Monday, November 28, 2011

Winter '11 / '12 - Arctic Oscillation and Eurasia Snow Cover

New research from Judah Cohen of sCast fame shows a strong correlation between the rate of change in Eurasia snow cover during October and the state of the Arctic Oscillation (AO) during the coming winter.

The paper detailing the results of the study were published in the 05-NOV-11 issue of Geophysical Research Letters Vol 38.

The "...develop(ment) a snow advance index (SAI) derived from antecedent observed snow cover...explains a large fraction of the variance of the winter AO."  The new index follows work relating monthly areal snow cover to NOAM temperatures and the implicit implication for the state of northern annular oscillation modes (AO / NAO).  The snow advance index (SAI) is the regression coefficient of the least square fit of the daily Eurasian SCE equatorward of 60°N calculated for the month of October.

Note how the observed AO and surface temperature values for meteorological winter (D-J-F) are correlated in c) above.  The positive correlation in the eastern CONUS (red) means when the AO is positive...temperatures are warmer (positive anomalies) and vice versa.  When the AO is negative...temperatures are colder (negative anomalies).

A strikingly similar depiction appears in d).  Cohen's new snow cover index derived in October is highly correlated (r = 0.86) with temperature anomalies in the eastern CONUS and all but mirrors the observed AO / T correlations during D-J-F.

Cohen and Jones...
"The implications of this discovery are potentially significant. Currently the AO is considered a product of the stochastic behavior of internal atmospheric dynamics and therefore chaotic.

"The fact that we discovered a single predictive index that explains close to 75% of the variance of the winter AO (though the period is short and the degraded SAI over a longer time period explains less of the AO variance) is inconsistent with this thinking and demonstrates that the AO, while thought to be unpredictable, may in fact be one of the most easily predicted  phenomenon known in the climate system.  [emphasis added]

"Even the most sophisticated GCMs achieve only marginal skill on the seasonal time scale in the extratropics. Implementation of the SAI in winter seasonal forecasts could potentially be a sea change in operational seasonal forecasts." 
Full paper here (.pdf 920 kb).
Harvey Leonard's (WCVB Boston) interviews with Dr. Cohen here and here.

Thursday, November 24, 2011

Winter '11 / '12 - Great Appalachian Storm of 1950


Kevin Myatt at the Roanoke Times reminds us of a blockbuster snow storm from a Thanksgiving weekend long ago. 

Known at the time as 'The Storm of the Century' ... Miller 'B' cyclogenesis occurred along shore of the Carolinas on the 24th ... rapidly deepened (26mb / 12hrs) over the Chesapeake Bay  / Delmarva on the 25th...then became trapped beneath a strong 1042 mb blocking HIGH to its north.

Kevin notes...

"The Great Appalachian Storm of 1950 left behind 9 inches of snow in Roanoke and 14inches in Wytheville, with 4 to 12 inches common across our region (ED:  SW VA).

"But it was far worse elsewhere. Parts of West Virginia, eastern Kentucky and western Pennsylvania saw 2 to 4 feet of snow. Winds topped 100mph in New York. The storm is blamed for more than 300 deaths and about 1million people cut off from electricity."
Some other interesting minutia about this historic early season snow storm...

Snowfall totals...
Youngstown...OH - 29"
Pittsburg...PA - 30"
Elkins...WV - 34"

Notable wind gusts were observed at...
Newark (EWR)...NJ - 108 mph
Bear Mountain (N of NYC) - 140 mph
Concord (CON)...NH - 110 mph
Mount Washington (MWN)...NH - 160 mph

Other storm effects...
- Coastal flooding breached dikes and flooded runways at LaGuardia Airport in New York.
- Crop damage and record minimum temperatures in the deep south (Birmingham:  5°F; Atlanta:  3°F;  Nashville:  -1°F)
- Some coastal areas in New England reported greater damage than the famed '38 hurricane.
- "...(T)he Ohio State-Michigan football game went on as scheduled (in Columbus), despite blizzard conditions. Nearly 50,000 fans showed up to watch the “Blizzard Bowl,” with a Rose Bowl berth the payoff. Michigan won the game 9–3 without making a single first down and only gaining 27 yards on offense." (Weatherwise Mar-Apr 2011)

Charts and graphs after the jump.

Tuesday, November 22, 2011

Winter '11 / '12 - The North Atlantic Oscillation and Sea Surface Temperatures

Conventional wisdom states it's next to impossible to skillfully forecast the state of the North Atlantic Oscillation (NAO) more than a week to ten days in advance.

Even that's a stretch given the generally low correlation between forecast and observed conditions.  The current ten-day and 14 day correlations are 0.72 and 0.55...respectively which means the forecast explains 52% and 30% of the NAO's variability.  There's some skill at ten-days...not so much at 14.

The UKMET Office produced an NAO winter-season forecast up until the fall of 2009.  It was based on sea surface temperature anomalies (SSTA) in the mid-Atlantic Ocean.  They claimed a 67% skill level.  It's no longer published...possibly b/c of recent controversies and adverse publicity surrounding their 'failure' to forecast the severity of the past two winters.

Image courtesy UKMET Office
"By taking the observed SST anomaly for May and calculating how it projects onto the predictor pattern we (ed:  UKMET Office) can make a prediction for the winter NAO.

"If the projection is positive (i.e., the anomaly pattern looks similar to the predictor pattern shown above) then the prediction is for a positive winter NAO.

"Conversely, if the observed May SST anomaly projects negatively onto the predictor pattern (i.e., it looks like the reverse of the predictor pattern) then we (ed:  UKMET Office) would predict a negative NAO."
---
So...where were SSTAs lurking in the Atlantic this May?
What...if anything...might they predict about the state of Winter '11 / '12 NAO?

If you're rooting for Team NAO to come through and deliver a halfway decent winter...then you can't be happy about the positive-negative-positive SSTA pattern along the northwest rim of the Atlantic Ocean last May.

The observed pattern predicts low 500mb heights will prevail over Greenland this winter and therefore...a predominantly positive NAO.
---
More about SSTAs and the NAO during winter in a few days.

Thursday, November 10, 2011

Winter '11 / '12 - Sunspots and The North Atlantic Oscillation

From 'Bits of Science'...

"Recent Nature Geoscience publication shows correlation between sunspot cycle and the AO and NAO index. This in turn influences northern hemisphere temperature distribution. Current NASA observations meanwhile show the 2008-2010 sunspot minimum has come to an end – shown NASA forecast predicts new peak around the year 2013."

"Judging by solar activity a pattern of consecutive cold winters [2008-2009, 2009-2010 & 2010-2011] has come to an end. High pressure blockades (over the North Atlantic, Scotland or Scandinavia) are therefore unlikely to prevail and for Western Europe there could even be another episode of some 5 or 6 years of mild and rainy winters ahead.

"The theory is quite straightforward – and has been known to climatologists for many years: during the winter months solar activity influences air pressure patterns over the Arctic and lower northern hemisphere latitudes.
"In this the number of sunspots is key. When there are few, northern hemisphere westerlies are weaker – and cold air is capable of escaping from the Arctic towards for instance the USA and Europe."
More here.

Tuesday, November 8, 2011

Winter '11 / '12 - Late October Snow - NESIS


NCDC has given the late October snow storm a preliminary score of 1.75 (Category 1) on the NESIS scale.  This is the only October storm to have a NESIS rank.

Multiple daily snowfall records were set during this event (29th - 44; 30th - 103).  Many stations recorded their first ever snowfalls in October.  Still others received the highest daily October snowfall in their recorded history.

The long wave flow regime changed promptly in the storm's aftermath from ridge west-trof east to trof west - ridge east.


Once again...a negative NAO was missing in action during an historic east coast snowfall.
Daily AO...NAO...and PNA indexes here.

Related post:  October Snow Cover

Thursday, November 3, 2011

Winter '11 / '12 - Eurasia Snow Cover - October

The latest monthly data from the Rutgers Snow Lab shows the observed areal snow cover over Eurasia was normal during October.


The chart's orange line marks the month of October's long-term average snow cover.  The dashed yellow lines are one standard deviation above or below the average.

From a statistical process control (SPC) perspective...little variation above or below the mean has occurred in the region since the early '90s.  The early '90s marked the end of a 'warm' convergence resulting from predominantly +ENSO and +PDO conditions in the tropical and extra-tropical Pacific ocean that began in the mid '80s.

Long range seasonal forecasting schemes based on autumnal Eurasia snow cover have received a fair amount of attention in recent years.  The working hypothesis centers around the idea that 'above normal' snow cover in this region during the early fall is a strong leading indicator of colder than normal winter surface air temperatures over the eastern CONUS.

These colder surface air temperatures are thought to be a consequence of upward propagating Rossby waves -- initiated by above normal Eurasia snow cover -- entering the stratosphere which...in turn...weakens or reverses the polar vortex (PV).  A weaken PV is a favorable condition for negative index values of the northern annular oscillations (AO; NAO) which generally lead to below normal temperatures and increased storminess over the eastern CONUS.

The current state of Eurasia's autumnal snow cover does not suggest the eastern CONUS will experience below normal surface air temperatures and/or increased storminess during the upcoming winter. 

Areal snowfall data courtesy Rutgers Snow Lab.

Tuesday, November 1, 2011

Winter '11 / '12 - Season-total Snowfall Forecast Contest - Call for Forecasts!

Central Park - NYC - 1958
NE.Wx's 11th annual ‘Season-total' Snowfall Forecast Contest is the absolute best...biggest...and probably ONLY chance to be recognized for your long-range forecasting acumen.

All you have to do is forecast the season-total snowfall at 25 stations from RDU to CAR.

Deadline: WED...30-NOV-11 @ 11:59 PM EST

Visit the Contest's website to enter your forecast.  Follow the link from 'Enter Season-total Forecast.'

As always...there are NO costs...fees...advertising...or annoying requests for personal information to enter the contest.
---
Forecast element:  sum-total season snowfall @ each station
Forecast period:  01-DEC-11 through 31-MAR-12

Verification:  NWS preliminary climate reports (CLM or F6)
Error statistic:  total absolute error [Σ abs(forecast - observed)]

Update your forecast as often as you want.  Only your last entry will be verified.

---
The reigning NE.Wx ‘Season-Total’ Snowfall Forecasting Champ-een is Roger Smith.  Last year's 'Season-total' forecast summary...verification...and final results ==> here.

---
The Contest is open to amateur and professional forecasters; broadcasters with or without trained Seals; any and all weather-biz types and / or wanna-bees; wish-casters...astrologers...along with any other universally recognized classes of dreamers; Pollyannas or Cassandras...registered Nostradamusts...non-violent megalomaniacs...woolly-bear caterpillars or their agents...pest detectives...NE.Wx NG regulars and lurkers; refugees from AmericanWx and/or USWeather...including self-imposed exiles from Eastern...and of course...meteorologists.

In honor of ne.weather's patron saint Mr. Joseph Bartlo's final request (RIP)...trolls...goats...hat3-lsiters...and psests need not apply.