gymengineer

They also have access to HCCA and FSSE, which often are the top performers and even sometimes outperform the official forecast track. There was the whole reveal this past fall of the 2020 contract between NOAA and RenaissanceRe that keeps the HCCA forecasts, with proprietary techniques from the insurance risk firm, from the public for 5 years.

I just don't remember- what were the models showing during the "gone" period for the two January snowstorms last year that came back in under 48-hours? Did they send the snow to our north or to our south in the medium term?

Yup, I have the Philly weather book by his former coworker which describes in detail the fiasco, including the TV crawl he put out way too early to hype up the storm.

I feel like we'd all avoid the peaks and valleys of modeling at range if we adopted the mindset of a NWS meteorologist doing their job. Like no watches issued day 3 and beyond. Snow maps only to 72 hours. Only the range of possibilities discussed in the medium range. Use of probabilities.

At least there was a trend in the medium term and the bust was not literally the day of, which we do regularly.

That’s why I was puzzled when someone brought up 2/89 as an analog as if it were a good thing.

I counted 7 that were 4”+ for parts of this sub forum- what I think you’re referring to as the DMV jackpots. 3 had 10”+ amounts (I didn’t count the 1/19 date because it was too far in date from 1/12-13/19). But only one, 2/06, was a megalopolis storm. That storm as most of us know was very significant, though, as it reached NESIS Category 3.

6.0”

1/1940- 21” at Richmond, 9.5” in DC, 12” in southeast DC, 24” into PG county.

That was the measurement issue I was referring to. It was the 1/2016 blizzard. Here's the service assessment report where they acknowledged the snowboard issue but refused to change the result (see page 20): https://www.weather.gov/media/publications/assessments/2016_blizzard_snowfall_evaluation.pdf

And any storm since Knickerbocker that could have cleared 20" at DCA had small flaws. 2/10 was above freezing there at the start, so lost some precipitation to melted snow. 1/96 had the dry slot/sleet lull. 1/16 had the dry slot lull (and measurement failure). Of course, 2/03 was mix-y. And just not enough precip in 2/79, 2/83 and 12/09. I think in terms of modeling, 1/16 had the best shot for 20", but I remember how bleak it was in here for downtown DC folks during the lull- the duration of it was unexpected.

Pivotal is free for the Euro: https://www.pivotalweather.com/model.php?m=ecmwf_full

The surface depiction on the GFS at hr 144 is a complete mess with the double low structure. All the other globals have one coherent low.

It’s not about any model run, IMO. It’s that any model that catches the phasing spits out a historic East coast storm up through the northeast. That isn’t a high probability outcome, but the potential is still being shown on every model suite.

I was thinking "Thank goodness the Super Bowl is literally going on right now" when wxtrix posted about her cataract appointment.