/snowier winter ahead. November big start out of the gate was a big sign that winter will be cold and snowier than average.

This brings me to the present "break" in the arctic air.

High temps across the eastern US trend slightly above normal by the weekend. Typically when November temperatures average as cold as last month did, bounce-backs are not that uncommon in December.

What has me concerned is the Typhoon in the eastern Pacific. Two models have it recurving back to the northeast next week with it eventually getting absorbed into the Alaskan low which boosts the western ridge making eastern US troughs/colder intrusions that much easier.

The American Model also shows the recurvature by December 10th

 The European is the only outlier with a more westerly track.

Remember the old Typhoon Rule? It states that when Typhoons recurve north and east near Japan, this signifies the overall pattern is conducive for below normal temperatures across the eastern US. See my post a few weeks back on this HERE.

The PNA pattern is showing a strong tendency toward positive which indicated the ridge is rebuilding in the west which drops the trough/colder air east.

I ran a composite of all December years where the PNA index was PLUS ONE OR HIGHER and MINUS ONE OR LOWER. You can see how these correlate strongly to extreme temperatures cold and warm.

PNA at -1 or Below correlates to MILDER EAST

PNA at +1 or Higher correlates to COLDER EAST
Sure enough, the European and Canadian models both develop the beginnings of an eastern trough by December 16-17.

So unless the typhoon tracks further west and doesn't recurve and the PNA pattern (which has trended colder across the US for a while now does a complete 180 degree flip), I would expect temperatures to trend colder the week before Christmas with higher chances of lake effect snow.

Wednesday, November 26, 2014

East Coast Thanksgiving Snow Update

East coast storm continues to churn northeast up the coast. Current radar loop below.

Central Great Lakes sector loop

I'm sticking with National Weather Service snowfall forecasts at this stage of the game. Heavy snow should stay inland away from major cities like New York City and Boston. Philadelphia will see the most accumulation.

Tuesday, November 25, 2014

Major East Coast Snow For Thanksgiving

Significant east coast storm will impact travel tomorrow and Thanksgiving. Expect significant flight delays out of all major hubs tomorrow and early Thursday.

Central Great Lakes sector loop

Probability of at least 6 inches of snow is greatest from Allentown, PA northeast into southern New Hampshire.

The newest projections are bringing the heaviest snowfall further west.

National Weather Service snowfall forecasts.


NWS Philadelphia
NWS State College, PA
NWS New York City
NWS Binghamton
NWS Boston

Washington, DC area

Central PA
Central New York

New York City Metro

Boston, Connecticut, Rhode Island

Wednesday, November 19, 2014

Buffalo Lake Effect Events: Oct 12-13, 2006 vs Current Event

I remember the October 2006 lake effect snow event well. I was sitting in the WJW FOX 8 weather center watching a band of snow moving parallel to the Lake Erie shoreline about 15-20 miles offshore near Ashtabula, Ohio.  The leading edge of the snow barreled through Buffalo, New York.

This is the only lake effect event comparable--when looking at the temperature parameters--to the current Buffalo event I can remember. It started on October 12, 2006 and finished the next day.

The Lake Erie water temperature was 62 degrees. The air aloft (5000 feet)--the key level for determining lake effect--was -8 celsius (or 17 degrees F).

The temperature difference: 45 degrees from the surface to 5000 feet.

Here were the 2 day totals:

In the current (November 18-19, 2014) event, the lake water temperature in the eastern basin of Lake Erie is between 50 and 54 degrees. The air aloft (5000 feet) was -18 celsius (or zero degrees F)! 

The temperature difference: 54 degrees from the surface to 5000 feet!

This eastern basin warm pocket fueled the snow development as the wind paralleled the shoreline before it made landfall south of Buffalo. The key has been the wind direction. The SW wind (now WSW) carried the arctic air directly over the ENTIRE warm pool of water. The fetch over the warm pocket was more than 100 miles! Normally in northern Ohio, the fetch is between 50 and 60 miles from Canada to Cleveland.

Notice the intensity and uniformity of the lake effect snow band compared to the "edgier, disjointed-looking" radar loop from the October 2006 snow event

November 18, 2014 Event

October 12-13, 2006 event
This type of cold in November (850mB temps at -18) has occurred a dozen times in Buffalo (upper air soundings) since 1950, the most recent was last year.
Snow forecast map from the National Weather Service office in Buffalo.

Snow totals as of mid evening Tuesday 11/18. Final snow totals coming later in the week.