2016 Global Model Runs Discussion (Out to Day 16)

[LarryWx]

The Euro is, indeed, King. If this chart were extended back to the 1990's, it would show the Euro doing better than the GFS. Over any period of note, the Euro has had a lower average 5 day error at 500 mb. However, I also want to point out the readers that the difference in accuracy, especially over the last two years, isn't so great that the Euro is going to be the closest in anywhere near all cases. If the Euro were .915 but the GFS something like only 0.750, that would be a far different story and I'd bet on the Euro over the GFS almost every time I'd be betting if I were a betting man. As it is, the gap is shrinking.

This is only one metric as well. I find it very interesting that overall both models seem to track each others skill score. If the Euro is having a hard time so is the GFS.

Very good point. They definitely do track each other. Related to this, I just noticed something that seems very strange to me. The models have every single year on this chart done have done best in winter and worst in summer. I would have thought it would be the opposite, especially with this addressing 20N-80N. I'm guessing it is due to the big misses related to the tropics. I guess that a good number of the misses are so big near TC's that these misses dominate the averages.

[ThetaE]

The Euro is, indeed, King. If this chart were extended back to the 1990's, it would show the Euro doing better than the GFS. Over any period of note, the Euro has had a lower average 5 day error at 500 mb. However, I also want to point out the readers that the difference in accuracy, especially over the last two years, isn't so great that the Euro is going to be the closest in anywhere near all cases. If the Euro were .915 but the GFS something like only 0.750, that would be a far different story and I'd bet on the Euro over the GFS almost every time I'd be betting if I were a betting man. As it is, the gap is shrinking.

This is only one metric as well. I find it very interesting that overall both models seem to track each others skill score. If the Euro is having a hard time so is the GFS.

Very good point. They definitely do track each other. Related to this, I just noticed something that seems very strange to me. The models have every single year on this chart done have done best in winter and worst in summer. I would have thought it would be the opposite, especially with this addressing 20N-80N. I'm guessing it is due to the big misses related to the tropics. I guess that a good number of the misses are so big near TC's that these misses dominate the averages.

Ryan Maue actually addressed this exact topic in a reply to that tweet. This only represents accuracy in the northern hemisphere. During the summer, convection and smaller-scale features are more prominent, which make weather forecasting less accurate. Larger-scale patterns are more prominent in the winter months, and thus models are more accurate. I don't really think it has much to do with TCs in particular... Honestly, I think we are more prone to notice model inaccuracies when the exact forecast is much more important, such as when a hurricane threatens land. Day-to-day weather events are mundane, and we just don't look at them with the same amount of scrutiny, so I think in those situations errors tend to go unnoticed.

This is just speculation on human psychology, though, and it's a bit off-topic. It's a neat graphic, and it's very informative in many ways.

[LarryWx]

^Thanks. I follow you, but I must say that this still seems counterintuitive due to the much larger changes from day to day in the 500 mb heights in winter vs summer, especially down in the 20N-50N area. Does the convection of summer and do the smaller scale features of summer cause more variance/unpredictability at 500 mb than the always changing large scale 500 mb patterns of winter? That is hard for me to accept even though it may be right based on more stable 500 mb heights in the temperate regions.
If an area gets thunderstorms on day 5 vs not getting them, are the 500 mb heights going to vary all that much between the two scenarios?

I just thought of something else. Keep in mind that this chart covers 20N-80N. Maybe way up in the 50-80N area 500 mb hts are much more changeable in summer vs winter, thus perhaps more than negating the 20n-50N less changeable summer vs winter at 500 mb? We tend to think more about the temperate regions since we live there but what about the polar regions?

[otowntiger]

Any models showing anything coming out of the western Caribbean like they were last week?

[AJC3]

Any models showing anything coming out of the western Caribbean like they were last week?

It has its own thread...

http://storm2k.org/phpbb2/viewtopic.php ... b61a0cb3c5

[Digital-TC-Chaser]

two area's so posted here.

[otowntiger]

Any models showing anything coming out of the western Caribbean like they were last week?

It has its own thread...

http://storm2k.org/phpbb2/viewtopic.php ... b61a0cb3c5

Ok, thanks!

[blp]

I expect we may see some activity first week of November. I have been watching the MJO for several days and it has been trending toward an upward motion push into our region during this time. The GFS is very aggressive with the MJO and probably too aggressive while the Euro is also leaning toward entering our basin but much less agressive. The GFS which is the only model right now going out that far in that long range has been showing development last several runs in the SW Carribean. Let's see if we get some late season action.

GFS:


Euro:


GFS different view:

[StormHunter72]

cold front coming in this weekend. Looks like the GOM will be closing soon.

[LarryWx]

Interestingly, 10 Oct's since 1995 have had some days in phase 8 (I mean outside the circle) but none had a genesis during that phase that later became a TS+. So, it has actually been the least active for genesis of all of the MJO phases since 1995 with a shutout of 0 for 10: 1996, 97, 01, 03, 04, 06, 07, 09, 11, and 13. These 10 took up 44 days. So, 0 geneses in 44 Oct phase 8 days since 1995 for TC's that became TS+. (Phase 1 has been much more genesis friendly in Oct.)

In November, however, there were two phase 8 geneses since 1995 out of 9 instances of Nov having some days during phase 8 (28 total phase 8 days):

1) TS Gamma of 2005 in W. Caribbean: http://weather.unisys.com/hurricane/atl ... /track.gif

2) H Marco of 1996 in W. Caribbean: http://weather.unisys.com/hurricane/atl ... /track.gif

So, based on this, a W. Caribbean phase 8 TC genesis in November wouldn't be all that surprising should it occur. If it were to occur in Oct. it would be the first going back to 1985.

Going back to 1975, H Juan of 1985 is the only Oct genesis during phase 8 (that later became a TS+). There were 17 Oct's since 1975 that had at least a day during phase 8. So, only 1 in 17 Oct's since 1975 that had a portion in phase 8 that had a genesis during that phase:

H Juan in GOM: http://weather.unisys.com/hurricane/atl ... /track.gif

[WeatherEmperor]

GFS shows something around 240hrs developing deep in the SW Carib and here is how it ends in fantasy land. The GFS has shown this to some extent for a few runs now. Still far in fantasy range...but lets see if anything happens when the timeframe closes into reality. At the very least it is something to watch as the season winds down.






Sent from my iPhone 7 using Tapatalk

[blp]

GFS shows something around 240hrs developing deep in the SW Carib and here is how it ends in fantasy land. The GFS has shown this to some extent for a few runs now. Still far in fantasy range...but lets see if anything happens when the timeframe closes into reality. At the very least it is something to watch as the season winds down.

[http://uploads.tapatalk-cdn.com/20161021/a01dde79d8650436e3b4b2193bda3e28.png[/img]

[]http://uploads.tapatalk-cdn.com/20161021/8adaa0996ab2c692ebf6b96cd4535b78.png[/img]


Sent from my iPhone 7 using

Yes I have been watching as well and the CMC has been showing something the last three days straight. The MJO is predicted to be favorable. I think the GFS has done well this year in the long range picking up development. Not as many phantoms as years past. The Euro has been a little slower to catch up this year on cyclogensis so we will see if it starts showing something this weekend. Still early but might be our last shot for development.

[WeatherHoon]

Both the CMC (multiple runs now) and GFS are picking up on a system trying to develop starting this upcoming Wednesday in the WCarib. The GFS depicts nothing more than a weak low, but tries to get it going a bit near the Carolinas before it's ultimately shredded up. Then, of course, we have the CMC blowing it up into a formidable Hurricane Given the not-so-far-out timeframe, it seems that it bears some watching.

[GCANE]

Both the CMC (multiple runs now) and GFS are picking up on a system trying to develop starting this upcoming Wednesday in the WCarib. The GFS depicts nothing more than a weak low, but tries to get it going a bit near the Carolinas before it's ultimately shredded up. Then, of course, we have the CMC blowing it up into a formidable Hurricane Given the not-so-far-out timeframe, it seems that it bears some watching.

It about the right time in terms of climo to develop something here and given that the upper troposphere is cooling off, its not out of the question this may spin up IMHO.

[Frank2]

An election day hurricane - that'd throw things in a spin if it threatened the USEC by Nov 8 (per he GFS it'd be in the SW Carib on Nov 6)...

[WeatherEmperor]

GFS has a hurricane in the SE Gulf. The Canadian has a hurricane hitting eastern Cuba. Timeframe about 10 days out.






Sent from my iPhone 7 using Tapatalk

[Tampa Bay Hurricane]

GFS appears to be throwing random cyclones around lol. Yesterday it had the WCarib system on Days 15-16. Today it has a WGOM system much earlier in the run. Will watch for persistence. With all the fronts being forecasted to sweep through by the GFS it will be a challenge to get a system into the N or C GOM.

[floridasun78]

Mid/Late Next Week...moisture levels will trend upward through
the end of next week as moderate/strong low level NE-E flow
increases and mid/upper level low pressure over the Gulf and NW
Caribbean influences our regional pattern. my question this non tropical low miami weather office talking about or tropical low have models show any thing this area by next week those area

[Steve]

I think 12z GFS isn't necessarily that bad. It's too far west, but it wouldn't be impossible to see something come up from the BoC with some of the analog years. I think it would likely be more like Opal than a hit on SWLA and bounce back into the gulf. I wouldn't rule out a later season crossover or pile up of energy down there sort of in the Juan (referenced above) or winter storm type scenario.

But it seems in that timeframe it would be more likely to see something come up on the eastern side of the Yucatan. 18z is running now, but it may be too soon for it to more accurately hone in on any Genesis farther east.

[WeatherHoon]

I'm more interested in what ends up happening with the low-pressure system that both the CMC (shows a hurricane further in the run) and GFS(shows weak low) see forming this Wednesday.




While I doubt it's going to amount to much, it's more believable than election day phantom cane.

On the subject of said phantom cane, could one just push through a front like that?