Winter '18 / '19 - 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 '18 /'19 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 '18 / '19 ... '84 / '85 was the leading analog followed by '72 / '73 ... '59 / '60 ... '89 / '90 ... and '91 / '92.  Winter '18 / '19 NAO started weakly positive ... approached zero at meteorological winter's end ... then climbed above 1 in MAR.

A qualitative assessment of the analog forecast's accuracy would rate the analogs #2 or #5 as 'best' with #1 ... 3 ... and 4 'poor.'

A quantitative assessment of the other '72 / '73 and '91 / '92 teleconnections ... not so much.

Read more »

Winter '18 / '19 - Arctic Oscillation (AO) Analog Verification

The analog forecasting technique seeks associations to the AO state in the run-up to the coming winter with AO run-up states of past winters.  Presented here is the verification of AO analogs for the '18 /'19 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.

In the run-up to Winter '18 / '19 ... '55 / '56 was the leading analog followed by '64 / '65 ... '84 / '85 ... '99 / '00 ... and '91 / '92.  Observed AO values for Winter '18 / '19 began neutral in DEC ... turned weakly negative in JAN ... then surged positive at meteorological winter/s end and into MAR.

A qualitative assessment of the analog forecast/s accuracy rates '84 / '85 as best.

A quantitative assessment of the other '84 / '85 teleconnections ... not so much.

Read more »

Winter '19 / '20 - NCDC Issues Updated Period-of Record Snowfall Climatology

Here

It's only been four years since the last update.

More NCDC climo data here

Winter '18 / '19 - 20th Annual 'Regular Season' Snowfall Forecast Contest: FINAL Results

After five snow storm forecasting contests ... under the ‘two-thirds’ rule … forecasters who entered at least FOUR 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).

Read more »

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

UPDATE (08-APR-19)
Forecaster verification table at the Contest web site (direct link)
Season-total snowfall v period-or-record normal trend chart

---
Original post:  (07-APR-19 @ 8:30 PM EDT)

Congratulations to Donald Rosenfeld for his 1st Place finish.
Donald also placed first in '05 / '06 and '12 / '13.

Forecasters ranked by their total absolute error.

---

Season-total snowfall from all stations (784") came in 14% below the period-of-record normal (914").
Winter '18 / '19 ranks 11th among other Contest years since winter '04 / '05 ... the season when BGR and BGM were added to the station list.

---
Apologies for the abbreviated recap.
Will add to the pile a little at a time this week including each forecaster's station-by-station verification.

Winter '18 / '19 - Season-total Snowfall Forecast Contest: Preliminary STP Verification

UPDATE (08-APR-19 @ 11:55 AM EDT)
Added links to monthly snowfall totals

---
Original post (06-APR-19 @ 6:42 PM EDT)

Preliminary sum-total season snowfalls for the verification period 01-DEC-18 though 31-MAR-19 collected from monthly climate bulletins (CLMxxx; CXUS51 ... CXUS52).

Seven forecast stations (28% of all stations) with bonus season-total snowfalls [more than Period-Of-Record-Normal (P-O-R-N)].

Rank ordered descending by percent of P-O-R-N.

Green ==> 75th percentile
White ==> Interquartile range (< 75th & > 25th percentiles)
Red ==> 25th percentile

---
Please report any errors in Comments along with a link to the correct data.
FINAL results expected NLT SUN evening.

---
Winter '18 / '19 monthly snow totals
DEC:  http://newxsfc.blogspot.com/2019/01/winter-18-19-season-total-snowfall.html
JAN:  http://newxsfc.blogspot.com/2019/02/winter-18-19-season-total-snowfall.html
FEB:  http://newxsfc.blogspot.com/2019/03/winter-18-19-season-total-snowfall.html
MAR:  http://newxsfc.blogspot.com/2019/04/winter-18-19-season-total-snowfall.html

Winter '18 / '19 - Season-total Snowfall Forecast Contest: MAR Totals

MAR-19 snowfall summary by Contest forecast station.
Rank ordered descending by percent of monthly period-of-record-normal (P-O-R-N).

---
Green ==> 75th percentile
White ==> Interquartile range (< 75th & > 25th percentiles)
Red ==> 25th percentile

---
MAR Forecast Station Highlights
Eleven stations above P-O-R-N
MA: 5
NE:  6

BDR ... NYC ... ABE ... MDT
At least 200% of monthly P-O-R-N

Biggest losers
Observed v P-O-R-N (% P-O-R-N)
RDU:  0" v 1.0"
ORH:  0" v 0.9"
RIC:  0" v 2.2"
SBY:  T v 1.8" (3%)
DCA:  0.3" v 2.1" (14%)

Winter '18 / '19 - Regular Season: Status

Philly

Winter '18 / '19 getting long in the tooth for another contest-worthy snow storm; although ... it/s happened four times previously in April ('02 ... '03 ... '07 ... and '11) which works out to a five-year 'return period' so maybe we get lucky.

Winter '18 / '19 - Regular Season Interim Standings: #3

CORRECTED (09-MAR-19)
Don Sutherland had been identified incorrectly as the Chief 'Regular Season' forecaster when that honor belongs to Brad Yehl.  We regret the error and thank Don Sutherland for bringing it to our attention.

---
After five snow storm forecasting contests ... under the ‘two-thirds’ rule … forecasters who have entered at least FOUR forecasts are included in interim standings #3.

Complete interim statistics table and chart at the Contest/s web site HERE (direct link)
Forecaster statistics for Winter '18 / '19 contest snow storms HERE (direct link)

---
SUMSQ errors for each contest snow storm are normalized (i.e., standardized) with a 'Z-score' ... then averaged to compute the standings. If a forecaster has issued forecasts for more than two-thirds of all snow storm contests ... then Z-scores from their 'best two-thirds' forecasts are used to calculate the Interim and Final standings.

Similar idea as dropping the lowest test score before computing the final grade.

Winter '18 / '19 - Snow Storm #5: FINAL Results

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

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