NWMSGuy
Member
just like to see the cold have a little more staying power... it don't have to be 0 degrees for week. lol give me week of highs in the 30 s and let the southern stream continue... than we can talk than...The pattern has changed as far as temperatures go. Just a winter storm isn’t there of course it’s always subject to change
Sent from my iPhone using Tapatalk
You can probably understand why most on this board don’t trust anything out past 200hr. I’m speaking of the WAR. What if it never leaves? I’m sure there are some solutions on the ensembles where that ridge hangs out til the end of timeEps continues it’s cold look inside day ten now so it continues to move forward for all you negative people that think things are being delayed cause it’s not showing snow and below zero temps at your house
Sent from my iPhone using Tapatalk
You continuously speak of storms being pushed back - there has not been one legitimate winter storm threat this entire winter outside of the December NC storm and the ice storm we just finished with. So I'm not sure what you're referring to - if you're talking about these fictitious D10+ winter storms you're just wasting your time on erroneous model noise. I'm not sure why anyone even looks at an operational run beyond 7 days out for any purpose. Just look at what happens to model performance even beyond D4!The storms themselves are being pushed back, or maybe I should say storms that were showing up before are now rain. So, it is taking longer for storms to show up, at least on the op runs. Hopefully, the ensembles are correct and the ops will start to show them soon and we will have an actual storm to track before the end of the month. The pattern might be coming along, but we still want it to produce winter storms, or it is a waste. When I talk of things being pushed back, I speak of the actual storms showing up to track.
I have nothing to base this on. But I could see this winter going like last winter. We saw December storm followed by a clipper type on system in February which dumped 1-3 inches. I could see something like that happening. Of course I’m in upstate sc so I can’t speak for areas west and east and north
Sent from my iPhone using Tapatalk
The storms themselves are being pushed back, or maybe I should say storms that were showing up before are now rain. So, it is taking longer for storms to show up, at least on the op runs. Hopefully, the ensembles are correct and the ops will start to show them soon and we will have an actual storm to track before the end of the month. The pattern might be coming along, but we still want it to produce winter storms, or it is a waste. When I talk of things being pushed back, I speak of the actual storms showing up to track.
What happened last winter, does not have any correlation with this winter. Clippers are always possible every winter. There's always the possibility of clippers during cold northwest flows. Clipper systems can also phase with pieces of energy in the southern stream, which would mean a stronger storm system.I have nothing to base this on. But I could see this winter going like last winter. We saw December storm followed by a clipper type on system in February which dumped 1-3 inches. I could see something like that happening. Of course I’m in upstate sc so I can’t speak for areas west and east and north
Sent from my iPhone using Tapatalk
Good post. Brick and I had almost this exact same conversation last night. He thinks because an Op showed fantasy snow and then lost it, but the GEFS mean showed fantasy snow, the Op should start showing it again if you believe it's actually going to snow. What seems to be impossible to grasp is that just because a model or ensemble mean shows fantasy snow, doesn't mean it's actually a high or even medium probability event. If you take a very marginal pattern and run a model 50 times with different initial conditions, doesn't it seem logical that at least a few of the runs will show snow for your area? That will give you at least SOMETHING on the mean. That doesn't mean that snow is likely at all. And the op may never show it again because the pattern REALLY isn't all that conducive. None of that means you can't thread the needle, but that's beside the point.You continuously speak of storms being pushed back - there has not been one legitimate winter storm threat this entire winter outside of the December NC storm and the ice storm we just finished with. So I'm not sure what you're referring to - if you're talking about these fictitious D10+ winter storms you're just wasting your time on erroneous model noise. I'm not sure why anyone even looks at an operational run beyond 7 days out for any purpose. Just look at what happens to model performance even beyond D4!
I think one thing that may be tripping some people is when we talk about the pattern is supportive of a storm, and then one pops up for a run or two, that somehow that is a legitimate storm threat. Unless it's inside D7, that simply isn't true and I or anyone else who says otherwise has become lost in the hype. Right now we are trying to determine the big pieces of the pattern - the ridge out west, the big vortex over southeastern Canada, and the position, strength and timing of the STJ.
This is all the more true on the GEFS which has less than half the members of the EPS. 2 or 3 skewed members can really impact the mean if they are showing insane 20-30" totals and we've seen that countless times already this winter. Additionally, it's important to remember that the ensembles should also be used primarily for patterns -- especially at long lead times -- because they are lower-resolution than the operationals they are derived from. Using lower-resolution models to detect individual storm threats 10 days out - good luck with that too!I think people have a misconception of the Ensemble mean. It is a smoothed combination of all the averages of the individual ensemble member runs. It can be skewed by very strong members and also by timing. So it is very hard to determine storm threats until you get closer and the members start agreeing. It can also mask things because one extreme member skews it. All this can make it appear things are moving back because a number of members may have been a few days too fast.
And to add to the point about the pattern, often, the ensembles do a good job at longer leads depicting the general nature of how the pattern is likely to evolve. But they may shift important features such as, ridge placements, trough axes, location of blocking, etc. slightly enough that the great snowstorm pattern that they showed for your location two days ago doesn't look quite as great now, even though the general pattern they show turns out mostly correct. This is just the nature of the chaotic state of the atmosphere.This is all the more true on the GEFS which has less than half the members of the EPS. 2 or 3 skewed members can really impact the mean if they are showing insane 20-30" totals and we've seen that countless times already this winter. Additionally, it's important to remember that the ensembles should also be used primarily for patterns -- especially at long lead times -- because they are lower-resolution than the operationals they are derived from. Using lower-resolution models to detect individual storm threats 10 days out - good luck with that too!
In summary, beyond D7 the discussion should be PATTERN PATTERN PATTERN focused and what COULD happen. If I go off the rails, someone can feel free to call me out, I'm sure at some point I've been a hypocrite of this advice and guilty of getting too caught up in details beyond a week myself before and should be called out for it if I have!
And for this reason I think it's futile to post 384 hr GEFS snow mean maps (no offense to those who post them and like to look at them), I just think it sets some up for disappointmentI think people have a misconception of the Ensemble mean. It is a smoothed combination of all the averages of the individual ensemble member runs. It can be skewed by very strong members and also by timing. So it is very hard to determine storm threats until you get closer and the members start agreeing. It can also mask things because one extreme member skews it. All this can make it appear things are moving back because a number of members may have been a few days too fast.
And for this reason I think it's futile to post 384 hr GEFS snow mean maps (no offense to those who post them and like to look at them), I just think it sets some up for disappointment
This is what I have never completely understood about ensembles. How can you load different initial conditions into an ensemble model as initial atmospheric conditions, weather observations, measurements, etc. are already measured and defined at a current point in time? It's like taking a snapshot of the current weather conditions, varying or changing them to something different, and then loading them into the ensemble. If the ensembles start out with initial conditions that are varied or possibly incorrect, how the ensembles can be expected to create an accurate forecast further out in time. Are these variations that are loaded into the ensemble being done under the assumption that the current weather conditions may not have been accurately measured or sampled and therefore may actually be incorrect to begin with?