Monday, February 15, 2010

Winter '09 / '10 - Interim Standings

After five contest-worthy snow storms...the first interim standings have been posted here.

Under the 'Two-thirds' Rule...forecasters who/ve made at least four forecasts were included in the Interim Standings.

Ordinarily...the first interim standings would have been posted after three storms...but the storms just kept a'comin.'

Image:  Boston...1910.

Winter '09 / '10 - Snow Storm #6 - The Forecasts

Forecaster Summary
Forecasts in the table below are ranked by storm total precipitation (STP) in ascending order. Blue (red) STP values are below (above) the 25th (75th) percentile of all forecast STPs.

Average STP (93.8") per average number of stations (25) forecast: 3.8".



Nine entries...including two Rookies... one Intern...and six Senior forecasters...issued 229 station forecasts for the winter/s 6th contest snow storm.

Everyone's station forecasts have been posted on the Contest's web site.


Consensus for a modest SNE event.



Arctic Oscillation (AO) stuck well below zero. North Atlantic Oscillation (NAO) remains more than one standard deviation below its mean in response to hi-latitude omega block in the vicinity of Iceland. Pacific-North American (PNA) about one standard deviation indicating amplified westerlies across NOAM.


Disorganized area of surface LOW over OH Valley and the mid-Atlantic.  Storm b/comes better organized with height where closed circulations begin near 700 mb and continue throughout the remainder of troposphere.

Verification period ends 11:59 PM EST...WED 17-FEB-10.
Preliminary storm-total snowfall amounts will be posted the evening of THU 18-FEB-10.

Sunday, February 14, 2010

Winter '09 / '10 - Snow Storm #6 - Raw Forecasts

Raw forecasts here.

All forecasts will be posted to the Contest web site and a summary on the web blog Monday evening.

Saturday, February 13, 2010

Winter '09 / '10 - Snow Storm #6 - Call for Forecasts

Northern stream short-wave forecast to intensify as it swings onto the east coast from the Upper Plains. Surface LOW expected to follow a familiar path through BWI and along coastal SNE where NWP models suggest it may linger in the Gulf of Maine...blocked by strong HIGH pressure over Greenland.

-----
Deadline: 10:30 PM EST SUN...14-FEB-10

Forecast element: storm-total snowfall
Verification begins: 12:01 AM EST MON...15-FEB-10
Verification ends: 11:59 PM EST WED...17-FEB-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).

Friday, February 12, 2010

Winter '09 / '10 - Snow Storm #5 - Final Results

NEWxSFC Chief forecaster donsutherland1 has won this season/s 5th snowfall forecast contest.

SUMSQ error:232.5"
Z-score: -0.996

Improvement over the average error:  57% (542")

Perfect forecast @JFK

Lowest station errors @JFK and ORF.

Congratulations donsutherland1!

2nd Place: wxfox51* (260.5"; -0.905)
3rd Place: weatherT (302.4"; -0.771)
Honorable Mention: Raven (312.7"; -0.737)
*Rookie

Full forecast verification...storm summary...and final results at the Contest web site.

Thursday, February 11, 2010

Winter '09 / '10 - BWI/s Historic Snowfall Season

History in the making....with several more weeks of winter left on the clock.

Baltimore / Washington International airport (BWI) has measured 79.9" of snow as of 10-FEB...the most ever observed since record-keeping began in 1883.

How unusual is it for BWI to see this much snow in one winter?

This amount of snow would be considered a statistical outlier...more than four standard deviations (4.322) above the period-of-record mean of 21.9".

It's more snow than 99.994% of all other winters.

Other notable seasons at BWI were 2002/3 (58.1")...1995/6 (62.5")...1963/4 (51.8")... and 1898 (51.1")....all of which were more that two standard deviations above the mean (more snow than 95% of all other winters).
Interesting to note...three of the largest season-total snowfalls have occurred in the last 15 years...during a time when NHEMI areal snowfall has been generally well below average...especially during summer.
How rare is this winter/s snowfall?
How often can an extreme season such as this be expected?

Since it/s never happened during the period-of-record...a statistical technique using a Gumbel distribution can be applied to determine the 'return-period' for BWI receiving this amount of snow in one winter. The distribution and return-period are calculated using the 126-year historical season-total snowfall data from LWX.


The 126-year Gumbel distribution for BWI season-total snowfall is shown above.

The calculation...based on the line's slope...intercept...and a crisp double-inverse natural log function...shows the return-period as once every 675 years! The return-period for the runner-up 'Winter of 1995/96' is only 115 years.

Image: Baltimore...MD FEB-1899 (credit:  Baltimore Sun)

Winter '09 / '10 - Snow Storm #5 - Preliminary Verification

Preliminary storm-total snowfalls for Tuesday and Wednesday from CDUS41...METARs...and PNS bulletins.

No climate report for SBY at post-time. AKQPNS carried a report of 8" from law enforcement. The snowfall total seems reasonable when plotted along with other reports from the immediate area. This value may be revised if CDUS41 or CF6 bulletins are issued before final verification.

No spotter report for HYA in BOXPNS. The 11-FEB-10 METAR at 12z carried a 7/group report of 0.78". That was reduced by 0.01" to account for a brief period of rain which fell during the verification period. Storm-total snowfall was estimated at 6.8" by applying an 8.8:1 ratio...the observed ratio at PVD.

Eleven new daily records.
TUE...09-FEB-10
PHL - 6" (5.5"; 2006)
IAD - 11.5" (2"; 1980)

WED...10-FEB-10
BDR - 4.8" (1.9"; 1967)
BGM - 4" (2.3"; 2003)
ISP - 9.2" (T; 2008)
JFK - 11.1" (2.5"; 1967)
ABE - 17.3" (4.2"; 1926)
PHL - 9.3" (8.7"; 1926)
EWR - 13.2" (2.6"; 1964)
BWI - 15.5" (9.8"; 1926)
IAD - 9.3" (0.5"; 1986)

New daily snowfall records at ABE...PHL...and BWI were established on Wednesday. The old records were set in 1926.  Storm evolution was similar for both events with decaying parent LOW over OH and secondary development / Miller 'B' forming off the Carolina coast.

Here are the surface maps from FEB 9th and 10th...1926...courtesy NOAA's Daily Wx Maps archive.


Note the parent LOW over OH and the initial evidence of secondary / Miller 'B' development off the Carolina coast on the 09-FEB chart.


The 10-FEB chart shows the coastal LOW at 40°N with a minimum central pressure below 29.20" Hg (988.8 mb). Nantucket's 8 AM pressure was reported as 29.16" Hg (987.5 mb) with a 12-hour change of 0.58" Hg (19.6 mb) and a wind speed of 52 MPH (45 KTS).


Please report errors and a link to the correct data in Comments.
Final results and storm summary Friday evening.

Tuesday, February 09, 2010

Winter '09 / '10 - Snow Storm #5 - The Forecasts

Forecaster Summary
Forecasts in the table below are ranked by storm total precipitation (STP) in ascending order. Blue (red) STP values are below (above) the 25th (75th) percentile of all forecast STPs.

Average STP (181.5") per average number of stations (24) forecast: 7.6".


21 entries...including five Rookies... one who has entered their first contest this winter...three Interns...two Journeyman...and 11 Senior forecasters...issued 509 station forecasts for the winter/s 5th contest snow storm.

Two entries were disqualified; one b/c the forecast included one station only and one was submitted well past the deadline.

Everyone's station forecasts have been posted on the Contest's web site.


Consensus for a northern mid-Atlantic and coastal SNE event.


Arctic Oscillation (AO) remains well below zero.  North Atlantic Oscillation (NAO) has dipped to more than one standard deviation below its mean in response to hi-latitude blocking in the vicinity of Iceland.  Pacific-North American (PNA) above one standard deviation indicating amplified westerlies across NOAM.



Parent LOW over OH put down light snowfall today.  Triple-point / secondary development over GA at 21z has moved to the coastal waters of South Carolina at post time.  Miller 'B' storm progged to come up the coast tonight and tomorrow...side-swiping coastal sections of the mid-Atlantic and SNE.

Verification period ends 11:59 PM EST...WED 10-FEB-10.
Preliminary storm-total snowfall amountst will be posted the evening of THU 11-FEB-10.

Monday, February 08, 2010

Winter '09 / '10 - Snow Storm #5 - Raw Forecasts

Raw forecasts here.
 
All forecasts will be posted to the Contest web site and a summary on the web blog TUE evening.

Winter '09 / '10 - Snow Storm #4 - Final Results

NEWxSFC Rookie forecaster Roger Smith has won this season/s 4th snowfall forecast contest with a SUMSQ error of 392.2"...a 55% improvement over the average SUMSQ error of 876". His Z-score was a respectable -1.218.


Roger also placed 1st in the 'average absolute error' category with an average error of 2.7" per station (24 stations forecast).
He had the lowest station errors @BOS...MDT...BWI...and ORF.


Congratulations Roger Smith!

2nd Place: catatonia  (SUMSQ error: 493.2"; Z-score: -0.964)
3rd Place:  wxfox51*  (500.5"; -0.946)
Honorable Mention: herb @ maws (521"; -0.894)

*Rookie

Full forecast verification...storm summary...and final results at the Contest web site.