Texas Winter 2024-2025

[snownado]

It’s the worst performing global by a long shot, well documented in this forum over the years

Actually did well during 2021 before the rest of the globals even sniffed cold was coming (especially the "King"). It has its moments, but it was on to that event early on if I recall.

Yes, it did well with temps with that one but outside of that event, we’ve been burned so many times by that model even more so than the GFS

Broken clocks...

[txtwister78]

SREF mean accumulations

[wxman22]

Keep in mind these are experimental.

[Ralph's Weather]

18z RGEM shows exactly why the finale may end up being where the big transition occurs into DFW putting down several inches before precip shuts off. Keeps everything mostly sleet though throught most of the event and all snow just NW

https://images.weatherbell.com/model/rgem-all/tx/instant_ptype/1736186400/1736488800-3CSXjL6FJUg.png

The Canadian is an outlier with its temp profile though. It has cold surface and warm 850mb while all others are pretty steady temp wise throughout the column.

[Quicksilver17]

I'm pretty confident we'll see a cold moderate to heavy rain Thursday into Friday. Any ice and snow accumulations will likely remain north of the metro area. I suppose Tomball could briefly see some sleet/freezing rain at the onset but it should quickly become all rain.

STS,

Tks for the insight. Fingers crossed that we get something this far south before end of Winter.

We might get our turn down here later this month. It’s looking interesting.

Hope so, it would be nice to see some white gold falling here again!

[txtwister78]

18z GFS looks u-g-l-y. Boy better start rooting for these short-range models I tell ya with these global trends.

[South Texas Storms]

I disagree, I think the map you posted is depicting totals exclusively for snow. For Dallas proper, for example, the icon depicts over 1” qpf over the course of the storm, of which 0.1” fall before temps fall to freezing. If the remaining >0.9” qpf that does fall below freezing were being lumped in with snow, Dallas’ storm total would be depicted on that map as 9-10”, not 5-6”. I think nw shift and reduction in totals is because it’s depicting more precip as “not snow” this run.

For the 18z ICON, it's both warmer and drier.

But it's always been an unreliably wonky model, so...

It’s the worst performing global by a long shot, well documented in this forum over the years

The ICON is far from the worst global model, in my opinion. It handled the tropical cyclone threats to TX very well, if not the best, of the globals last summer. It also typically has a good handle on temperatures, with no well known cold or hot bias.

[TeamPlayersBlue]

18z RGEM shows exactly why the finale may end up being where the big transition occurs into DFW putting down several inches before precip shuts off. Keeps everything mostly sleet though throught most of the event and all snow just NW

https://images.weatherbell.com/model/rgem-all/tx/instant_ptype/1736186400/1736488800-3CSXjL6FJUg.png

Im not a fan of this one, because the major models have the 850 temps below freezing where this map shows frz precip.

[HockeyTx82]

18z GFS looks u-g-l-y. Boy better start rooting for these short-range models I tell ya with these global trends.

Very vague, what are you seeing and for where? We have most onboard with a narrow line slicing the DFW Metro, so what specifically are you seeing and not liking?

[TeamPlayersBlue]

18z GFS looks u-g-l-y. Boy better start rooting for these short-range models I tell ya with these global trends.

850 frz line is waaaayy north of other runs

[snownado]

For the 18z ICON, it's both warmer and drier.

But it's always been an unreliably wonky model, so...

It’s the worst performing global by a long shot, well documented in this forum over the years

The ICON is far from the worst global model, in my opinion. It handled the tropical cyclone threats to TX very well, if not the best, of the globals last summer. It also typically has a good handle on temperatures, with no well known cold or hot bias.

In a world where the UKMET, NAVGEM, JMA and CFS exists, true...

[txtwister78]

18z RGEM shows exactly why the finale may end up being where the big transition occurs into DFW putting down several inches before precip shuts off. Keeps everything mostly sleet though throught most of the event and all snow just NW

https://images.weatherbell.com/model/rgem-all/tx/instant_ptype/1736186400/1736488800-3CSXjL6FJUg.png

Im not a fan of this one, because the major models have the 850 temps below freezing where this map shows frz precip.

Well definitely not the GFS that's for sure if you're including that one as a major. Take a look at those 850 temp swings over the past several runs for the Thursday timeframe. Yikes.

[Stratton23]

In my opinion the HRRR/ NAM models probably have a better handle on what surface temperatures will be like, at this range the short range guidance gets the most weight

[Ntxw]

For the 18z ICON, it's both warmer and drier.

But it's always been an unreliably wonky model, so...

It’s the worst performing global by a long shot, well documented in this forum over the years

The ICON is far from the worst global model, in my opinion. It handled the tropical cyclone threats to TX very well, if not the best, of the globals last summer. It also typically has a good handle on temperatures, with no well known cold or hot bias.

I've felt it does well regarding situations where the surface plays a lead role, cold, lower level features etc. at 500mb it's had questionable fairly late shifts even with this current event, and some others the past couple of seasons. No model is perfect just have to know and account for each's strengths and weakness.

[bubba hotep]

In my opinion the HRRR/ NAM models probably have a better handle on what surface temperatures will be like, at this range the short range guidance gets the most weight

Just a smidge of difference...

18z GFS at 48hrs


vs.

18z HRRR

[txtwister78]

In my opinion the HRRR/ NAM models probably have a better handle on what surface temperatures will be like, at this range the short range guidance gets the most weight

They may very well, but tough to ignore the trends from the others considering we were locked on to those just 24 hours ago with accumulation totals. GFS looks much different right now.

[HockeyTx82]

In my opinion the HRRR/ NAM models probably have a better handle on what surface temperatures will be like, at this range the short range guidance gets the most weight

They may very well, but tough to ignore the trends from the others considering we were locked on to those just 24 hours ago with accumulation totals. GFS looks much different right now.

Globals leading up. Short range trust from here, no?

[snownado]

In my opinion the HRRR/ NAM models probably have a better handle on what surface temperatures will be like, at this range the short range guidance gets the most weight

Just a smidge of difference...

18z GFS at 48hrs
https://m1o.pivotalweather.com/maps/models/gfs/2025010618/048/sfct-imp.us_sc.png

vs.

18z HRRR
https://m2o.pivotalweather.com/maps/models/hrrr/2025010618/048/sfct-imp.us_sc.png

Pretty sure the long-range HRRR also has a cold bias, FWIW...

[South Texas Storms]

It’s the worst performing global by a long shot, well documented in this forum over the years

The ICON is far from the worst global model, in my opinion. It handled the tropical cyclone threats to TX very well, if not the best, of the globals last summer. It also typically has a good handle on temperatures, with no well known cold or hot bias.

I've felt it does well regarding situations where the surface plays a lead role, cold, lower level features etc. at 500mb it's had questionable fairly late shifts even with this current event, and some others the past couple of seasons. No model is perfect just have to know and account for each's strengths and weakness.

Yep that's why taking a blended model approach typically works best when making a forecast. The NBM model does a good job and that's why it's the NWS model of choice.

[cheezyWXguy]

18z ICON is another example. Look at the adjustment in snowfall accumulations between one run. This is why in my opinion it's best to go with blends because these are wild swings here still showing up run to run and when you get mixed precip types involved I think that can create a false accumulation output. 12z to current 18z. Also high resolution will do a better job of sorting out where bands of precip setup so these totals won't be as broad.

https://images.weatherbell.com/model/icon-all/tx/total_snow_10to1/1736164800/1736521200-Qwrlgf0j3XY.png

https://images.weatherbell.com/model/icon-all/tx/total_snow_10to1/1736186400/1736521200-24L5lGCDsPM.png

I disagree, I think the map you posted is depicting totals exclusively for snow. For Dallas proper, for example, the icon depicts over 1” qpf over the course of the storm, of which 0.1” fall before temps fall to freezing. If the remaining >0.9” qpf that does fall below freezing were being lumped in with snow, Dallas’ storm total would be depicted on that map as 9-10”, not 5-6”. I think nw shift and reduction in totals is because it’s depicting more precip as “not snow” this run.

I don't disagree with that and that was my main point about the various precip types. You can still get snow above freezing (surface temps) btw so it's not about needing temps to get below freezing, but I used the euro as an example the other day in that on weatherbell (not sure if it does this on the other sites) it may depict sleet as the preciptation type but it converts that to accumulated snowfall and that's why you can't tell on some of these model outputs what isn't snow unless you go look at the details (850mb/thickness values etc at the time of the precip). It sure isn't rain at that point up there regardless of how much QPF you have so that was my point in reference to the sleet also getting mixed into the accumulation output. Heck even if your QPF varied that much run to run and thus impacted totals that way, it would still be a wild model swing. I don't think we disagree unless I'm reading your take differently?

Maybe I just misinterpreted your point. I thought you were saying that the icon map did have sleet accumulations lumped in with snow on that map.