• Hello, please take a minute to check out our awesome content, contributed by the wonderful members of our community. We hope you'll add your own thoughts and opinions by making a free account!

Learning Which Model Won?

Storm5

Storm5

Administrator
Moderator
Joined
Jan 4, 2017
Messages
14,116
Reaction score
26,643
Location
Birmingham Alabama
Let the great debate begin . Which model won in regards to this winter storm .

For me the NAM , RGEM, gefs and GFS did best with the precip shield placement. They all sucked in regards to thermals.

The CMC, EURO and EPS come in a tie for last In my opinion. The euro wasn't even close with the precip shield or temps for my area . The cmc is well, the cmc

Sent from my SM-G900V using Tapatalk
 
The NAM pretty much nailed it with timing and the GFS had the general idea but was too far south until almost go time.
 
I know who didn't win, the Euro. Far from being the "King" this time aorund. I just think the complexity of southern winter storms, always give the models fits. I do agree the NAM, RGEM were the best (better) of the other models.
 
The NAM sucked at first, but once we got into that 24-48 hr window, it handled it well for GA. Not sure about other states. The GFS started to get a little better on the day of the event. It was still aggressive with the snow but I think it did okay for some areas.

Sent from my SM-J700T1 using Tapatalk
 
DLI4SCwx said:
I know who didn't win, the Euro. Far from being the "King" this time aorund. I just think the complexity of southern winter storms, always give the models fits. I do agree the NAM, RGEM were the best (better) of the other models.
Click to expand...

On the tropical weather board I frequent, some of the mets were saying that Euro underwent some changes that weren't for the better. It wasn't the king (that it usually is) in the 2016 hurricane season either.
 
I know one thing, the American models at least failed miserably with QPF. The GFS had me with 1.0-1.5 inches of precip. I received around 0.35 inches.

My fears on the low level cold not being deep enough happened though. We stayed ZR until right at the end. If the GFS was correct, would have had a crippling ice storm.
 
NorthGAWinterWx said:
The NAM sucked at first, but once we got into that 24-48 hr window, it handled it well for GA. Not sure about other states. The GFS started to get a little better on the day of the event. It was still aggressive with the snow but I think it did okay for some areas.  

Sent from my SM-J700T1 using Tapatalk
Click to expand...

That's what I kind of saw as well. The GFS always over performs with snow. You know what did fail? The HRRR. Kept showing the convective banding being constant with no gaps and the snow ending later than it did. However, I think it did okay on P-types. Showed the boundary south of me, which is where it actually set up.
 
Models all nailed the timing. Pretty poor with the warm nose. 4k nam did the best

Sent from my SM-G928V using Tapatalk
 
NAM for me the gfs and euro both kept doing funky less qpf just east of the blue ridge in the northern foothills of NC. 8-12" is a lot more than some of the isolated 2-5" amounts I was seeing leading up to the storm
 
The GFS was the clear winner in my eyes for setting the stage from a long term perspective showing as much QPF for our area. None of the short range models I observed showed the degree of sleet lasting as long as it did for my area right up until the last minute. Bottom line - even in the age of technology, we don't know until it falls from the sky
 
I think the 4km nam really showed the more likely outcome. Once we in its range it showed changes that none of us really liked, I thought it was too aggressive, but it ended up being right. Or the most correct. I actually think the euro op and eps did the worse.
 
I think from an idea standpoint, the GFS held it's ground along with it's ensembles the best. While it was too cold for some, and South, the idea was there. The details got worked out by the higher resolution models like the NAM suite.

There were a string of days where we could not figure out if the GFS was being dumb or if it really was onto something it was so consistent and different than other guidance.
 
The 12k and 4K NAM nailed the northern stream storm for Tn on Friday . The GFS never had any snow within 2 counties of me on any run. Ended up with 2 inches. The GFS was consistent on the area of snow coming in but had the coverage area way underdone.
 
The NAM nailed the fact that this was going north and would leave the area from Anderson SC up through Raleigh NC out of the major snow. That line got to within 15 miles of me and never got closer all night. We did manage a dusting out of a final band just after daylight Sat morning.
 
Hey y'all, thanks for this thread too. I will be intrigued to see how the models do with major storm events (winter, spring or summer).
 
There needs to be an algorithm for snow total forecasting in certain major metros like Charlotte and Raleigh where if

A) deduct 1-2" off your forecast if temps are not supportive at the onset
B) deduct 1-2" off your forecast if a warm nose is within a 3 county radius
C) et cetera

I feel like this could really help mets like Brad Panovich...would take some research to see what works...but I believe it would further help with accuracy instead of going off the models entirely.
 
You must log in or register to reply here.
Back
Top