phil882 Posted August 10, 2012 Share Posted August 10, 2012 I should have been more specific before I set off a firestorm. I am referring to the global models BEFORE Ernesto was in the Caribbean. The models did a much better job when Ernesto was around 70W and beyond, and the GFS preformed superior to the ECMWF in my view of things. The point I was trying to make is that the models improperly forecasted the CCKW that was moving through the Caribbean / Atlantic that interacted with both Ernesto and Florence. When Ernesto was still a depression, the ECMWF was weakening the system in the Caribbean, and the GFS, while slightly stronger, also dissipated the system on several runs. Once the CCKW moved over the system and was just past it, Ernesto had become a tropical storm and was now located in the Eastern Caribbean. Due to the stronger than expected reflection of the storm, the models became more agressive in the Caribbean and had Ernesto surviving and then going on to intensify over the Western Caribbean. We saw a very similar situation with Florence, which developed abruptly with a poor lead time from both the GFS and ECMWF which only had a weak disturbance. Link to comment Share on other sites More sharing options...
dtk Posted August 10, 2012 Share Posted August 10, 2012 Ed is correct. Certainly for the GFS which has an equivalent grid spacing of ~25-30km in the tropics. However, the ECMWF model has struggled occasionally (this is opinion only here) in the tropics despite its superior resolution (on the order of equivalent grid spacking of ~12km or so, I think). This may not be entirely surprising, as they are in a resolution zone known to be a bit of a grey area (i.e. not high enough resolution to be convection permitting, but still high enough resolution that convective parameterizations get quite messy). Link to comment Share on other sites More sharing options...
am19psu Posted August 10, 2012 Share Posted August 10, 2012 I should have been more specific before I set off a firestorm. I am referring to the global models BEFORE Ernesto was in the Caribbean. The models did a much better job when Ernesto was around 70W and beyond, and the GFS preformed superior to the ECMWF in my view of things. The point I was trying to make is that the models improperly forecasted the CCKW that was moving through the Caribbean / Atlantic that interacted with both Ernesto and Florence. When Ernesto was still a depression, the ECMWF was weakening the system in the Caribbean, and the GFS, while slightly stronger, also dissipated the system on several runs. Once the CCKW moved over the system and was just past it, Ernesto had become a tropical storm and was now located in the Eastern Caribbean. Due to the stronger than expected reflection of the storm, the models became more agressive in the Caribbean and had Ernesto surviving and then going on to intensify over the Western Caribbean. We saw a very similar situation with Florence, which developed abruptly with a poor lead time from both the GFS and ECMWF which only had a weak disturbance. And the proper lesson should be: global models have a difficult time forecasting the effects of Kelvin waves (which if you are good at forecasting these things, you already knew), not that global models are poor at forecasting tropical cyclones. Link to comment Share on other sites More sharing options...
mappy Posted August 10, 2012 Share Posted August 10, 2012 Bendy Mod ftw Link to comment Share on other sites More sharing options...
am19psu Posted August 10, 2012 Share Posted August 10, 2012 P.S. Phil, that was royal "you," not directed at you Link to comment Share on other sites More sharing options...
phil882 Posted August 10, 2012 Share Posted August 10, 2012 And the proper lesson should be: global models have a difficult time forecasting the effects of Kelvin waves (which if you are good at forecasting these things, you already knew), not that global models are poor at forecasting tropical cyclones. Yes agreed... but my offshoot to that is that there is a big ??? when a tropical cyclone interacts with said Kelvin Wave... which is what TD#7 will be doing in about 24-36 hours, right when the shear is expected to increase. I am not discounting the possibility that this system will dissipate, that is certainly possible. I just don't think its set in stone yet, especially given that the models might not be properly accounting for the beneficial effects of the CCKW. Link to comment Share on other sites More sharing options...
H2O Posted August 10, 2012 Share Posted August 10, 2012 No, they obviously didn't perform well. Unfortunately, the vast majority of mets in this forum are sellouts, just like every other business. They work for companies that RELY on GFS and ECMWF output, and can't say that they didn't perform well, or that they are useless with one thing or another, for fear of losing their job, or not getting that coveted promotion. It's sad in this day and age that people value selling out to the man more than making a correct forecast, but I can kind of understand it in a bad economy, too. lolz to the first part and megalolz to the second. Clearly they should rely on the JMA more. Link to comment Share on other sites More sharing options...
am19psu Posted August 10, 2012 Share Posted August 10, 2012 Yes agreed... but my offshoot to that is that there is a big ??? when a tropical cyclone interacts with said Kelvin Wave... which is what TD#7 will be doing in about 24-36 hours, right when the shear is expected to increase. This is like Ernesto again all over... Yep, we're on the same page. Link to comment Share on other sites More sharing options...
TUweathermanDD Posted August 10, 2012 Share Posted August 10, 2012 I should have been more specific before I set off a firestorm. I am referring to the global models BEFORE Ernesto was in the Caribbean. The models did a much better job when Ernesto was around 70W and beyond, and the GFS preformed superior to the ECMWF in my view of things. The point I was trying to make is that the models improperly forecasted the CCKW that was moving through the Caribbean / Atlantic that interacted with both Ernesto and Florence. When Ernesto was still a depression, the ECMWF was weakening the system in the Caribbean, and the GFS, while slightly stronger, also dissipated the system on several runs. Once the CCKW moved over the system and was just past it, Ernesto had become a tropical storm and was now located in the Eastern Caribbean. Due to the stronger than expected reflection of the storm, the models became more agressive in the Caribbean and had Ernesto surviving and then going on to intensify over the Western Caribbean. We saw a very similar situation with Florence, which developed abruptly with a poor lead time from both the GFS and ECMWF which only had a weak disturbance. This is how it should be said, not like amp who says you're not even close to right. Why dont you use the meteorological knowledge you might have to say the convectively coupled kelvin wave in tandem with the westerly motion wasn't adequately picked out initially by the globals. I give the GFS moderate scores overall, doing very well on track for the most part and in terms of intensity it may have been low but it picked out the strengthening trend well. All and all the GFS was solid, though the Euro wasnt as good. Enough from me. Link to comment Share on other sites More sharing options...
MississippiWx Posted August 10, 2012 Share Posted August 10, 2012 I should have been more specific before I set off a firestorm. I am referring to the global models BEFORE Ernesto was in the Caribbean. The models did a much better job when Ernesto was around 70W and beyond, and the GFS preformed superior to the ECMWF in my view of things. The point I was trying to make is that the models improperly forecasted the CCKW that was moving through the Caribbean / Atlantic that interacted with both Ernesto and Florence. When Ernesto was still a depression, the ECMWF was weakening the system in the Caribbean, and the GFS, while slightly stronger, also dissipated the system on several runs. Once the CCKW moved over the system and was just past it, Ernesto had become a tropical storm and was now located in the Eastern Caribbean. Due to the stronger than expected reflection of the storm, the models became more agressive in the Caribbean and had Ernesto surviving and then going on to intensify over the Western Caribbean. We saw a very similar situation with Florence, which developed abruptly with a poor lead time from both the GFS and ECMWF which only had a weak disturbance. I think it is important to emphasize that the models weren't "dissipating" the system in the Caribbean. They kept the wave/trof/low alive throughout the Caribbean. If you were watching recon when it investigated Ernesto in the Central Caribbean, you will remember that recon revealed that Ernesto's wind field was very similar to what the global models were showing: a strong open wave. The HH literally found no west winds on consecutive investigations. However, due to Ernesto's appearance, strength, track, and predicted strength in the Western Caribbean the NHC never downgraded it which was a good call. How can a global model show a tropical cyclone when in reality there was not one there due to no closed circulation being in place? The GFS/Euro showed strengthening back into a tropical cyclone in the Western Caribbean, which proved to be correct. Link to comment Share on other sites More sharing options...
mappy Posted August 10, 2012 Share Posted August 10, 2012 No, they obviously didn't perform well. Unfortunately, the vast majority of mets in this forum are sellouts, just like every other business. They work for companies that RELY on GFS and ECMWF output, and can't say that they didn't perform well, or that they are useless with one thing or another, for fear of losing their job, or not getting that coveted promotion. It's sad in this day and age that people value selling out to the man more than making a correct forecast, but I can kind of understand it in a bad economy, too. Didn't you quit the business of being a met? Why do you even have a red tag? Link to comment Share on other sites More sharing options...
NittanyWx Posted August 10, 2012 Share Posted August 10, 2012 No, they obviously didn't perform well. Unfortunately, the vast majority of mets in this forum are sellouts, just like every other business. They work for companies that RELY on GFS and ECMWF output, and can't say that they didn't perform well, or that they are useless with one thing or another, for fear of losing their job, or not getting that coveted promotion. It's sad in this day and age that people value selling out to the man more than making a correct forecast, but I can kind of understand it in a bad economy, too. My company doesn't give a **** which model I use as long as I'm right. Link to comment Share on other sites More sharing options...
Mencken_Fan Posted August 10, 2012 Share Posted August 10, 2012 One thing I've notice over the years is people's expectations often exceed model capability no matter how many times they're improved. The only thing that remains the same is human nature. Chasers and those in a forecast cone may get frazzled nerves but for the rest.... relax! Link to comment Share on other sites More sharing options...
H2O Posted August 10, 2012 Share Posted August 10, 2012 Guess the NWS is a bunch of sellouts Link to comment Share on other sites More sharing options...
Ed Lizard Posted August 10, 2012 Share Posted August 10, 2012 Didn't you quit the business of being a met? Why do you even have a red tag? I'm not naming names, I made a slight mention to red tags coming in different varieties. I believe board policy is anyone with a met degree is entitled to have a red tag. Some are clearly better than others. Like any profession. Some red taggers will start threads after a model bust on a snow storm and earn Weenie of the Year honors, some red taggers, like Adam and B_I, are so respected they are elevated to staff. Why I mentioned somewhere, even though it is Storm2K-ish, certain amateurs who have demonstrated clear understanding of concepts, Jorge and Don Sutherland come to mind, I'm finding GaWx useful in the ENSO threads, might be worthy of a tag for "well qualified amateur", perhaps decided among the mods and red taggers and out of site sight of us regular users. Their/There/They're edit... Edit II for below - this may be as close as I ever come to an OT thread again. Link to comment Share on other sites More sharing options...
Analog96 Posted August 10, 2012 Share Posted August 10, 2012 lolz to the first part and megalolz to the second. Clearly they should rely on the JMA more. Nobody mentioned the JMA, we were talking about tropical models. Name some for me, waterboy. Link to comment Share on other sites More sharing options...
Analog96 Posted August 10, 2012 Share Posted August 10, 2012 Didn't you quit the business of being a met? Why do you even have a red tag? Is a retired doctor still worthy of being called a doctor? Link to comment Share on other sites More sharing options...
Analog96 Posted August 10, 2012 Share Posted August 10, 2012 My company doesn't give a **** which model I use as long as I'm right. Bravo for your company then, seriously! Link to comment Share on other sites More sharing options...
Analog96 Posted August 10, 2012 Share Posted August 10, 2012 Guess the NWS is a bunch of sellouts A lot of them, yes, although over there it has gotten better recently. I heard from a reliable source that up until somewhere in the 1990s they were encouraged to use the AVN/MRF over the ECMWF, just because it was American, regardless of accuracy. Link to comment Share on other sites More sharing options...
Ed Lizard Posted August 10, 2012 Share Posted August 10, 2012 I have a non-argument question about CCKW and TD #7 in the TD #7 thread for Phil, BTW. Link to comment Share on other sites More sharing options...
phil882 Posted August 10, 2012 Share Posted August 10, 2012 I think it is important to emphasize that the models weren't "dissipating" the system in the Caribbean. They kept the wave/trof/low alive throughout the Caribbean. If you were watching recon when it investigated Ernesto in the Central Caribbean, you will remember that recon revealed that Ernesto's wind field was very similar to what the global models were showing: a strong open wave. The HH literally found no west winds on consecutive investigations. However, due to Ernesto's appearance, strength, track, and predicted strength in the Western Caribbean the NHC never downgraded it which was a good call. How can a global model show a tropical cyclone when in reality there was not one there due to no closed circulation being in place? The GFS/Euro showed strengthening back into a tropical cyclone in the Western Caribbean, which proved to be correct. I am going to try to add a bit of objectivity to this discussion. Here is a dprog/dt map animation verifying on 1200 UTC 7 August. It goes though all of the GFS solutions from 180 hours to the analysis verifying time. 925-850 layer mean relative vorticity is contoured in black, with the more closely packed the contours are, the stronger the vorticity. Notice that the solutions up to 54 hours out are all over the place.. some developing a decent TC, while others had the system degenerating into an open trough that was shunted southward into Central America. The point is the GFS didn't figure out this system was going to be a substancial TC in the Western Caribbean for sure until about 48 hours before verification. There were model runs that hinted this possibility, but there was no consistency. Right now the GFS has TD#7 weakening to an open trough in the Central Caribbean, right around where the GFS was dissipating Ernesto in this time frame on some of the runs.. I'm certainly not saying its not going to happen (in fact its probably more likely than TD#7 intensifying substantially in the east or central Caribbean). However, it leaves the door open to speculate on other features influencing the organization of TD#7 that aren't typically captured well in the modeling. Link to comment Share on other sites More sharing options...
isohume Posted August 10, 2012 Share Posted August 10, 2012 A lot of them, yes, although over there it has gotten better recently. I heard from a reliable source that up until somewhere in the 1990s they were encouraged to use the AVN/MRF over the ECMWF, just because it was American, regardless of accuracy. Lol! Thanks for the hearty laugh! As if this thread wasn't funny enough, I stumbled upon this post. Link to comment Share on other sites More sharing options...
mappy Posted August 10, 2012 Share Posted August 10, 2012 Is a retired doctor still worthy of being called a doctor? Yes. I'm still waiting for you to out forecast Adam Link to comment Share on other sites More sharing options...
Ed Lizard Posted August 10, 2012 Share Posted August 10, 2012 From Ernesto thread BTW- a forecast from the Eastern Caribbean for a Nicaragua landfall as compared to an actual landfall near the MX/BZ border isn't terribly wrong, if one looks at the geography... There's some discussion of verification for Ernesto in the other thread. We should probably move it to here guys. I saved this one from 18Z on the 4th. The GFS (green line, labeled AVNI) absolutely nailed the track. Maybe the GFS intensity was a bit off, but correctly depicting pulsing convection was never the GFS's strong point. Link to comment Share on other sites More sharing options...
Hoosier Posted August 10, 2012 Share Posted August 10, 2012 Anyhoo, I see what a lot of you are saying, and I was probably too severe on those two models-- point taken. But I stand by my original point which is that I'm not writing TD7 off because the Euro and the GFS said so. And you can't make me. I'm far from a tropics guy but isn't intensity (forecasting and modeling) much more prone to major error than track? If that's the case, then you, being a leading tropics enthusiast here, should know that model and observational trends are of paramount importance. Link to comment Share on other sites More sharing options...
wxsmwhrms Posted August 10, 2012 Share Posted August 10, 2012 A lot of them, yes, although over there it has gotten better recently. I heard from a reliable source that up until somewhere in the 1990s they were encouraged to use the AVN/MRF over the ECMWF, just because it was American, regardless of accuracy. This is crap. I have worked for NOAA/NWS since the late '80s, and we have never been "encouraged" to use ANY one model, and I have never heard of such a thing. And I also want to agree with MS Weather about the model depicition of Ernesto. In spite of comments to the contrary at the time, the models did not dissipate that system. In fact, the ECMWF correctly showed better organization and intensification of the low level vorticity center as it approached the Yucatan. Did it forecast a 975 mb hurricane? Of course not. Did it get the trend right? Yes. And it did much better on the trend than the LGEM/SHIPS which consistently showed intensification way too early. Link to comment Share on other sites More sharing options...
HurricaneJosh Posted August 10, 2012 Author Share Posted August 10, 2012 Well, I have to apologize for inadvertently adding gasoline to this firestorm with some rather dismissive remarks Re: the Euro and the GFS. Adam, MississippiWx, and Ed made some good points and I see what y'all are saying. Don't mind me-- I'm just a chaser dude. P.S. I still ain't trusting them for TD7. Link to comment Share on other sites More sharing options...
Ed Lizard Posted August 10, 2012 Share Posted August 10, 2012 Well, I have to apologize for inadvertently adding gasoline to this firestorm with some rather dismissive remarks Re: the Euro and the GFS. Adam, MississippiWx, and Ed made some good points and I see what y'all are saying. Don't mind me-- I'm just a chaser dude. P.S. I still ain't trusting them for TD7. The good news is that one of the globals, the 12Z Canadian, at least holds out hope for a Tampico chase. And the remnants of Ernesto 94E will be approaching SoCal, per that model. Some of the GFS ensemble members at 192 have a closed low in the BoC as well. Link to comment Share on other sites More sharing options...
andyhb Posted August 10, 2012 Share Posted August 10, 2012 No, they obviously didn't perform well. Unfortunately, the vast majority of mets in this forum are sellouts, just like every other business. They work for companies that RELY on GFS and ECMWF output, and can't say that they didn't perform well, or that they are useless with one thing or another, for fear of losing their job, or not getting that coveted promotion. It's sad in this day and age that people value selling out to the man more than making a correct forecast, but I can kind of understand it in a bad economy, too. Forgive me, but what in the hell are you talking about? Edit: Nevermind, it has been made clear in succeeding posts. I'm pretty sure most of us in this forum and most mets elsewhere would have no problem pointing out when a model blows/is inconsistent in forecasting a setup of a tropical system, severe weather outbreak, snowstorm or otherwise. Link to comment Share on other sites More sharing options...
Analog96 Posted August 10, 2012 Share Posted August 10, 2012 Yes. I'm still waiting for you to out forecast Adam I'm still waiting for someone to issue an official challenge. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.