And the winner of the challenge is...
I have gathered as much information as I can regarding snow totals. Most are exact, some are close estimates unfortunately. I cannot believe some of these 'official' reporting stations do not have a snow report. So, I did accept these 'few' estimates from the NWS pages. You will notice the storm totals next to the location. Despite the few estimates, I really do not think it would have made much of a difference. There was a clear winner for this challenge. In a mild upset, the winner is....
GFS
----------------------------------------------------------
That's right. Welcome to the MikJournal Snow Challenge between the superior Euro model and the doormat of forecast models, the GFS.
Who will win this challenge? The Euro, notorious for 'holding back energy from the Southwest'? Or the GFS, the model that seems to get lost in mid-thought?
Both claim medium-range forecasting as their niche. That's why these models should only be used at 4-7 days out from a storm system's ETA. I suggest neither of these models be used for short range forecasting. Leave that to the higher resolution and short-range forecast models like the NAM or HRRR.
Now, it's my rules. I get to choose the maps, the dates, the locations. After Sunday night, come Monday morning let's see how these 'snapshots' of this potential winter storm system unfolded by focusing on snowfall totals ONLY. I don't care about the rain, freezing rain, and sleet, just snowfall totals through Sunday night.
In this corner, here is a map of one of the Euro's runs on Tuesday, well into getting past that so-called Southwest Bias and still within an acceptable range for providing its 'superior' forecast.
And in this corner, The GFS' picture from the same day and applicable to the same time period from Friday through Sunday....
These are the maps I am using. No other maps.
Next, I choose the locations. As you can see, there are many locations represented by the amounts listed. For time purposes, I will select a FEW locations from Kentucky, Indiana, Illinois, and Missouri. These are the ONLY states I will use.
The locations listed below, I have no idea how they will do in terms of snowfall. But, I will gather the official snowfall readings from the NWS for Friday - Sunday night ending at Midnight(which would technically be Monday the next day). Remember, snowfall does not equal snow depth. You can actually have a dusting on the ground and still have 1" of snowfall recorded by the NWS.
I do not care how well one model did for other locations not listed. These are my rules. I choose the locations. I am grading this challenge solely on the amount of snowfall listed from the NWS through Sunday night, starting whenever on Friday, and only the locations I have chosen. I will not accept measurements from any other sources but the NWS offices for each location.
You will see each model's interpretation or estimate in inches. Then come sometime Monday or Tuesday, I will gather the official totals from the NWS offices and add up the differences. This challenge is not based on how many locations each model wins, but the grand sum of the differences. The model with the overall least difference wins this challenge. It's my rules.
Louisville: 2.9
Euro...6
GFS...2
Lexington: 2.1
Euro...7
GFS...6
Jackson: 0.4
Euro...5
GFS...4
Paducah: 0.6
Euro...7
GFS...3
Covington: 8.7
Euro...6
GFS...5
Terra Haute: 6
Euro...3
GFS...4
Evansville: 3.4
Euro...6
GFS...2
Indianapolis: 7.2
Euro...5
GFS...4
Muncie: 6
Euro...3
GFS...4
Bloomington IN: 3.6
Euro...4
GFS...4
Vincennes: 4
Euro...5
GFS...4
Effingham: 6.5
Euro...6
GFS...3
Carbondale: 4
Euro...8
GFS...3
Salem IL: 5.0
Euro...7
GFS...4
Cape Girardeau MO: 2.3
Euro...7
GFS...4
St Louis: 11.4
Euro...9
GFS...2
Springfield MO: 0.1
Euro...6
GFS...4
Rolla: 5.5"
Euro...9
GFS...7
I'll say right now, as I write this, St Louis is expecting nearly 12" snow, according to latest forecast. GFS might get slammed on that one alone.
I'll determine the winner with the data provided by the NWS offices for each of these locations on Monday or Tuesday.
MS
No comments:
Post a Comment