Thursday, December 31, 2009

Winter '09 / '10 - Snow Storm #2 - Call for Forecasts


Two weeks to the day since the last contest-worthy snow storm. Latest NWP forecasts indicate this event will affect mainly northern forecast stations over a two-day period as the surface LOW meanders around the Gulf of Maine.

This storm won/t spawn from the typical winter wx regime where the mid-upper level westerlies amplify...the wavelength shortens...and a nor-easter gains latitude as it comes up the east coast. If anything...the long-wave pattern is de-amplifying...evidenced by decreasing PNA index values. This weekend/s storm is a child of a down-stream Rex block formed by a strong...closed anti-cyclone over Greenland. HPC mentions 'phasing' a few times in HPC/s afternoon 'Heavy Snowfall Discussion' (HSD)...so...a call goes out for forecasts.

-----
Deadline: 10:30 PM EST FRI...01-JAN-10

Forecast element: storm-total snowfall

Verification begins: 12:01 AM EST SAT...02-JAN-10
Verification ends: 11:59 PM EST SUN...03-JAN-10

Enter your forecast at the Contest web site here.
Follow the link from 'Enter Storm Forecast.'

As always...there/s no cost...or fee...or annoying requests for personal information to enter.  Just a fun exercise to see who can make the best synoptic-scale snowfall forecast.

If you are issuing your first forecast this year...or you entered the 'season-total' forecast contest...you/ll need to create an account (user name / password / valid e-mail...if you want a copy of your forecast sent to your Inbox) before submitting your entry.



Image:  Millville...NJ

Saturday, December 26, 2009

Winter '09 / '10 - ECMWF/s 'Happy New Year' Xmas Present or Coastal Teaser?

Arctic air everywhere.  Happy New Year!

Photobucket

Center of upper LOW progging farther S+W w/time ==> surface development @lower latitude.
Trof appears to take on a negative-tilt.

Fits the pattern.
Don't want it looking too good this far out.

Friday, December 25, 2009

Wednesday, December 23, 2009

Winter '09 / '10 - WSI - December Update 2

Dr. Todd Crawford - WSI seasonal forecaster:
""The combination of the current El Niño event, cold north Pacific, and weakened stratospheric vortex are favorable for a continuation of widespread below-normal temperatures across the US for the upcoming season"...

""There may be a relaxation of the current cold pattern in the Northeast during January, followed by a return to more consistent cold in February and March."

"In January, WSI predicts regional temperature anomalies as follows:
Northeast - Warmer than normal, except PA/NJ/DE

"In February, WSI sees the monthly breakdown as:
Northeast - Colder than normal

"In March, WSI forecasts:
Northeast - Colder than normal"

The stratospheric vortex began to weaken and split late last month continuing into early DEC; however...the two PV circulations have since merged and the upper level anticyclone has dissipated.  The effects of the weakened PV took about three weeks to reach the bottom of the troposphere...as evidenced by near-record low AO index.

Today/s 10 mb geo-potential height / air temperature analysis (courtesy U of WY) shows a strong... consolidated PV centered near the pole.



Tuesday, December 22, 2009

Winter '09 / '10 - Snow Storm #1 - Final Results

NEWxSFC Chief forecaster donsutherland1 picked up where he left off last winter with another top-ranked synoptic-scale snowfall forecast.

His 27 station forecasts verified with 344.2 SUMSQ error points (Z-Score = -0.915)...a 67% improvement over the average error of 1056".  The 220" storm-total precipitation (STP) forecast for all stations came within 5% of the observed 213.4" snowfall.


Donsutherland1/s issued perfect forecasts @ BOS and BGM.

Congratulations donsutherland1!





2nd Place:  Donald Rosenfeld (SUMSQ error: 372.9"; Z-score:  -0.878)
3rd Place:  Newa2010 (379.16"; -0.870)
Honorable Mention:  Herb @ MAWS (383.2; -0.865)

Full forecast verification and results summary here.

Winter '09 / '10 - Arctic Oscillation


The Arctic Oscillation (AO) was headed toward near-record levels during the past weekend while an historic storm buried many locations in the mid-Atlantic and southern New England with record snowfalls. 

A minor stratospheric warming event preceded the east coast snow storm by three-to-four weeks. The polar vortex (PV) weakened after splitting in two during late NOV and early DEC...as reported on the NEWxSFC web log here...here...and here.

The PV/s stronger half shifted from the eHEMI to wHEMI...where opposing flow between the two circulations created weak ridging near the pole.  With time...the ridging aloft made its way to the surface where an arctic anti-cyclone pushed the AO to levels seldom seen.

The last time the AO was less than or equal to 21-DEC-09/s -5.668 value was over 20 years ago on 01-JAN-85.

The AO has been this low...or lower...on just 15 days since modern record-keeping began in 1950  (NOV-59, JUN-64, JUL-64, MAR-70, JAN-77, JAN-85).  The odds of an AO index value at or below -5.668 are 1 in 1460 (0.07%).

The record negative value of -7.433 was set 15-JAN-77 during a borderline +ENSO (MEI = 0.484)... strong +PDO (1.65)...and an east QBO (-14.91).



Monday, December 21, 2009

Winter '09 / '10 - Snow Storm #1 - Preliminary Verification


Preliminary storm-total snowfalls for Saturday and Sunday from CDUS41...PNS...and CF6 bulletins.

Nine new daily records.

SAT...19-DEC-09
ISP - 9.6" (5.6"; 1995)
PHL - 22.5" (9.2"; 1945)
ACY - 11.4" (4.5"; 1979)
DCA - 15" (7"; 1945)

SUN...20-DEC-09
BOS - 10" (7.7"; 1995)
PVD - 14.3" (6.3; 1995)
ISP - 14.3" (4.2"; 1995)
JFK - 4.7" (3.8"; 1995)
EWR - 4.3" (3.2"; 1964)

Please report errors in Comments.

Final results and storm summary Tuesday evening.



Surface analysis:  19-DEC-1945.
Daily snowfall records last set in 1945...fell at PHL and DCA during Storm Day 1.  The 1945 Miller 'A' nor'easter originated in Galveston Bay...TX...then tracked NE to a position near HSE.



Surface analysis:  20-DEC-1995
Daily snowfall records set in 1995 at ISP...JFK...BOS...and PVD during Storm Day 2.  The 1995 MIller 'B' nor'easter formed near HSE...passing just south of the 'benchmark' at 40/70.

Sunday, December 20, 2009

Winter '09 / '10 - Snow Storm #1 - ECMWF Verification

Outstanding NWP performance!
ECMWF/s 96-HR (4-day) Forecast

Miller-hybrid 996 mb surface LOW south of HSE.
Open 5H trof over OH river valley
Occluding 2ndary surface LOW over NE OH


Observed
Miller' A' 989 mb surface LOW due E HSE.
Closed 5H LOW over OH river valley
2ndary surface LOW over southern OH / NE KY


GooFuS didn/t have a clue...projecting the storm off-shore near the VA Capes...then out to sea.

Winter '09 / '10 - Snow Storm #1 - Blizzard Conditions


Blizzard conditions were observed at Islip...NY between 10:30 PM EST...19-DEC-09 through almost 2 AM EST...20-DEC-09.

Three conditions are needed for a blizzard:
1 - Visibility frequently less than 1/4 SM (< 1320')...
2 - Sustained wind speed or frequent gusts of 30 KTs (35 MPH) or higher...and
3 - Restricted visibility and high-wind conditions observed for at least three hours.
Note the METAR reports of SNINCR 3/xx (three inches-per-hour followed by the amount on the ground...rounded to closest whole number) in the 0456z...0556z...and 0656z observations (Sunday midnight - 3 AM EST)

SPECI KISP 200328Z 36017G27KT 1/8SM +SN BLSN FZFG VV001 M04/M07 A2955 RMK AO2 PK WND 01030/0300 P0001 $

METAR KISP 200356Z 36020G30KT 1/8SM +SN BLSN FZFG VV001 M04/M07 A2952 RMK AO2 PK WND 36031/0330 SLP998 SNINCR 1/7 P0001 T10441067 $

METAR KISP 200456Z 36019G30KT 1/16SM +SN BLSN FZFG VV001 M04/M07 A2949 RMK AO2 PK WND 01034/0441 SLP987 SNINCR 3/10 P0003 T10441067 410111050 $

METAR KISP 200556Z 36020G35KT 1/8SM +SN BLSN FZFG VV001 M04/M07 A2948 RMK AO2 PK WND 35035/0554 SLP985 931114 4/013 SNINCR 3/13 P0002 60013 T10441067 11039 21050 56023 $

METAR KISP 200656Z 35021G32KT 1/8SM +SN BLSN FZFG VV001 M04/M07 A2950 RMK AO2 PK WND 36032/0649 SLP989 SNINCR 3/15 P0003 T10441067 $

METAR KISP 200756Z 35016G22KT 1/4SM +SN FZFG BLSN VV002 M04/M07 A2953 RMK AO2 PK WND 36030/0717 SLP999 SNINCR 2/17 P0003 T10441067 $
 
Blizzards are wind storms where visibility is restricted by falling or blowing snow.
 
Blizzards defined here and here.

Winter '09 / '10 - Snow Storm #1 - Snow Cover

Snow cover as of 20-DEC-09 at 12z.  Heavy snow axis 18 - 21" runs through ROA - DCA - BWI - DOV - southern NJ - eastern LI.



Image courtesy NOHRSC

Saturday, December 19, 2009

Winter '09 / '10 - Snow Storm #1 - The Forecasts



26 forecasters...including 10 rookies...issued 550 stations forecasts for this season/s inaugural snow storm.  Forecasts are ranked by storm total precipitation (STP) in ascending order.  Blue (red) values are in the 25th (75th) percentile.

Forecasts here.



Consensus heavy snowfall axis from IAD - DCA - BWI - PHL - EWR - JFK - ISP -ACY - IAD.



Amplifying flow regime evidenced by rising PNA and cold air source indicated by strongly negative values for NAO and AO.  Oscillations come on the heels of a minor stratospheric warming event that slowed and split the PV.  Arctic region surface pressure rose as the one piece of the PV moved from eHEMI to wHEMI...where opposing flow between the two circulations created weak ridging near the pole.



986 surface LOW over HSE SAT 19-DEC-09 at 12z crawls to a position abeam the southern end of Delmarva peninsula 12 hours later laying down a wide swath of heavy snow from VA to MD on its way to SNE.

Verification period ends 11:59 PM EST...SUN 20-DEC-09.  Preliminary storm-total snowfall totals will be posted the evening of MON 21-DEC-09.

Thursday, December 17, 2009

Winter '09 / '10 - Early Snow...Cold Snap Hits NW Europe


"The iconic Eiffel Tower was closed to visitors, and train and bus services were delayed in many areas, as daytime temperatures dropped below zero for the first time in the year and black ice coated northern roads.

"Snow also fell in Britain, Germany, Belgium and the Netherlands and weather experts forecast worse to come, with disruption especially in Scotland and northern England.

"The northern Netherlands experienced the worst falls in four years, with snow up to 25 centimetres deep in places..."

More...


Winter '09 / '10 - Snow Storm #1 - Call for Forecasts


Today's PROBABILISTIC HEAVY SNOW AND ICING DISCUSSION from HPC:
"THE 12-18Z NAM NOW SHOW SNOW OF HISTORIC PROPORTIONS IN THE CORRIDOR FROM SOUTHWEST VA TO WASHINGTON DC AND BALTIMORE."
Snow may begin falling late FRI evening over southern forecast stations.  Any snow that accumulates before the verification period begins will not be included in the amount of storm-total snowfall for those stations.

Deadline: 10:30 PM EST FRI...18-DEC-09

Forecast element:  storm-total snowfall

Verification begins: 12:01 AM EST SAT...19-DEC-09
Verification ends: 11:59 PM EST SUN...20-DEC-09

Enter your forecast here.
Follow the link from 'Enter Storm Forecast.'

As always...there/s no cost...or fee...or annoying requests for personal information to enter.

If you are making your first forecast this year or you entered the 'season-total' forecast contest...you/ll need to create an account (user name / password / valid e-mail...if you want a copy of your forecast sent to your Inbox) before entering a forecast.

Image:  NYC 1917


Tuesday, December 15, 2009

Winter '09 / '10 - Coastal Teaser #2


The European starting showing signs of life in the eastern Gulf of Mexico and offshore from Georgia the last few runs.  Today...GooFuS came on board after several days of sending the storm out to sea.

Prospects for a Miller 'A' storm coming up the coast now looms in the medium range. 

Extension of the parent arctic air mass... centered south of Hudson Bay...will be over land as far south as CHS on Saturday... providing an excellent baroclinic zone along the SE coast for cyclogenesis as short-wave dives into base of trof on NW flow aloft.

Progressive westerlies currently progged suggest storm will skirt the M-A coast and not affect SNE.  More energy entering into the trof from the NW than forecast may cause a more northerly trajectory of the surface LOW.

Winter '09 / '10 - WSI - December Update

WSI Energycast Outlook Calls for Colder-Than-Normal Period in US, Especially in Northern States

Dr. Todd Crawford - WSI seasonal forecaster:
"The general pattern of cold-north and warm-south observed during the winter so far will generally continue into early spring. The warmest temperatures, relative to normal, will be in the south-central and southeastern states while the coldest temperatures will continue to be observed in the north-central states.  The cold Pacific Ocean suggests that the upcoming aggregate three-month period will be relatively cold nationwide, relative to normal, especially in March and April.  An incipient stratospheric warming event, which appears to be historic in magnitude, may continue to favor more Arctic air masses in mid-latitudes and increases confidence in the widespread cold forecast of the US these upcoming few months."
In February:
Northeast - Colder than normal

In March:
Northeast - Colder than normal

More...

Monday, December 14, 2009

Is -NAO Over-valued as an Antecedent Condition for Significant NE Snowfall?


Introducing another round of posts in a series of posts... challenging the conventional wisdom about the relationship between a negative North Atlantic Oscillation (NAO) and north-eastern US snowfall.

The analysis will use data from a paper by Paul Kocin and Dr. Louie Uccellini published in the Bulletin of the American Meteorological Society (BAMS - 2004).

The Northeast Snowfall Importance Scale (NESIS) categorizes and ranks snowstorms affecting the NE CONUS by snowfall and population distribution.

More about NESIS here and here.

Data for the analysis comes from Table 6 in the BAMS article.  Storm-day NAO values...as well as NAO values for the three days preceding the event will be examined to determine their correlation to notable snowfalls.  Contributions from the Arctic Oscillation (AO) and the Pacific-North  American (PNA) index will also be assessed.

Previous analysis on the subject have shown a weak...yet statistically significant relationship... between increased snowfall @ stations in the NE and mid-Atlantic regions during winters (D-J-F) where the average NAO is negative but no statistically significant relationship between -NAO and individual snowfalls of any amount.

Forecasters and wx enthusiasts often pine for the westerlies to align is such a way for the NAO to b/come negative in the mistaken belief this pre-condition is an essential requirement for decent snows over NE CONUS.

NAO values leading up to and those observed during the so-called 'Storm of the (20th) Century) stands this idea on its head.

Saturday, December 12, 2009

Wednesday, December 9, 2009

Winter '09 / '10 - Strike 2


Second contest-worthy storm in less than a week to be poorly forecast with enough lead-time to make a call for forecasts.

Early NWP forecasts indicating three-to-four hour snow burst followed by change-over to rain were way underdone.  HPC/s winter weather products also failed to indicate a decent event was coming for New England until it was too late.


ALB:  7.3" between 0836z and 1551z  (~7.5 hrs)  New daily record.
BGM:  6.4" between 0553z and 1216z  (~6 hours)  New daily record.

BDL:  6" between 0906z and 1251z  (4 hours)
ORH:  7.5" between 0933z and 1503z  (4.5 hours)

CON:  8.3" between 1117z and 1848z  (7.5 hours)
PWM:  4.3" between 1248z and 1839z  (6 hours)
BGR:  4" between 1534z to post time  (7 hours and counting; 8 - 12" forecast)
CARL  6 - 10" forecast

At least three other forecast stations measured nuisance snowfall amounts.

Image:  Winter in VT 1933

Tuesday, December 8, 2009

Winter '09 / '10 - C-C-C-O-L-D December - Part Deux

Amplification of the long-wave patten over NOAM appearing more likely with every model run.

The long-advertised deepening of the Arctic Oscillation (AO) index to near record-levels...aided by on-going weak stratospheric warming event weakening the polar vortex... appears set to get under way the next day or two along with an off-the-chart...near record-setting rise in the Pacific-North American (PNA) index.  Essential ingredients poised to fall in place for a stormy end of the month.

The return-period for an AO index less than or equal to -5.0 is 3.9 years for a total of 32 days since 1950 with no days observed that low since OCT-02.  The record value of -7.433 was set 15-JAN-77.





The return-period for a PNA index greater than or equal to 2.5 is 5.6 years for a total of 24 days since 1950 and no days observed during winter.  The record value of 3.397 was set 24-SEP-08.



Week 2 forecasts show broad positive (negative) 500 Z anomalies over the North Pole (across the western Atlantic) a perfect inverse match to the AO's loading pattern.  Note the negative anomalies south of Alaska marking the position of the Aleutian LOW and an east coast CONUS storm track custom-made for Miller 'A' cyclogenesis.



Week2 positive sea-level pressure anomalies located at high latitude along the Prime Meridian and south-central Canada allowing arctic air to drain into low-latitudes creating a strong baroclinic zone along the Gulf of Mexico and SE CONUS coasts.

Daily teleconnection index values here.

Earth System Research Lab - Physical Sciences Division/s Week 2 Tercile Probability Forecasts here.

EDIT:  corrected dates of record AO and PNA index values

Sunday, December 6, 2009

Winter '09 / '10 - Opening Day

Winter/s inaugural snow storm was poorly forecast by NWP up until just before the event began.

NCEP/s HPC had a 10% probability of +4" over SE VA near the time a decision would be made whether the storm was contest-worthy.  Storm-total snowfall amounts were marginal...even if the an accurate numerical forecast had been more timely.





Saturday, December 5, 2009

Regular Season Contest Overview


DC 1917
Here/s the sequence of events whenever there/s a contest-worthy snow storm.

A 'Call for Forecasts' is made ~24 hours before the deadline for entries. The deadline is usually 10:00 PM ET the day before snowfall is expected.

The 'verification period' begins @midnight on the day snow is expected to begin over any forecast station. The period continues until snow has stopped falling over the entire forecast area or a hard deadline for the Contest/s end has been set.

Once the snow has stopped and the NWS has reported the snowfall amounts in their daily climatological bulletins ... a 'preliminary verification' data table with the storm-total snowfall amounts is posted to the web log. These numbers are subject to challenge and / or modification for ~24 hours ... if they are found later to be in error and / or new information becomes available.

All forecasts are then verified and ranked. The forecaster with the smallest sum of squared-errors (SUMSQ) is the winner.  Storm summary statistics and individual station forecast results are posted to the Contest/s web page. Forecasters are notified by e-mail when the final results are posted.

Summary
Day 1 - Call for Forecasts
Day 2 - Deadline - 10:00 PM Eastern (edited:  15-NOV-18)
Day 3 - Snow happens.  Forecasts posted to web site.  Forecast summary posted to web log.
Day 4 - Preliminary snowfall totals posted to web log ... if storm duration is one day.
Day 5 - Verified forecasts and Final Results posted to web site. Brief storm wrap-up posted to web log.

Complete contest rules and more information here.

Friday, December 4, 2009

Thursday, December 3, 2009

Winter '09 / '10 - Pattern Change - Part Deux

ECMWF 100-mb forecast advertising a split and dramatic shift in the polar vortex (PV) over the next 10 days...with a minor stratospheric warming thrown in for good measure.

PV is forecast to move from its present position in the eastern HEMI north of Siberia...



...to extreme NW Canada in the western HEMI.


Winter '09 / '10 - Coastal Teaser CANCEL

FORECAST


OBSERVED

Winter '09 / '10 - NHEMI Snowcover - November

Could be better.  Much better.  Not only is much of Canada still below normal but so is a fair chunk of the Upper Plains and mid-West.



Orange (blue) areas are below normal (above).  White areas are normal.



Late OCT rally lasted through early NOV before fading rapidly.



State of the cryosphere at month/s end and the start of meteorological winter shows there/s a lot of ground to be made up.  Good hi-latitude snow pack is an important element for creating polar and arctic air masses.

Wednesday, December 2, 2009

9th Annual 'Season-total' Snowfall Forecast Contest - The Forecasts

Welcome rookie forecasters and welcome back veterans to the start of another NE.Wx snowfall forecasting season.  19 entries this year.  Only three expecting below-normal snowfall.

Forecasts in the table below are ranked in ascending order by sum-total snowfall for all stations.

Sum-total period-of-record normal (PORN) snowfall for all stations during D-J-F-M is 916.9".
The percentage displayed next each forecaster/s 'season total' snowfall forecast is its PORN percentage.

Forecasts colored in BLUE (RED) indicate the 25th (75th) percentile.



Minimum sum-total forecast is 576.6" (63% of PORN).
Maximum sum-total forecast is 1,298" (42% above PORN)

Forecaster average is 1,042" (14% above PORN)
Average of all 25-station forecasts is 16% above PORN.

Consensus station forecasts for snowfall at least 25% above PORN:
BGR - 139%
BDR - 135%
SBY - 133%
PHL - 132%
ACY - 128%

® = rookie
© = defending champion

Detailed forecasts here.

Tuesday, December 1, 2009

9th Annual 'Season-total' Snowfall Forecast Contest - UPDATE

Deadline for entries has passed.  Thanks to all who submitted a forecast.
 
Expected to have the forecasts posted this evening but that will have to wait until tomorrow.