Jump to content
  • Member Statistics

    17,588
    Total Members
    7,904
    Most Online
    LopezElliana
    Newest Member
    LopezElliana
    Joined

Feb 24th threat part deux


Typhoon Tip

Recommended Posts

Meh...I think really the solution itself has entirely changed from what models were predicting days ago.  I said as much to another member offline - just figured was better to let sleeping dogs lay until we got closer.

 

Let's see what the models bring about tonight, I think we've got to detach from what was being shown 2-3, even 18 hours ago because it doesn't really matter.  Analogs are surely useful, but only to the point that the underlying predicted pattern is close.

 

Everything aside of the blizzard developed later than we would have liked.  I would think the bumps we've seen continue, where that leaves us? 

 

OT, but actually shocked how little snow there is OTG in New England.  Around Plymouth NH there's some decent cover, less on the road up to Loon.  Loon itself has reasonable cover, but compared to normal I have to think it's really low.   Great skiing though.

 

Rollo agree. Posted similar thoughts earlier today: writing was on the wall as soon as Euro looked like crap for several consecutive runs beginning at least 12z Wednesday. Issue is, as this appeared synoptically a day ago, this was extremely close to a much bigger hit had that southern vorticity moved quickly enough. And NAM / CMC / SREFs / UK continued to show something along those lines.

 

You'll remember we thought the 18z NAM depiction last night was off (I even wondered what the Model Diagnostic Disco would comment) because it was so dramatically different at H5 compared to 6z, 12z, and even the subsequent 0z runs. Turns out it handled that vorticity over New Mexico / Texas the best.

 

Not sure what lesson to take from this. Hard not to conclude what we always seem to conclude: that the Euro trumps all, and if it disagrees significantly with NAM / CMC / SREFs / UK / GFS, then those other models must be missing something. I know Euro itself also trended southeast, but it had the best handling of this the soonest.

 

 

And with that last thought... the 12z Euro 2/22/13 for next week... lock it, frame it, put some armed guards and motion-detector lasers around it, and let's hope the other models follow.

Link to comment
Share on other sites

  • Replies 2k
  • Created
  • Last Reply

Well the models are only as good as the data that gets input into them.  I think the model physics is pretty good.  And now the NAM looks like the GFS... possibly with lighter inverted trof QPF.

 

That's not true.  While it is true is that "garbage in, garbage out", that doesn't mean good data equal a good model.  You can have all the 'right' data going in, but if the model was designed poorly, it's going to give sucky results.  

Link to comment
Share on other sites

Well the models are only as good as the data that gets input into them.  I think the model physics is pretty good.  And now the NAM looks like the GFS... possibly with lighter inverted trof QPF.

 

The NAM has been god awful this entire winter. The hard data 18 hours before the Feb 8 system gave my location over 30 inches of snow and I received 12. It's track of the storm at that 18 hour lead time was off by hundreds of miles. Here, it busted horribly on a forecast inside 42 hours once again.

Link to comment
Share on other sites

EURO and SREFs have been consistent with 5-7" here for 3 runs. NAM and GFS have jumped around from 1" qpf  to .25" qpf back to .5" to .25". All over the place. I'm relatively confident in 4-8" from up here down to ORH.

 

I agree with you that BOS is the real wild card, could be non accumulating crap.

In a small little region, the SREF and Euro have shown relatively consistent QPF.  But the ensemble has jumped around just like the NAM and the Euro solution changed every run too, but the QPF effects are more apparent outside of your local region.  The reality is, models change each run.  People expect too much exactness over a very small region.  If a model shifts 50 miles in 6 hours people call it garbage... but this is a GLOBAL domain we are talking about.  Guidance has consistently shown the same basic synoptic evolution for several days, which is remarkable.  Tiny shifts obviously cause large differences in actual weather, but unfortunately the scale of those shifts is smaller than model skill beyond a day or two.

Link to comment
Share on other sites

The NAM gets a bad wrap, but it really depends on the situation/pattern/scenario. It has moments to shine, like the Blizzard or most cold air damming situations for example. I'm not sure if it's this season or just people micro-analyzing model runs, but it *seems* like the models have had a very rocky season. In baseball terms, perhaps only the Euro is above the Mendoza line.

Link to comment
Share on other sites

What an awful 24 hrs of modeling.

 

I'd say about 48 hours.

 

The NAM gets a bad wrap, but it really depends on the situation/pattern/scenario. It has moments to shine, like the Blizzard or most cold air damming situations for example. I'm not sure if it's this season or just people micro-analyzing model runs, but it *seems* like the models have had a very rocky season. In baseball terms, perhaps only the Euro is above the Mendoza line.

 

Quincy I defend the NCEP stuff as much as anyone.  The GFS has had some good moments this winter, but the NAM is just atrociously bad.   It's become less reliable than the NGM which they flipped the switch on years ago.

Link to comment
Share on other sites

That's not true.  While it is true is that "garbage in, garbage out", that doesn't mean good data equal a good model.  You can have all the 'right' data going in, but if the model was designed poorly, it's going to give sucky results.  

I agree.  I'm not defending the NAM.  To my eye, which appears to be the consensus, the NAM is the least reliable of the major guidance.  But I think the biggest problem is the QPF parameter.  It is usually close to other guidance in every other respect, often leads trends, and usually has one of the better thermal profiles.  Unfortunately, to most of us the most important parameter is QPF.

 

I still think the limitations on all the models is the input, not the physics or computing power.  Ensemble modeling suggests that many systems are relatively stable to small perturbations, yet run to run changes are often large.  That's clearly a problem of inperfect data input.  If we had perfect input, we would have wonderful models.  Obviously that is very difficult and unlikely to significantly improve.

Link to comment
Share on other sites

I'd say about 48 hours.

Quincy I defend the NCEP stuff as much as anyone. The GFS has had some good moments this winter, but the NAM is just atrociously bad. It's become less reliable than the NGM which they flipped the switch on years ago.

I think we can agree that it has become very unreliable. Run to run continuity is virtually non-existent and although it has had its moments, when it's bad, it's clueless. I recall one event a couple if years ago it was spitting out 30" for NYC like 30 hours out and maybe 1-3 inches fell.

I'm probably just biased against the GFS now because it has had a few IMBY epic failures.

Link to comment
Share on other sites

I honestly can see why the general public gets pissed. No matter how many times an on air met gives caveats, this type of situation can truly suck.

What can you do though...I wish we could bust the other way more often

If anything, this is a great reason why it's dangerous to show what models are saying 4 days out.
Link to comment
Share on other sites

The NAM has been god awful this entire winter. The hard data 18 hours before the Feb 8 system gave my location over 30 inches of snow and I received 12. It's track of the storm at that 18 hour lead time was off by hundreds of miles. Here, it busted horribly on a forecast inside 42 hours once again.

I'm not a fan of the NAM.  But I agree with Quincy.  Every model has bad runs.  The Euro and GFS were taking this storm from Atl City to ACK (or closer) several runs ago.  And for the blizzard they were hundreds of miles south while the NAM was much closer.  People remember the huge QPF bombs of the NAM, which never verify, but they more easily forget the large track or height errors of other models that do not include insanely high QPF.

Link to comment
Share on other sites

If anything, this is a great reason why it's dangerous to show what models are saying 4 days out.

 

guess this is why you "stay the course" and don't chuck numbers early. 

 

the blizzard had remarkable run-to-run consistency from the euro, and really everything else was on board despite wobbles here and there. that was an easier call / higher confidence situation way in advance. this had some nice support at about 4 days...and then actually got worse closer in which is always a bad sign, imo. 

Link to comment
Share on other sites

I'd say about 48 hours.

 

 

Quincy I defend the NCEP stuff as much as anyone.  The GFS has had some good moments this winter, but the NAM is just atrociously bad.   It's become less reliable than the NGM which they flipped the switch on years ago.

That's silly.  The NGM was primative compared to today's models.  It's placement of features and QPF distribution was often off by hundreds of miles in the short term.  I remember it fondly however, and I bet others do too, because we had low expectations of its abilities, and if it was even halfway wet it meant a crushing.

Link to comment
Share on other sites

I miss the NGM....  It was a nice stable model. If you knew it's biases and deficiences you could compensate for it.  Also it was good to compare against and fact check the other models.

I'd say about 48 hours.

 

 

Quincy I defend the NCEP stuff as much as anyone.  The GFS has had some good moments this winter, but the NAM is just atrociously bad.   It's become less reliable than the NGM which they flipped the switch on years ago.

Link to comment
Share on other sites

I'm not a fan of the NAM.  But I agree with Quincy.  Every model has bad runs.  The Euro and GFS were taking this storm from Atl City to ACK (or closer) several runs ago.  And for the blizzard they were hundreds of miles south while the NAM was much closer.  People remember the huge QPF bombs of the NAM, which never verify, but they more easily forget the large track or height errors of other models that do not include insanely high QPF.

the NAM absolutely SUCKED with the blizzard. sucked. i know because 24 hours out the model, on the friday 12z run, took the low over my head. you can ask scooter because he and i texted back and forth about it. it verified SOUTH of the benchmark 24 hours later. 

Link to comment
Share on other sites

the NAM is terrible. sorry. 

The NAM is an absolutely useless model this winter. It got people in NJ especially up in a bunch before the 2/8 event when it had a couple of runs that nailed them, just to back off completely later. It just can't seem to get the progressive overall pattern right and it keeps blowing up and tucking in these storms prematurely. I wouldn't use its output for anything more than toilet paper the rest of the season.

Link to comment
Share on other sites

It has it's good points with mesoscale features such as lake effect snow bands. The other plus is it's handling of cold air damming situations.

 

 

 

 

I agree.  I'm not defending the NAM.  To my eye, which appears to be the consensus, the NAM is the least reliable of the major guidance.  But I think the biggest problem is the QPF parameter.  It is usually close to other guidance in every other respect, often leads trends, and usually has one of the better thermal profiles.  Unfortunately, to most of us the most important parameter is QPF.

 

I still think the limitations on all the models is the input, not the physics or computing power.  Ensemble modeling suggests that many systems are relatively stable to small perturbations, yet run to run changes are often large.  That's clearly a problem of inperfect data input.  If we had perfect input, we would have wonderful models.  Obviously that is very difficult and unlikely to significantly improve.

Link to comment
Share on other sites

the NAM absolutely SUCKED with the blizzard. sucked. i know because 24 hours out the model, on the friday 12z run, took the low over my head. you can ask scooter because he and i texted back and forth about it. it verified SOUTH of the benchmark 24 hours later. 

Down here for the blizzard it offered us the worst c*ck-teases for quite a while and had the main CCB way west of where it verified. For people like us who were on the edge through the whole thing it was a bad experience for many, lol.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...