CONTEST STATUS - Updated: SAT ... 01-DEC-18 @ 7:55 PM EST

Winter '18 / '19 - Snowfall Forecast Contests

20th Annual 'Regular Season' Snowfall Forecast Contest

What's A 'Contest-worthy' Storm?

Snow Storm #2
Synoptiscope in VCP32

Snow Storm #1
FINAL results here

18th Annual 'Season-total' Snowfall Forecast Contest
Deadline for entries has passed
Forecasts here

---------------------------------------------------------
Winter '17 / '18 - Snowfall Forecast Contests
---------------------------------------------------------
19th Annual 'Regular Season'
FINAL results here

---
17th Annual 'Season-total'
FINAL results here

Saturday, December 15, 2018

Winter '18 / '19 - Sudden Stratospheric Warming: Watch #1

Polar vortex (PV) displacement event appears to be in the cards come Christmas.



Full flow reversal + warming ==> major sudden stratosphere warming (SSW) event ...



BLUE:  into the board ==> east wind (-U)
ORANGE:  out of the board ==> west wind (U)

"A major SSW occurs when the 10 mb 60°N zonal mean zonal wind reverses from westerly to easterly and the 10 mb zonal mean temperature gradient increases poleward of 60°N." (.pdf)

---
Last winter's SSW event crash-landed over the UK (weeks of late season snow + cold)

This year ... round and round she goes.
Where she stops ...
Nobody knows.


Wednesday, December 5, 2018

Winter '18 / '19 - November/s AO Predicts The State Of Winter/s AO

NOV/s Arctic Oscillation (AO) Index:  -1.116

Conventional meteorological wisdom holds the state of winter/s AO is next to impossible to forecast.

Some use analogs.  The analog forecasting methods seeks similarities between the AO state in the run-up to the coming winter with AO run-up states of winters past.  The NEWxSFC method ranks analog years by their the sum of square errors (SSE) statistic

Lower SSE errors ==> stronger analog

Constraining the number of analog winters for analysis to five is arbitrary.
See this year's analogs below.

---

An alternative forecasting technique looks at the AO/s 'sign' (i.e., positive or negative) for any calendar-year's month preceding the pending winter as a potential leading indicator of the AO/s sign for upcoming D-J-F period.

Results from a chi-square 'test for independence' infers a statistically significant relationship ... at the 95% confidence level and a p-value < 0.05 ... between NOV/s AO sign and AO/s sign of the upcoming D-J-F period.

IOW ... if NOV/s AO is negative (positive) ... then the average AO state during the upcoming winter will also be negative (positive); although the classification model is stronger ... i.e., lower false alarm rate ... for the predictor's month with negative signs than positive.
True + ==> prediction is True
False + ==> prediction is False


Given NOV-18/s negative AO ... the 2x2 contingency table predicts a 73% likelihood (27/37) the three-month average AO (D-J-F) will also be negative this winter.  The results says nothing about the magnitude of the winter/s AO or which months will be negative ... only the sign of its average.

During a +ENSO ... this bodes well for a colder than average winter over southern portions and snowier than average winter over just about everywhere across the NEWxSFC forecast area.

Image courtesy NCEP Climate Prediction Center/s ATLAS No. 8 - Relationships Between El Niño-Southern Oscillation and the Arctic Oscillation: A Climate-Weather Link
 
---
The leading analog year for the AO in the run-up to the coming winter is 1955.    It tracks closely with the AO/s 2018 index year-to-date.  The analog saw winter's AO/s minimum extent in FEB then rebounding sharply into meteorological spring.

The average AO for Winter '55 / '56 was -1.226.  AO for all three months were negative.
Dec:  -0.444
Jan:  -1.204
Feb:  -2.029


Major flaw:  Winter '55 / '56 was a moderate La Nina (ONI:  -1.1).  The only year in the 'Top 5' AO analogs with a +ENSO year was '91 / '92 (ONI:  1.7) where the season-average AO floated high above the rest.

Another example how other teleconnection indexes ... in this case ENSO/s sign and strength ... can confound the potential predictive value of a single index.

Multi-variate indices of varying weights can sometimes offer additional insight when the ranked AO analogs are paired with teleconnection indices such as PDO ... QBO ... MEI ... SOI ... NAO ... and sun spots; however .. that's not happening this year ... either.

BOTTOM LINE:
- The signal from the 2x2 contingency model suggests strongly the coming winter/s AO will average below zero.
- AO analogs offer little insight for Winter '18 / '19.

Saturday, December 1, 2018

Winter '18 / '19 - 18th Annual 'Season-total' Snowfall Forecast Contest - The Forecasts!

20 forecasters + P-O-R-N + CONSENSUS

Two first-time forecasters
Welcome bdougherty and The Snowman

A lot of familiar faces ... too.  Many with 17 years experience.
Welcome back!

Forecaster table ranked by STP


BLUE - <= 25th percentile
RED - >= 75th percentile
ORANGE - Chief forecaster (1st Place - 17th Annual 'Season-total' Forecast Contest)
P-O-R-N - period of record normal
CONSENSUS - average station forecasts

 ---
Station forecasts:  550 (including P-O-R-N & CONSENSUS)
Stations forecasts for ...
BELOW average snowfall - 84 (15%)
AVERAGE snowfall - 25 (5%)
ABOVE average snowfall - 441 (80%)


Confidence (at least 67% of forecasts) for stations with ...
BELOW average snowfall @ NONE
ABOVE average snowfall @ all stations except ACY
All forecasters expect above average season-total snowfall @ BGR & BDR

---
All forecaster entries at the Contest/s web site here (direct link to forecasts).

---

NEWxSFC/s regular season 'snow storm' forecasting contest began 16-NOV.
'Call for Forecasts' for subsequent snow storms are issued at NEWxSFC/s web site and blog (you are here) ... via e-mail ... and Facebook.

Web sitehttp://www.newx-forecasts.com/index.htm
Facebookhttps://www.facebook.com/newxsfc/

Winter '18 / '19 - Outlooks

Thread updated periodically as new outlooks and forecasts are issued from a variety of credible sources.

---
Cohen (AER)

---
Little change from last month's outlook.
Climate Prediction Center (CPC)

Accuwx

Millersville University Weather Center's Eric Horst
Eric speak.  You listen.

Monday, November 26, 2018

Winter '18 / '19 - 18th Annual 'Season-total' Snowfall Forecast Contest - Call for Forecasts!

NEWxSFC Meteorologist-in-Charge
Seasonal Forecasts
Here comes winter with its oscillating Arctic and North Atlantic indexes ... suddenly warming stratosphere ... and an endless parade of coastal teasers.

NE.Wx/s 18th annual ‘Season-total' Snowfall Forecast Contest is the absolute best ... biggest ... and probably ONLY chance to be recognized for your long-range forecasting acumen ;/

And it's s-o-o easy.
Cool prizes ... too!

All you have to do is issue the best forecast for the season-total snowfall at 25 east coast stations between RDU and CAR!

Deadline: FRI ... 30-NOV-18 @ 11:59 PM EST

---
Visit the Contest's website to enter your forecast.
Follow the link at the top of the page to 'Enter Season-total Forecast.'

1st place prize (delivered post-paid right to your front door):
"The Snow Booklet" by Nolan J. Doesken and Arthur Judson (paperback) and
"Snow in America" by Bernard Mergan (hardcover)

---
As always ... NO cost ... NO fees ... NO advertising ... NO annoying requests for personal information to enter the contest.  NEWxSFC is just a fun exercise for winter wx enthusiasts to see who can make the best season-total snowfall forecast.

---
Forecast element: sum-total season snowfall @ each station
Forecast period: 01-DEC-18 through 31-MAR-19

Verification: NWS preliminary climate reports (CLM or CF6)
Error statistic: total absolute error [Σ abs(forecast - observed)]

Sunday, November 25, 2018

Winter '18 / '19 - ENSO ... Arctic Oscillation ... & Season-total Snowfall @NEWxSFC Stations (Wonkish)

Roosting snow crows rooting for bonus snows thanks to expected weak-to-moderate El Niño conditions.

Typical analysis ignores influence of other controlling factors ... such as the Arctic Oscillation (AO).

BLUF:  ENSO state does not provide robust guidance for season-total snowfall (STP) outcomes @NEWxSFC stations unless AO is considered.

--- 
Analysis period:  '66 / '67 to '17 / '18 (D-J-F-M)
n = 52

For all winter ENSO states ... if AVG D-J-F-M AO < 0 … then STP @NEWxSFC stations > AVG
Weak correlation:  R = -0.311

Correlation (R) >= +/- 0.6 considered generally to have useful predictive power

ONI = Oceanic Niño Index

Sunday, November 18, 2018

Winter '18 / '19 - Snow Storm #1: FINAL Results

Forecasters' station verifications and the storm summary at NEWxSFC/s home page.
http://www.newx-forecasts.com/

1st - donsutherland1  
SUMSQ: 38  
SUMSQ Z: -1.400  
STP: 7.1  (1)
TAE: 18.9  (1)
AAE: 0.95  (1)
     
2nd - Brad Yehl  
SUMSQ: 54  
SUMSQ Z: -1.215  
STP: 9.2  (2)
TAE: 21.2  (2)
AAE: 1.06  (2)
     
3rd - WeatherT  
SUMSQ: 85  
SUMSQ Z: -0.860  
STP: 16.2  (4)
TAE: 33.4  (5)
AAE: 1.52  (3)
     
HM - Donald Rosenfeld  
SUMSQ: 113  
SUMSQ Z: -0.541  
STP: 20.3  (8)
TAE: 33.0  (3)
AAE: 1.57  (4)

SUMSQ:  sum of square error (")
SUMSQ Z:  Z-score
STP:  storm total precipitation error (")
TAE:  total absolute error (")
AAE:  average absolute error (")
(#):  category rank


Forecast by Observed Snowfall Scatterplots for Top Forecasts
Dotted blue line below (above) solid red line ==> under (over) forecast


Station by Station Comparison of Top Forecasters

Saturday, November 17, 2018

Winter '18 / '19 - Snow Storm #1: Preliminary STP Verification

Table of preliminary storm-total snowfall by station for THU and FRI from CDUS41 (CLI) ... CXUS51 (CF6) and ... SAUS41 (METAR) bulletins.

Exceptions:
BDR - daily snowfall reports missing.
STP estimated from METAR 6/nnnn group @ SLR 10:1

SLR not reported for some stations with measurable snowfall b/c liquid and / or freezing precipitation also occurred during the verification period.

---
Stations observing at least Trace - 20 (74%)

Given a forecast station observed measurable snowfall ... stations observing at least:
4" - 10 (50%)
6" - 5 (30%)
8" - 3 (15%)
10" - 0

Max snow melt-water
CAR - 0.70"
BTV - 0.62"
CON - 0.60"
JFK - 0.91"

BGM SLR 6:1 (no liquid precipitation reported)

Max precipitation:  ACY - 1.73"

---
New daily records:
15-NOV-18 (1)
ORH - 7.8" (5.7"; 1967)

Eleven daily records set this day -- some held since 1906; however ... snow fell on ten stations before the deadline; therefore  ... they are not included.

16-NOV-18 (4)
CAR - 9.3" (5.4"; 1973)
BGM - 4" (1.8"; 1987, 1999)
ABE - 0.7" (0.05"; 1995)
MDT - 0.5" (0.05"; 1985)


SFC analysis:  00z ... 16-NOV-18

---
Please report errors in Comments along with a link to the correct data.
FINAL Results post expected SUN evening ... 18-NOV-18

Friday, November 16, 2018

Winter '18 / '19 - Snow Storm #1: The Forecasts

Forecaster class
Rookie 2
Journey 1
Senior 9
TOT 12

- Brad Yelh is NEWxSFC/s reigning 'Chief Forecaster' for 2019 ... having placed 1st in last year's 'Regular' season contest.

- Welcome and Good Luck to our Rookie forecasters NJWeather & Weather Wizard!

- No entry for ER NWS WSOs b/c of snowfall forecasts for many of their stations occurred before the verification period started.

All forecasts posted to the Contest/s web site. (direct link)


CORR:  quagmireweathercentral return to the fray for his third season warrants promotion to 'Journeyman' status. 

Forecast table ranked MIN-to-MAX by their verification period storm-total precipitation (STP).
BLUE = 25th percentile
RED >= 75th percentile
White / Grey- STP and 4cast cells range between the 25th and 75th percentile


Heaviest snowfall (>= 6") consensus expected along and to the right of a line from CAR - CON - ORH - ALB - BGM - BTV - CAR.  Lollipop @ BGM.



Flat PNA ... POS NAO ... Bi-polar AO.
Another NEWxSFC contest-worthy snow storm defies conventional teleconnection wisdom.

---
Challenging verifications ahead given some lower latitude forecast stations observed much of their storm-total snowfall prior to the deadline.

Look for the forecast station's  preliminary verifying STPs NLT tomorrow (SAT) evening.

Wednesday, November 14, 2018

Winter '18 / '19 - Snow Storm #1: Call for Forecasts!

Brooklyn ... NY
26-NOV-38
Cyclogenesis along the Mid-Atlantic progged to develop into an early season Nor'easter & poised to advance on a slowly retreating Arctic HIGH ... shows real promise for plowable snowfalls over northern portions of the forecast area.

Contest-worthy snow storms in NOV are quite rare with only two since NEWxSFC started in 1999 (2003 & 2013).

Earlier than usual deadline (8 PM EST) b/c of the expected pre-midnight on-set of accumulating snows.

Deadline: 8:00 PM EST … THU … 15-NOV-18
Verification begins: 8:00 PM EST … THU … 15-NOV-18
Verification ends: 11:59 PM EST … FRI … 16-NOV-18

Enter your forecast at the NEWxSFC/s home page @ http://www.newx-forecasts.com/
Follow the top center-of-page link from 'Enter Storm Forecast.

Sunday, November 11, 2018

Winter '18 / '19 - Eurasia Snow Cover: OCT

Eurasia/s areal snow cover for OCT-18:  ~10,391,000 km2

Last year:  ~12,052,000 km2

1% above 53-year P-O-R-N (~10,260,000 km2)
Rank: 22nd

Lowest in eight years

Seven of the last ten years above normal

Analog years for winter '18 / '19
Rank Winter ENSO
1 82-83 W+
2 00-01 C-
3 69-70 W-
4 01-02 nada-
5 84-85 C

---
Data courtesy Rutgers University Global Snow Lab
http://climate.rutgers.edu/snowcover/index.php

Saturday, November 10, 2018

Winter '18 / '19 - Gains and Losses: 10-NOV-18

Minor snow event over extreme northern portions of the forecast area today forced by a lifting 5H LOW centered @12z over SE Canada and 2ndry SFC LOW occluding INVOF the Gulf of ME.

Gains:  Blue
Losses:  Reds

Image courtesy NOHRSC

Winter '18 / '19 - Snow Advance Index and Season-total Snowfall in the NE and Mid-Atlantic

The Snow Advance Index (SAI) measures the mean daily rate-of-change in Eurasia's areal snow cover at latitudes equatorward of 60°N during OCT.  Published research suggests rapid increases in Eurasia/s OCT areal snow cover are associated with increased season-total snowfall (STP) in the eastern U.S.

Does the SAI provide useful guidance for season-total snowfall at NEWxSFC forecast stations?

Meh ... me thinks.

This analysis looks at the relationship between the SAI and NEWxSFC stations' STP for the winters between '70 / '71 and '17' / '18 (n = 48).

PCT increase in areal coverage is proxy for the mean rate of change of snow cover extent (SCE) from daily snow cover data in the above referenced study. (Cohen and Jones 2011).

AVG PCT increase in weekly Eurasian areal snow cover between Week 40 and Week 44 (OCT):  416% (ORANGE LINE)

OCTs ... where SAI > AVG:  20

Accordingly ... if OCT/s SAI is above AVG ... then STP at NEWxSFC forecast station should also be > AVG.

---
AVG 'season-total' snowfall (STP) for NEWxSFC forecast stations:  1,042" (GREEN LINE)
Winters ... where STP > AVG:  23

---
The question the analysis wants to answer:
Does the SAI-proxy - PCT increase in Eurasia/s OCT snow cover - predict whether NEWxSFC forecast stations' STP > AVG.

---
Years ... where SAI > AVG & Years ... where NEWxSFC stations' STP > AVG:  11 (RED DOTS)
Recall ... there were 20 years when OCT/s SAI > AVG which means nine years the STP < AVG.

Probability of  OCT SAI > AVG:  42% (20/48)
For any given OCT ... there's a 42% chance of SAI > AVG.

Probability NEWxSFC stations STP > AVG:  48% (23/48)
For any given year ... there's a 48% chance of STP > AVG.

What's the probability given SAI > AVG ... the STP will also be > AVG?
Cumulative probability for NEWxSFC stations STP > AVG ... given SAI > AVG:  42%

---
BOTTOM LINE:  In any given year ... if the SAI > AVG ... there's a 42% chance the 'season-total' snowfall ... for all NEWxSFC forecast stations ... will be above the period of study's AVG 'season-total' snowfall.

Looking at it another way ... there's a 58% chance the 'season-total' snowfall ... for all NEWxSFC forecast stations ... will _not_ be above the period of study's AVG 'season-total' snowfall.  This isn't to say SAI doesn't offer useful guidance for individual NEWxSFC forecast stations.

OCT-18/s PCT snow cover increase is just under 500% or ~80 PCT-points > AVG; therefore ... there/s a mere 42% chance this season the STP from NEWxSFC/s forecast stations will be > AVG.

Saturday, November 3, 2018

Winter '18 / '19 - Predictive Value of NW Atlantic Ocean's SST Anomalies for Winter's Phase of the North Atlantic Oscillation

Up until a few years ago ... the UKMET office issued a long-lead forecast for the phase of the upcoming winter's North Atlantic Oscillation (NAO).  The forecast was based on a statistically significant correlation between May's sea-surface temperature anomalies (SSTA) in the western Atlantic ocean (depicted in the image below left) and NAO's future D-J-F phase (as depicted in the Winter Z500 pattern image below right).


To apply this technique for Winter '18 / '19; take note of the SSTA pattern in the NW Atlantic at the end of MAY-18 shown below.

- Positive anomalies off the mid-Atlantic and NE coasts.
- Negative anomalies in the Labrador Sea off Greenland's southern coast.
- Positive anomalies north of Iceland in the Norwegian Sea.

https://www.ospo.noaa.gov/data/sst/anomaly/2018/anomnight.5.31.2018.gif

Compare these anomalies with the 'May SST pattern' map.
See how well they align?

MET Office research concluded if May's SSTA aligns generally with the prediction pattern in the western Atlantic ocean ... then the expected dominant NAO state would be positive during the upcoming winter.

Above normal SST in the offshore waters of the NE coast ... below normal water temperatures SW of Iceland ... and positive anomalies to the northeast point to a +NAO.  When the pattern is reversed a -NAO is forecast.

Based on May-18 SSTA ... the dominant NAO state prediction for Winter '18 / '19 is positive.
There's even empirical evidence the D-J-F NAO will average above zero if NOV's NAO index is greater than zero.

Winter-season forecasters hitching their collective wagons to the +ENSO state might could want to stop-and-consider the effect of the Northern Annual Modes on their outlooks.

---
Earlier post about UKMET NAO predictive research here.

Saturday, October 20, 2018

Winter '18 / '19 - How Much For Philly?


A weak-to-moderate El Niño continues to be the consensus forecast for Winter '18 / '19.

Weak-to-moderate El Niño conditions -- defined generally as sea surface temperature anomalies at least 0.5°C above normal near the International Dateline in the tropical Pacific ocean -- correlates with more than a few more inches of snow than average in and around Philly.

One problem; average isn't always the best metric to describe what's typical or most likely to occur.  That honor frequently falls to the median.

As such ... the median snowfall is four inches less than the 23" 51-year average suggesting this winter's season-total snowfall in Philly could very be up to seven inches more than most other winters.

Saturday, October 13, 2018

Winter '17 / '18 - Correlation of Eurasia's OCT Snow Cover and Season-total Snowfall in NE and M-A Regions

REPOST from OCT-17 (lightly edited for clarity)
Updated with Winter '17 / '18 verifications for RIC and NYC and their outlooks for Winter '18 / '19

---
The correlation between the areal coverage of Eurasia/s OCT snow cover and season-total snowfall has become broad-brushed conventional wisdom (CW) following the innovative research by AER climatologist Dr. Judah Cohen.

But ... just how well does the CW hold up for NEWxSFC/s forecast stations across New England (NE) and the mid-Atlantic (M-A) regions?

To find out ... monthly period-of-record areal snow cover data for Eurasia from Rutgers Global Snow Lab were correlated with season-total snowfall data for the 27 NEWxSFC/s stations.

A positive and statistically significant correlation means the greater the areal snow cover over Eurasia in OCT ... the greater the season-total snowfall for the following winter.

An Excel radar chart shown below depicts the results of the analysis.

DISCUSSION:  Stations between the inner and outer rings have a positive correlation coefficient statistically different than zero.  The coefficients range between 0.289 (CAR) and 0.424 (ORH).   Correlation values in this range are classified generally as 'low' (moderate:  >= 0.5 - 0.7; strong:  >= 0.7 - 0.9).  Even though the correlations are weak ... they can still provide useful information for seasonal snowfall forecasts.

Translation:  greater season-total snowfall over select stations in the NE and M-A is correlated with greater OCT areal snow cover in Eurasia.

---
The analysis showed other significant correlations of interest.

- RIC/s season-total snowfall has a positive correlation with Eurasia/s AUG areal snow cover.
Eurasia/s AUG-17 snow cover was well below normal ==> lower season-total snowfall @RIC this winter.

VERIFICATION Winter '17 / '18:  RIC STP 12.4"  (AVG:  13.2")
OUTLOOK Winter '18 / '19:  Eurasia's AUG snow cover below average ==> STP below AVG

- NYC/s season-total snowfall has a negative correlation with Eurasia's JUN areal snow cover.
Eurasia/s JUN-17 snow cover was above normal ==> lower season-total snowfall @ NYC this winter.

VERIFICATION Winter '17 / '18:  NYC STP 35.4"  (AVG:  26.1")
OUTLOOK Winter '18 / '19:  Eurasia's JUN snow cover below average ==> STP above AVG

GREEN (RED):  positive (negative) correlation between monthly Eurasian areal snow cover and season-total snowfall.

VERIFICATION Winter '17 / '18
Eurasia's OCT-17 areal snow cover was greater than average (12,051,667 km^2 v 10,261,134 km^2).
Positive correlations for the stations listed below suggest Winter '17 / '18 STP would be above average

STN:  STP" / AVG"
ABE:  46 / 31.4
BOS:  58.6 / 41.6
BTV:  81.5 / 69.8
PWM:  91.1 / 63.8

---
FINDINGS:  data analysis supports the CW for NE forecast stations but not so much across the M-A.

Friday, October 5, 2018

Winter '18 / '19 - The Woolly Bear Caterpillar Guide to Winter Weather Forecasting

BGR
03-JAN-63
It/s that time of year again when long-range wx forecasters scour the bowels looking for ... well ... wadda you say we don/t go there.

More better ... consult the Woolly Bear Caterpillar Guide to Winter Weather Forecasting.

(Lightly edited re-post from USENET news group ne.weather   11-OCT-03)

---
In recent weeks ... a spate of news articles have appeared about the woolly bear caterpillar and its legendary ability to foretell the upcoming winter based on the color and width of its black and orange stripes.

With so much good information scattered across so many sources ... there appeared to be a need to compile the forecasting rules.

This handy forecasting guide can be printed and keep inside your coat pocket for easy reference.
 
The Woolly Bear Caterpillars Among Us
There is more than one kind of woolly bear caterpillar.  The one you use when forecasting is the banded woolly caterpillar ... which becomes the Isabella moth (Pyrrharctia Isabella) in the spring.  Pictures of the caterpillar and moth can be seen here: Caterpillar Moth Both
 
Science is Skeptical
It should come as no surprise ... entomologists pooh-pooh the very idea banded woolly bear caterpillars can predict future events.  These naysayers claim variations in band color and width are evidence of a worm/s age and the fall weather conditions when the worm reached maturity.

Worms exhibiting more black than orange are older and grew in wet conditions.  Worms exhibiting less black than orange are younger and grew in dry conditions.

Thursday, October 4, 2018

Winter '17 / '18 - Arctic Oscillation (AO) Analog Verification

The analog forecasting technique seeks similarities to the AO state in the run-up to the coming winter with AO run-up states of winters past.  Presented here is the verification of AO analogs for the '17 /'18 winter.

The NEWxSFC method ranks analog years by their the sum of square errors (SSE) statistic.
Lower SSE errors ==> stronger analog

Constraining the number of analog winters for analysis to five is arbitrary.

Arctic Oscillation (AO) Index Analog Forecast Verification
In the run-up to Winter '17 / '18 ... '73 / '74 was the leading analog followed by '08 / '09 ... '54 / '55 ... '95 / '96 ... and '75 / '76.  Winter '17 / '18 AO started weakly negative and remained weakly negative until meteorological winter's end.  MAR-18 AO (not shown) crashed to -0.941.
 
A qualitative assessment of the forecast's accuracy would rate all analogs except '73 / '74 as 'poor' ... IOW ... useless.
 
OTOH ... analog #1 mimicked the observed behavior of the AO associated with a weak La Nina; whereas ... .  strong La Nina conditions prevailed during the winters of '73 / '74.
 
CONCLUSION:  The analog forecasting technique provided useful guidance for Winter '17 / '18.
 
---
An alternative forecasting technique looks at the AO/s 'sign' (i.e., positive or negative) for any calendar-year's month preceding the pending winter as a potential leading indicator of the AO's sign for upcoming D-J-F period.

Results from a chi-square 'test for independence' infers a statistically significant relationship ... at the 95% confidence level and a p-value < 0.05 ... between NOV's AO sign and AO's sign of the upcoming D-J-F period.

IOW ... if NOV's AO is negative (positive) ... then the average AO state during the upcoming winter will also be negative (positive); although the classification model is stronger ... i.e., lower false alarm rate ... for the predictor's month with negative signs than positive.  This is opposite of the relationship found for the NAO.


BOTTOM LINE:  If NOV's AO is negative ... chances are good the AO state will average negative during the D-J-F period.
 
NOV '17 AO/s sign was negative (-0.078).  The 2x2 contingency technique predicted correctly AO state for Winter '17 / '18 would average less than zero.  AO for the D-J-F period averaged (0-.076).

Saturday, September 29, 2018

Winter '17 / '18 - North Atlantic Oscillation (NAO) Analog Verification

The analog forecasting technique seeks similarities to the NAO state in the run-up to the coming winter with NAO run-up states of winters past.  Presented here is the verification of NAO analogs for the '17 /'18 winter.

The NEWxSFC method ranks analog years by their the sum of square errors (SSE) statistic.
Lower SSE errors ==> stronger analog

Constraining the number of analog winters for analysis to five is arbitrary.

North Atlantic Oscillation (NAO) Index Analog Forecast Verification
In the run-up to Winter '17 / '18 ... '90 / '91 was the leading analog followed by '94 / '95 ... '54 / '55 ... '04 / '05 ... and '02 / '03.  Winter '17 / '18 NAO started strongly positive ... strengthened through the heart of the season then crashed below zero following meteorological winter's end
 
A qualitative assessment of the forecast's accuracy would rate the analogs #2 ... #3 ... and #4 as 'poor' ... IOW ... useless.
 
OTOH ... analogs #1 and #5 mimicked the observed behavior of the NAO associated with a weak La Nina.  Moderate El Niño conditions prevailed during the winters of '94 / '95 and '02 / '03.
 
CONCLUSION:  The analog forecasting technique provided useful guidance for Winter '17 / '18.
  
---
An alternative forecasting technique looks at the NAO/s 'sign' (i.e., positive or negative) for any calendar-year's month preceding the pending winter as a potential leading indicator of the NAO's sign for upcoming D-J-F period.

Results from a chi-square 'test for independence' infers a statistically significant relationship ... at the 95% confidence level and a p-value < 0.05 ... between NOV's NAO sign and NAO's sign of the upcoming D-J-F period.

IOW ... if NOV's NAO is negative (positive) ... then the average NAO state during the upcoming winter will also be negative (positive); although the classification model is stronger ... i.e., lower false alarm rate ... for the predictor's month with positive signs than negative.  This is opposite of the relationship found for the Arctic Oscillation (AO).
 
BOTTOM LINE:  If NOV's NAO is positive ... chances are good the NAO state will average positive during the D-J-F period.
 
NOV '17 NAO/s sign was zero (0) for the 1st time in the index's 68 year record.  The 2x2 contingency technique offered no useful guidance.

Saturday, April 14, 2018

Winter '17 / '18 - 19th Annual 'Regular Season' Snowfall Forecast Contest: Season Summary and Historical Perspective

FINAL results here.

EIGHT contest-worthy storms this season ... one more than average.
The storm count tied for 6th place with four other Winters ('07 / '08 ... '08 / '09 ... '10 / '11 ... and '13 / '14).

Contest-worthy Storm Count by Month
DEC - 1 [25-DEC-17]
JAN - 2 [04-JAN-18; 17-JAN-18]
FEB - 2 [07-FEB-18; 17-FEB-18]
MAR - 3 [07-MAR-18; 12-MAR-18; 21-MAR-18]
TOT - 8



 
 
Verified forecasts
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm2_verifications_04Jan18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm3_verifications_17Jan18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm4_verifications_07Feb18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm5_verifications_17Feb18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm6_verifications_07Mar18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm7_verifications_12Mar18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm8_verifications_21Mar18.htm
 
FINAL Results and Storm Summaries
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm2_summary_04Jan18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm3_summary_17Jan18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm4_summary_07Feb18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm5_summary_17Feb18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm6_summary_07Mar18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm7_summary_12Mar18.htm
http://www.newx-forecasts.com/NEWxSFC_19/storms/storm8_summary_21Mar18.htm

Interim Standings
http://www.newx-forecasts.com/interim_19/interim1.htm
http://www.newx-forecasts.com/interim_19/interim2.htm
http://www.newx-forecasts.com/interim_19/interim3.htm
http://www.newx-forecasts.com/interim_19/interim4.htm
http://www.newx-forecasts.com/interim_19/interim5.htm
http://www.newx-forecasts.com/interim_19/interim6.htm
 
 

Wednesday, April 11, 2018

Winter '17 / '18 - 19th Annual 'Regular Season' Snowfall Forecast Contest: FINAL Results

Under the ‘two-thirds’ rule … forecasters who entered at least SIX forecasts are included in this season's FINAL standings.

Full table with all other error statistics at the Contest/s web site here (direct link).
Individual forecaster's storm statistics here (direct link).

 
 
Best Forecasts by Storm

 
Top 10 Forecasts (Error statistic:  SUMSQ Z)

SUMSQ Error Z is the primary measure of forecaster skill (lower the better).
Accounts for the magnitude and distribution of the storm-total snowfall for all stations.

Top 10 Forecasts (Error statistic:  Total Absolute Error Z - TAE)
TAE Error Z is the secondary measure of forecaster skill (lower the better).
Accounts for the magnitude of snowfall at each station.
 
Top 10 Forecasts (Error statistic:  RSQ Z)
R-squared (RSQ Z) is a supplementary measure of forecaster skill (higher the better).
Accounts for how well the variability of the observed snowfall was accounted for by the forecast.

---
Eighteen unique forecasters submitted a total of 2,582 stations forecasts.
Eight forecasters entered all 8 contests.
Three forecasters entered 7 contests.
The remainder entered fewer than six

---
Hope to see y'all again next winter.