Thursday, January 12, 2017

Winter '16 / '17 - Snow Storm #2 - Another NWP and Public Forecast Failure

NYC
The season/s third contest-worthy snow storm out-performed not only the suite of numerical models but the professional meteorologists responsible for issuing the official forecasts.

As of Wednesday afternoon (04-JAN) ... the progs and public forecasts for the Friday/Saturday award-winning event (06-JAN thru 07-JAN) all indicated a low probability (40%) for a meager 4" snowfall across a narrow slice of SE VA and NE NC.

When all was said and done ... heavy snowfall was observed from SE VA (up to 12") ... the Delmarva (up to 13") ... LI (up to 7") ... and Cape Cod (where near-blizzard conditions were reported and up to 16").  Eleven contest stations measured at least 4".

Four new daily records:
PVD  10.7"
ISP  9.7"
RIC  7"
SBY   9"

SLRs where at least 4" of snow was observed ranged between ~12:1 and ~25:1

---
Some have suggested a 'Call for Forecasts' be issued whenever a potential snow storm looms.
That way ... a contest-worthy storm would seldom be missed.

Sounds like a good idea and it would be if there wasn't four hours of work involved when issuing the call.

Hundreds of email 'invitations' have to been sent (50 at a time ... 30' apart so as not to be flagged as spam or blocked by my ISP).

The contest/s web site has to be updated.
The contest/s web log has to be updated.
The contest/s Facebook page has to be updated.

---
Then it becomes a judgment call during the evening of the deadline ... deciding whether the storm will be contest-worthy (six to eight stations with more than nuisance storm-total snowfall i.e., >= 4").

Collecting surface ... upper air ... and remote sensing data used in the final analysis and storm summary reporting takes time while the event is unfolding. All for naught if the storm fizzles.

Believe you me ... no one is more disappointed than me when a good snow storm slips through the cracks.
Bottom line: a contest-worthy storm has to be reasonably well-predicted by the numerical weather prediction models and WPC 36-to-48 hours before the first flakes fall for a 'Call for Forecasts' to be issued.

1 comment:

Donald Rosenfeld said...

I agree with all of your thoughts and I think you should continue to do it your way, it's worked in the past and will work in the future.

No system is perfect and in our case, dealing with the atmosphere, some variability in indication must be accepted.

JMHO