Thoughts on Winter 2010-2011

Long range forecasting is a bit of a crap shoot. A 7 day forecast can be a challenge so a 1-4 month forecast can be even harder. That said there are some clues that we can use to piece together a prediction for the next couple months.

Current ENSO State
The biggest driver of winter 2010-2011 will be a strong La Nina over the equatorial Pacific. Cold sea surface temperature (SST) anomalies have developed and remain very impressive. It seems likely that this will be a moderate or strong La Nina event and will have a substantial impact on this upcoming winter.

This La Nina event is basin wide with negative temperature anomalies from South America all the way west past the International Dateline. This La Nina is coming on the heels of an impressive El Nino event which ended during the Spring of 2010 and model forecasts indicate that it could strengthen further.


La Nina’s impacts
La Nina generally feature a variable Pacific jet stream with a ridge in the southeastern United States. This promotes a drier than normal winter (DJF or December, January, February) in the southeastern part of the country and a wetter than normal winter across the Great Lakes. The storm track tends to be from the southern/central plains through the Great Lakes with a reforming low off the Mid Atlantic/New England coast. In general precipitation is close to average in New England.

The “typical” La Nina pattern becomes exaggerated during a moderate and strong event. Here are the temperature differences between a weak La Nina and a moderate or strong La Nina (using an ONI of -1.0 as a cut-off).


If you look at how strong La Nina winters tend to progress they start near normal or slightly cooler than normal on the east coast while they’re much warmer in the central plains and the northern Rockies. Notice the December warm anomalies are mainly centered in the middle of the country while a cool signal exists here in New England and the middle Atlantic.

However, by January and February a substantially warmer pattern begins to take shape across the east with notable positive temperature anomalies here in southern New England. It’s also important to point out that the only 2 years in the La Nina sample that have gone from strong El Nino to strong La Nina are 1973-1974 and 1998-1999 which were both exceptionally warm winters over almost all of the country. That’s obviously way too small of a sample size to draw any big conclusions but something to keep in mind.

Dynamical and statistical ENSO forecast models indicate that the La Nina event will hold its own (some even have the La Nina strengthening further) during the DJF time frame. Subsurface buoy data shows this La Nina is firmly entrenched and is going nowhere fast. Using a moderate/strong La Nina as a basis for this years winter forecast is probably a good start. From there it’s important to try to add some value by looking at other variables. Most important for New England will be the presence (or lack thereof) of high latitude blocking.

La Nina also features a stronger and colder than normal polar vortex over the North Pole. This makes can make the vortex more difficult to disturb and dislodge and keep cold air locked to the north. I’ll discuss this more in the next section.

Blocking (QBO/Solar/AO/NAO)
The Quasi-Biennial Oscillation is an alternating easterly and westerly wind regime in the stratosphere over the equator. The oscillation is between 24 and 30 months and can have significant impacts on the global circulation. Holton and Tan, 1980 showed that the equatorial wind oscillation in the stratosphere modulated the global circulation and that the stratospheric polar vortex is stronger during a +QBO (westerly) regime during northern hemisphere winter.

In papers since (especially many by Baldwin and Dunkerton) a link between QBO, the polar vortex, and high latitude blocking have been established. Baldwin and Dunkerton (2001) found that stratospheric warmings are much less likely during a +QBO phase while a strong PV is nearly 3 times as likely during a +QBO and a +NAO/+AO is more likely during a +QBO phase.

The link between +QBO and a stronger polar vortex was showed to be even stronger when solar forcing is low (Gray et al., 2001). More research showed that during a solar minimum/+QBO there is no favored time for a sudden stratospheric warming because the easterly wind anomaly in the upper stratosphere partially cancels the westerly wind anomaly in the stratosphere from the +QBO.

Given that the QBO is becoming increasingly westerly (during a solar min) it seems likely the polar vortex will be stronger and less disturbed than normal (stratospheric warming helps disturb the PV). This will lead to a winter that features less blocking than normal and likely a +AO/+NAO on average.

There are some signals showing that this winter won’t be a total loss, however. For one, the NAO has been substantially negative since March of 2008 with minor positive blips. In fact the NAO has been negative for the last 12 consecutive months! We appear to be in a negative decadal NAO oscillation which may provide us with a lower “base state” for this year’s NAO which should help the NAO from average too high.

This should ensure temperatures are above normal during most of the upcoming winter, possibly substantially so for a period of time.

Pacific Decadal Oscillation
Sea surface temperatures outside of the equatorial Pacific can be important as well. The PDO is an oscillation that results in changes in sea surface temperatures in the northern Pacific Ocean. The current SST anomalies show a classic -PDO look with cold anomalies over the Gulf of Alaska and across the west coast of the United States while there are warm anomalies south of the far western Aleutian Islands and off the coast of Japan.

The -PDO pattern can result in a large Gulf of Alaska trough (the exact location of this can be very important) and a large ridge of high pressure over the Aleutians. All of this teleconnects to a classic La Nina pattern with a big ridge of high pressure in the southeastern United States which can keep most of the east coast warm.

Madden-Julian Oscillation
The MJO is an oscillation in the tropics that propagates eastward and can result in areas of enhanced or depressed convection and precipitation in the tropics. Certain phases of the MJO teleconnect to certain weather patterns over the United States. Though the MJO is a good way of diagnosing weather patterns in the medium range, using it as a long range tool is not as useful.

However, there are some signs that we can use to predict where MJO forcing will most likely set up this winter (particularly in the next 1-2 months). Strong ENSO events (either La Nina or El Nino) generally result in less MJO activity and MJO waves frequently don’t totally cycle all the way through their 8 phases. This year with a strong basin-wide La Nina the colder than normal sea surface temperatures near the International Dateline may make it more difficult for the MJO to cycle into phase 7 or 8. Those phases tend to correlate to colder than normal and stormy weather in New England.

Given the current warm anomalies near Indonesia and cooler than normal water temperatures near the International Dateline it appears MJO forcing will remain prevalent in MJO phase 4, 5, and 6. In December, phase 6 actually correlates to cooler than normal weather in the northeast, but by January and February all 3 phases (4 through 6) have a decidedly warm look in New England.

In fact in the last 6 weeks phases 4, 5, and 6 have been favored and I expect some more MJO activity in these quadrants over the next 2 months. Though this isn’t a huge player in my winter forecast I think it helps reinforce a few other variables I’m looking at and all of these things are related to one another anyway.


Putting It All Together
It’s quite clear that La Nina will be a major driver in our weather this winter. In general I think a typical moderate/strong La Nina pattern will develop across the US during the December-February time frame. It’s possible that La Nina will begin to wane by the end of winter but by that point the pattern will be already established (and there’s a lag between a change in SSTs and a change in the global circulation).

La Nina climatology argues for a cool to near normal December followed by above normal temperatures in January and February. I’m forecasting at or below normal temperatures to start out the winter followed by much warmer weather for January and February. MJO activity that meanders into phase 6 supports this as does a lingering -NAO signal (and even a weak -QBO wave downwelling in the lower stratosphere).

Many winter-lovers are banking on a decadal negative NAO trend to “save” the winter. I think this is flawed. Once the wavelengths change in winter the QBO/solar factor is hard to ignore. This should flip the NAO to somewhat positive but I think a raging +NAO is unlikely. That said, the strength of the Nina and the -PDO should help build up the southeast ridge that will flex its muscle and bring a general warmth to southern New England. Any periods where blocking does develop, however, could feature substantial cold if the polar vortex gets displaced. This will be the exception rather than the rule.

Here’s the bottom line:

Windsor Locks (BDL)

  • December: -1.0
  • January: +3.0
  • February: +2.5
  • Snowfall: 35″ (Average 46″)

Boston (BOS)

  • December: -1.5
  • January: +2.5
  • Feburary: +2.0
  • Snowfall: 35″ (Average 42″)

New York City (CPK)

  • December: Normal
  • January: +3.5
  • February: +3.0
  • Snowfall: 15″ (Average 22″)
Advertisements

One thought on “Thoughts on Winter 2010-2011

  1. Hi Ryan – Just came across your website/blog and particularly this post. I am a climate dynamicist, so I am always interested in seeing people’s thoughts on seasonal forecasts. This is in no way a post from someone who wants to say “Bad Forecast!” (not my style). This winter, like last winter, were challenges to the people who rely almost exclusively on ENSO. The problem, I believe is two things:

    (1) Changing nature of ENSO – I did my dissertation on North Pacific and tropical Pacific decadal climate variability and climate change. One of the more underreported stories on ENSO recently is the changes on the types of warmings occurring in the Pacific. Our traditional views of ENSO are for the warmest waters to be in the far eastern tropical Pacific (along the Peruvian coast). In 2009-10, the warmest waters instead were in the Central Tropical Pacific. This new type of ENSO, called Central Pacific El Nino or El Nino-Modoki, has very different teleconnection patterns with the Northern Hemisphere wintertime climate. In fact, if you are interested, I can send along composites of surface air temperature, sea level pressure, etc. for CP El Nino vs. EP El Nino. The differences are remarkable across North America.

    (2) Siberian snow cover – Over the last 20 years, October snow cover extent over Eurasia has been increasing (i.e., a linear trend). The enhanced diabatic cooling from this snow cover increases the strength of the Siberian high, which amplifies the standing wave over Eurasia and the Pacific. The amplification of this wave yields stronger vertically propagating waves, which then shifts to disturbing the polar vortex and breaks it down more effectively. The changes in the polar vortex descend through the column, generating a -AO response from the stratosphere to the troposphere. As we know, the last two seasons have seen incredible -AO conditions in December and January. Last year, February also had it. This is something again that is not considered by seasonal forecasters.

    Just some tidbits from my end. I love seeing people getting more and more into large-scale climate modes!

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s