Jump to content
  • Member Statistics

    17,608
    Total Members
    7,904
    Most Online
    Vesuvius
    Newest Member
    Vesuvius
    Joined

18Z Model Thread


Recommended Posts

  • Replies 155
  • Created
  • Last Reply

Initial n. stream impulse misses phase, s. stream (being slower, is now almost phasing (or is late) with a second n. stream. Good stuff! The northern stream is really the bugger here....along with the speed of the s. stream.... :arrowhead:

southern stream is more key imo.. there are several vorts in the n stream that are possible for phase.

Link to comment
Share on other sites

Based off that one run, and i know it's only one (18z to boot), I think all the big cities still have a shot at something. You shift that 992 off Hatteras 50-100 miles west and....

Mid-Atlantic probably has a less than 5% chance except maybe delmarva and right along the coast. The phase as is in the 18Z likely the best it can possibly get. I think that might be the best solution one can get out of this given the weak nature of the southern PV anomaly.

Link to comment
Share on other sites

a valid stance...i feel like i've had my heart broken by the euro so i don't trust anything at this point....

but the optimist in me, looked at the 12z gfs ensembles and saw that a track up the east coast was at least in the envelope of possible outcomes...and that's important...it's why we do ensembles...you are trying to assess the probability distribution function...unfortunately we don't have the computing resources to do more than 20 (gfs?) to 50 (euro?) members...but in the not too far off future...we'll be running 100s and then 1000s of perturbed simulations and then you can really statistically assess the PDF....ahh exciting stuff

im going to vote blip for now.. gfs is due for one ;)

Link to comment
Share on other sites

when dealing with complicated wave dynamics...i'll put myself out there and say it's not 0% and its not 100%....does that help?

Mid-Atlantic probably has a less than 5% chance. The phase as is in the 18Z likely the best it can possibly get. I think that might be the best solution one can get out of this given the weak nature of the southern PV anomaly.

Link to comment
Share on other sites

Have to agree with one of the previous posters, this is well within the recently consolidating Euro/GFS Ensembles. As was noted in the 12z run both 12z op runs of GFS and Euro were slightly east of the ensemble mean. From what I saw of the GFS ensembles there were several members that showed hits similar to this 18z Op.

Link to comment
Share on other sites

Mid-Atlantic probably has a less than 5% chance except maybe delmarva and right along the coast. The phase as is in the 18Z likely the best it can possibly get. I think that might be the best solution one can get out of this given the weak nature of the southern PV anomaly.

Probably a very valid point. This makes things a lot more interesting to hang around for another cycle and see, but yes, the realist in me says this may be the best we can hope to see at this point.

Link to comment
Share on other sites

a valid stance...i feel like i've had my heart broken by the euro so i don't trust anything at this point....

but the optimist in me, looked at the 12z gfs ensembles and saw that a track up the east coast was at least in the envelope of possible outcomes...and that's important...it's why we do ensembles...you are trying to assess the probability distribution function...unfortunately we don't have the computing resources to do more than 20 (gfs?) to 50 (euro?) members...but in the not too far off future...we'll be running 100s and then 1000s of perturbed simulations and then you can really statistically assess the PDF....ahh exciting stuff

Maybe in the future they'll have more data points to work with as well...

Link to comment
Share on other sites

a valid stance...i feel like i've had my heart broken by the euro so i don't trust anything at this point....

but the optimist in me, looked at the 12z gfs ensembles and saw that a track up the east coast was at least in the envelope of possible outcomes...and that's important...it's why we do ensembles...you are trying to assess the probability distribution function...unfortunately we don't have the computing resources to do more than 20 (gfs?) to 50 (euro?) members...but in the not too far off future...we'll be running 100s and then 1000s of perturbed simulations and then you can really statistically assess the PDF....ahh exciting stuff

Yea, while obviously outliers, there were 2 or 3 ensemble members that were still MECS.

Link to comment
Share on other sites

when dealing with complicated wave dynamics...i'll put myself out there and say it's not 0% and its not 100%....does that help?

Haha, I will stick with my 5%, and most of that threat would be coastal and delmarva. Wave phasing is very complicated and models suck with it, yes, but one consideration here is the southern PV anomaly is so weak the northern stream is barely phasing anything now. In other words, this 18Z might be the best we get.

Link to comment
Share on other sites

good deal....i love snow but i love the process trying to figure it out too....

Haha, I will stick with my 5%, and most of that threat would be coastal and delmarva. Wave phasing is very complicated and models suck with it, yes, but one consideration here is the southern PV anomaly is so weak the northern stream is barely phasing anything now. In other words, this 18Z might be the best we get.

Link to comment
Share on other sites

Am I wrong to state that the OP is the extrapolation of data from all the simulations? If the average between phase and no phase is what we are seeing, than to me, the way the models have presented the data makes sense, when taken with this rudimentary approach.

Now the NAM leans toward a phase, as opposed to no phase. If the other models say phase, then the when comes into place? Right now the models have made a real mess of this!

Link to comment
Share on other sites

you finally get what looks to be an actual phasing of the southern stream impulse, instead of it getting completely sheared out....unfortunately it phases over N. FL, but you can see the result...a really intense storm just a bit too far out to sea....if you can get this phase to occur a bit further west...things will get much more interesting.....

like was just said wrt to models and phasing....and i assume most people here have not taken advanced fluid dynamics...but wave phasing is very complicated, and very small perturbations can lead to very drastic results in a final solution...

The low resolution of the GEFS may be helping to smooth out these complicated variables, hence a little more constant solution. Can't wait to see if the 18Z mean shifted any more west or holds serve.

Link to comment
Share on other sites

It's a good point....but it is also difficult to assess if that is having a net positive or negative effect on a given ensemble member...

low resolution can kinda work both ways...

The low resolution of the GEFS may be helping to smooth out these complicated variables, hence a little more constant solution. Can't wait to see if the 18Z mean shifted any more west or holds serve.

Link to comment
Share on other sites

I might be reading this wrong...but the OP and each ensemble member are single realizations of the model and have no effect whatsoever on each other...

Am I wrong to state that the OP is the extrapolation of data from all the simulations? If the average between phase and no phase is what we are seeing, than to me, the way the models have presented the data makes sense, when taken with this rudimentary approach.

Now the NAM leans toward a phase, as opposed to no phase. If the other models say phase, then the when comes into place? Right now the models have made a real mess of this!

Link to comment
Share on other sites

Here is a question that i have. On the 18z GFS, we can see the s/w coming up the western portion of NC

http://www.nco.ncep.noaa.gov/pmb/nwprod/analysis/namer/gfs/18/images/gfs_500_066l.gif

How does it jump this far off the coast only 12 hours later?

http://www.nco.ncep.noaa.gov/pmb/nwprod/analysis/namer/gfs/18/images/gfs_500_078l.gif

Link to comment
Share on other sites

I might be reading this wrong...but the OP and each ensemble member are single realizations of the model and have no effect whatsoever on each other...

So how do models come to the OP exactly, or should I say how does the NAM, GFS individually, since it may be different? Or does the computer have to literally choose amongst all the individual members for the operational one?

Link to comment
Share on other sites

a valid stance...i feel like i've had my heart broken by the euro so i don't trust anything at this point....

but the optimist in me, looked at the 12z gfs ensembles and saw that a track up the east coast was at least in the envelope of possible outcomes...and that's important...it's why we do ensembles...you are trying to assess the probability distribution function...unfortunately we don't have the computing resources to do more than 20 (gfs?) to 50 (euro?) members...but in the not too far off future...we'll be running 100s and then 1000s of perturbed simulations and then you can really statistically assess the PDF....ahh exciting stuff

Yes-- the wonders of parallel processing. Then you'll be able to get something like a quantum cluster probability chart.

Link to comment
Share on other sites

It's a good point....but it is also difficult to assess if that is having a net positive or negative effect on a given ensemble member...

low resolution can kinda work both ways...

lower res can lower the "noise" but it also dampens the signal.

Link to comment
Share on other sites

Best thing to do is take blend of all models.

I'd venture to say we get a scraper up the Mid Atl (Read: Two to sixteen flurries in dca) then probably a decent hit on New England. Don't think this is OTS; rather, probably a scraper up the MA then a hit somewhere on the NE coast. I could be wrong though - Hope I am - What do I know anyway? I'm just a snow weenie that takes up valuable bandwidth

Link to comment
Share on other sites

The operational runs of the models typically use the best researched and verified model physics and are run at highest spatial resolution as deemed relevant (depending on your application, global models are more coarse than mesoscale model for computational reasons...).

I guess the best way of thinking about it...is that the operational version of the GFS is NCEP's starting pitcher...the guy you'd want to pitch game 7 of the world series (and in most cases lose to the Europeans j/k)....

as computer resources have gotten better and faster, ensemble modeling was created, ensemble modeling takes all the data that goes into creating the initial conditions for the model and either uses different model physics or perturbs the initial conditions in a chosen statistical manner....(if someone else is more familiar with the actual perturbations used in the GFS, please feel free to further explain...i'm not familar).

why numerical weather prediction is difficult is because even though we have numerical schemes which can solve the non-linear physics of the atmospheres...producing an accurate forecast requires accurate initial conditions...something we will never likely achieve (not to mention non-representative model physics which also can lead to errors during integration of the model itself)...the use of ensemble modeling allows us to attempt to represent that uncertainty in the initial conditions and the model physics to see the range of solutions which can occur...

for example, an ensemble set with low spread...shows that very little uncertainty in the forecast...while assessing a large ensemble spread, should give a forecaster a better judgement on the level of uncertainty

So how do models come to the OP exactly, or should I say how does the NAM, GFS individually, since it may be different? Or does the computer have to literally choose amongst all the individual members for the operational one?

Link to comment
Share on other sites

The operational run is a distinct setup for whatever model you are looking at. There is one particular set of parameterizations for the OP GFS. The ensembles are comprised of perturbations from that operational setup, say a different land-surface model, or different cumulus parameterization. The OP NAM is the non-hydrostatic version of the WRF model, the WRF-ARW is the research version of the model, and would be one ensemble member. In the WRF-ARW, there are a variety of options of how ice is modeled, 3 phase, 5 phase, 6 phase, etc, all of which would be members in the overall ensemble.

The Operational version is the best guess at the best overall model for general use, but different parameterization setups could be better at specific type events, such as extratropical transition of a tropical system, or air-sea interaction over the Gulf Stream, versus a CAD event.

So how do models come to the OP exactly, or should I say how does the NAM, GFS individually, since it may be different? Or does the computer have to literally choose amongst all the individual members for the operational one?

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...