Winter '23 / '24 - 25th Annual 'Regular Season' Snowfall Forecast Contest: FINAL Results

BWI - Christmastime in Highlandtown (1950s)

After THREE contest-worthy snow storms ... under the ‘two-thirds’ rule … forecasters who entered at least TWO 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).

---
FINAL Standings

---
Best Forecasts by Storm (lowest SUMSQ Error Z-Scores)

---
Top 10 Forecasts - All Storms [Error statistic:  Sum of Squared Errors Z (SUMSQ Z)]

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

---
Top 10 Forecasts - All Storms [Error statistic:  Total Absolute Error Z (TAE)]

 

TAE Error Z:  the secondary measure of forecaster skill (lower the better).
Accounts for the magnitude of snowfall forecast errors at each station.

---
Top 10 Forecasts - All Storms [Error statistic:  R-squared Z (RSQ Z)]

RSQ: a measure of the how well the 'forecast' snowfall explained (captured) the variability of the 'observed' snowfall (higher the better).

---
Forecasters' Skill Scores [(Forecaster Z-score - NWS ER WFOs Z-score) / ER NWS WFOs Z-Score]

Skill scores measure a forecaster/s performance against a standard measure (NWS ER WFOs).  Positive (negative) values indicate better (worse) performance compared to the standard.  0% for NWS does not indicate 'no skill.'

---
22 unique forecasters submitted a total of 965 station forecasts.
10 forecasters entered all 3 contests.
7 forecasters entered 2 contests.
5 forecasters entered 1 contest.

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