Atlantic Tropical Action 2013

[TropicalAnalystwx13]

We may have something to watch over the next week after all.

 

The CMC and GFS are in agreement for the large upper-level low southwest of the Azores to become elongated as ridging develops from the west and east. A piece of energy from this trough is pinched off and retrogrades southwestward over the southwestern Atlantic. Continued ridging forecasts the low westward towards the Bahamas and Florida in the model plots.

 

Given the amount of the time, I could see something working down to the surface from it. 

 

GFS takes it into south Texas without consequence, while the CMC takes it into Louisiana as a 985 millibar hurricane.

[Floydbuster]

We may have something to watch over the next week after all.

 

The CMC and GFS are in agreement for the large upper-level low southwest of the Azores to become elongated as ridging develops from the west and east. A piece of energy from this trough is pinched off and retrogrades southwestward over the southwestern Atlantic. Continued ridging forecasts the low westward towards the Bahamas and Florida in the model plots.

 

Given the amount of the time, I could see something working down to the surface from it. 

 

GFS takes it into south Texas without consequence, while the CMC takes it into Louisiana as a 985 millibar hurricane.

 

If memory serves that is how Hurricane Erika formed in 2003.

[Mike.Ventrice]

I'm noticing a warm-cored low pressure system tracking westward from the western tropical Atlantic to over FL in the GFS this morning between July 1-3... anyone else seeing this feature? It is there before the GFS degrades in resolution so it doesn't appear the model is struggling from convective parameterization issues....

 

 

[Ed Lizard]

^ 12Z Euro that as a weak feature in SLP and as an 850 mb vorticity feature,  and OP 12Z and 18Z GFS showed that as a 700 m feature next week, approaching Florida at GFS resolution reduction.  12Z Euro carried it into Gulf, but without signs of intensification.

 

New Euro has a bend in the isobars and an 850 mb vort max at 9 days, which apparently moves into Louisiana and weakens on day 10 on PSU 10 day tropical SLP/850 mb vorticiy presentation.

 

I don't think anything actually develops, from what I see.

[Sunny and Warm]

with all the rain along the east coast during June, is that telegraphing that the mean trough will set up on the eastern 1/3 of the nation through August.  I was reading Rainshadow's post in the Philly forum where he stated that the 10 wettest June's on record there ALWAYS begat above normal rainfall in the July and August combo.  Something to consider as we go forward, especially as latest guidance seems to be hinting at a FL to Gulf Coast threat moreso than an EC one for later this season (before mid Sept). 

[Srain]

Mike, we've been monitoring the progression of that feature you mentioned as well as a robust tropical wave entering the Western Gulf late week. A combination of a 5H low over Texas nearly 4-5 standard deviations below normal as well as a surge of PW's in the 2.2/2.4 range as well as a stalled boundary and some potential for a tropical low to spin up in the Bay of Campeche tends to spell trouble and can  bring core rain events along the Central Texas Coast on E into Louisiana. Needless to say we along the Western Gulf are following the trends rather closely with the long Holiday weekend ahead. The wind will not be the main story line from this potential event, but sloppy tropical systems that brought copious rainfall amount events of the past such as Claudette 1979/Allison 2001 come to mind and certainly warrants some monitoring.

[Ed Lizard]

I'd still like to know what the above product keys off of, because individual GEFS members showing closed 1008 mb lows look nothing like that image above which suggests TC genesis is almost inevitable.

 

 

The GFS/Euro Day 8 to 10 wave crossing over Florida into the Gulf, apparently becomes a hurricane that hits Louisiana in 10 days, just in case one wonders whether the resolution increase helped the Canadian Global any.

[icebreaker5221]

I'm noticing a warm-cored low pressure system tracking westward from the western tropical Atlantic to over FL in the GFS this morning between July 1-3... anyone else seeing this feature? It is there before the GFS degrades in resolution so it doesn't appear the model is struggling from convective parameterization issues....

 

 

Those e-wall maps can be a bit deceiving.  That's actually an area of less than 570 dam thickness associated with a very cold upper-level trough that stands out as a closed contour within an otherwise very warm airmass.  This isn't to say that we can't see a tropical-transition type genesis event further down the line in the Gulf with said feature. 

 

This is a good site for thickness anomalies (and many other handy plots)

http://www.instantweathermaps.com/GFS-php/natlupper.php

 

12Z GFS still has it. 

[Ed Lizard]

At Hour 192, GFS system entering Gulf from Florida has a closed 500 mb low, a decent 700 mb reflection although relative humidities are low, a hint of a reflection in 850 mb wind fields, and some showers mainly East of the closed 500 mb low.

 

I'd guess it won't have a chance to work its way down to the surface before it runs out of water in the Northern or Northwestern Gulf.

 

But at least we have something to pay attention to during the July doldrums.

[Top of the Bay]

Curious for input on what appears to my untrained eye a circulation attempting to form in these frames.

http://www.myfoxtampabay.com/category/235474/gulf-of-mexico-enhanced-satellite-view

[Ed Lizard]

Curious for input on what appears to my untrained eye a circulation attempting to form in these frames.gulf_mer_enh_ir_sat_1.jpg

http://www.myfoxtampabay.com/category/235474/gulf-of-mexico-enhanced-satellite-view

 

 

There is an upper level low there.  It shows up well on WV imagery and CIMSS 200 mb vorticity product.

 

 

AS discussed on a previous post, upper level and lower level lows can eventually become tropical, but it is a very slow process.  The red tagged people can explain it better.

 

That Florida 192 hour thing, that'd be cool if something started North of the Gulf and well East of here and actually hit Louisiana as predicted by the Canadian.  Sadly, the Canadian is rarely correct.

[hurricaneman]

The trough split possibilities just went up to 30% with the reliable models jumping on board, an analog storm as a possibility would be Edouard in 2002, other storms that did become formidable to dangerous with similar trough splits are Bob in 1991 and Gaston in 2004

[Ed Lizard]

Sure it isn't 25% or 33%?

[bluewave]

Interesting post by Stu Ostro:

 

http://www.wunderground.com/blog/stuostro/comment.html?entrynum=28

[GaWx]

Interesting. I'm sorry if I came across aggressive- was in a hurry this morning as I wanted to talk a bit more about this than hit my deadlines on my forecasts.

 

I'd love to see some of your results if you don't mind sharing

 

Mike,

 Actually, it isn't nearly as flat as I had thought. My prior study was concentrated in Oct./Nov. in the western portion of the Atlantic basin. The normalized results did indicate a pretty flat situation vs. the MJO phases.

 

 However, I just completed an analysis of 1995-2012 for the entire hurricane season and for the entire Atlantic basin. I'll probably post more detail in a separate thread. Here is the gist of it:

 

 I studied 286 Atlantic basin tropical cyclone formation dates within 6/1-11/30 for the years 1995-2012 and looked at the respective MJO phases. To normalize, I also counted the number of days in each MJO phase 6/1-11/30 for the same years.

 

 Whereas there were 347 phase 1 days and 346 phase 2 days, there were only 123 phase 7 days! So, normalization was crucial!

 

 Results after normalizing:

 

- phase 2 has been the most enhancing phase on a % basis : 44 storms formed vs. expected value of 30. Phase 2 has had the 2nd largest # of days, only one day lower than phase 1, but this is taken into account via normalization.

- phase 3: 20 vs. 17

- phase 4: 24 vs. 19: I never hear about phase 4 being a "favorable" phase but this analysis suggests it is favorable

 

- phase 1 actually was fairly near the expected value: 27 storms vs. expected value of 30. Phase 1 is deceiving because it has the highest # of days of the 8 phases. I often hear it is favorable, but this analysis indicates it is at best neutral.

 

- remaining phases underperformed

- phase 5: 19 storms vs. expected value of 24

- phase 6: 14 vs. 18

- phase 7: 6 vs. 11; this underperformed the most %wise; phase 7 has had by a very large margin the smallest # of days in any one phase. However, that is taken into account via normalization.

- phase 8: 12 vs. 18...2nd worst on % basis

- Circle of death performed right at the expected value: 120 vs. 120

 

So, on a % basis, here is how they all stack up starting after normalization with most enhanced:

 

1) phase 2 by far the best

2) phase 4

3) phase 3

 

The above three phases overperformed.

 

 

4) circle of death: flat

 

5) phase 1 pretty flat

 

 

The rest underperformed though only phases 7 and 8 underperformed significantly %wise:

 

 

6) phase 5

7) phase 6

8) phase 8

9) phase 7 the worst even after normalization. It had less than half the % of phase 2.

[Mike.Ventrice]

Mike,

 Actually, it isn't nearly as flat as I had thought. My prior study was concentrated in Oct./Nov. in the western portion of the Atlantic basin. The normalized results did indicate a pretty flat situation vs. the MJO phases.

 

 However, I just completed an analysis of 1995-2012 for the entire hurricane season and for the entire Atlantic basin. I'll probably post more detail in a separate thread. Here is the gist of it:

 

 I studied 263 Atlantic basin tropical cyclone formation dates within 6/1-11/30 for the years 1995-2012 and looked at the respective MJO phases. To normalize, I also counted the number of days in each MJO phase 6/1-11/30 for the same years.

 

 Whereas there were 347 phase 1 days and 346 phase 2 days, there were only 123 phase 7 days! So, normalization was crucial!

 

 Results after normalizing:

 

- phase 2 has been the most enhancing phase on a % basis : 44 storms formed vs. expected value of 28. Phase 2 has had the 2nd largest # of days, only one day lower than phase 1, but this is taken into account via normalization.

- phase 3: 20 vs. 15

- phase 4: 24 vs. 17: I never hear about phase 4 being a "favorable" phase but this analysis suggests it is favorable

 

- phase 1 actually was near the expected value: 27 storms vs. expected value of 28. Phase 1 is deceiving because it has the highest # of days of the 8 phases. I often hear it is favorable, but this analysis indicates it is only neutral.

 

- remaining phases underperformed

- phase 5: 19 storms vs. expected value of 22

- phase 6: 14 vs. 16

- phase 7: 6 vs. 10; this underperformed the most %wise; phase 7 has had by a very large margin the smallest # of days in any one phase. However, that is taken into account via normalization.

- phase 8: 12 vs. 17...2nd worst on % basis

- Circle of death also underperformed though not by too much on a % basis: 97 vs. 110

 

So, on a % basis, here is how they all stack up starting after normalization with most enhanced:

 

1) phase 2 by far the best

2) phase 4

3) phase 3

 

The above three phases overperformed.

 

 

4) phase 1 flat

 

 

The rest underperformed though only phases 7 and 8 underperformed significantly %wise:

 

5) circle of death

6) phase 6

7) phase 5

8) phase 8

9) phase 7 the worst even after normalization. It had less than half the % of phase 2.

 

Ah ok this is more in lines with my previous research and thoughts. If you look at just the MDR, you will see an uptick in activity during phase 1, but phases 2 and 3 steal the show. This is a time where you MJO convection pushes northward over the Indian Peninsula.  

 

The real signal is in tropical cyclone intensity. If you do the same study, but for hurricane days, the signal is just awesome! Check out the final figure of this (hopefully) future manuscript:

 

http://www.atmos.albany.edu/student/ventrice/documents/VPM_Manuscript.pdf

[Mike.Ventrice]

GFS still dropping the ball in the short-medium range with regards to correctly representing the tropics. I'd be very wary using this model until is straightens itself out. 

 

[Drz1111]

Ah ok this is more in lines with my previous research and thoughts. If you look at just the MDR, you will see an uptick in activity during phase 1, but phases 2 and 3 steal the show. This is a time where you MJO convection pushes northward over the Indian Peninsula.  

 

The real signal is in tropical cyclone intensity. If you do the same study, but for hurricane days, the signal is just awesome! Check out the final figure of this (hopefully) future manuscript:

 

http://www.atmos.albany.edu/student/ventrice/documents/VPM_Manuscript.pdf

Why the discontinuity in hurricane days between phase 8 and phase 1?

[Mike.Ventrice]

Why the discontinuity in hurricane days between phase 8 and phase 1?

 

So in phase 8, the Atlantic is still in the upper-level westerly wind phase of the MJO, and presumably more enhanced shear over the Atlantic. But by Phase 1, your active phase shifts over Africa- it's more of a rapid process and shear will turn to more easterly over the Atlantic. This occurs during a time where trades reduce and westerly wind anomalies develop in the lower troposphere. Due to the westward tilted with height nature of the MJO/CCKWs over the Atlantic, this is also a time where mid-tropospheric moisture is increased.

[Ed Lizard]

Assuming it is just a shade too early in the season for the currently favorable phases to work any magic, when, in your estimation, does the favorable period return and peak in August.  (Actually, even if a surprise July storm forms, when is the next happy time closer to the meat and potatoes season?)

 

Is there a better way than just eyeballing the CPC filtered OLR history and guessing the return frequency?

[das]

Good news everyone! The Cayman Islands have installed a radar on Grand Cayman and the data is online!

 

http://www.weather.gov.ky/portal/page?_pageid=4421,7428417&_dad=portal&_schema=PORTAL

 

Glad to see that it's up.  It's a curious communications design for the data feed.  Wireless from the tower 10km to the SW to the local prison then via land line to the Government Admin building in Georgetown.  Radar data finally then rides fiber to the (Cayman) NWS site out at the airport.  It'll be interesting to see how it holds up during a significant storm.

 

Oh, and, they designated and trained one NWS support tech for the maintenance and calibration actions.  Two non-NWS personnel were trained but their primary work roles are at the airport and telecommunications offices.

 

At least it's a solid Selex-Gematronik system so it should be pretty robust.

[GaWx]

Ah ok this is more in lines with my previous research and thoughts. If you look at just the MDR, you will see an uptick in activity during phase 1, but phases 2 and 3 steal the show. This is a time where you MJO convection pushes northward over the Indian Peninsula.  

 

The real signal is in tropical cyclone intensity. If you do the same study, but for hurricane days, the signal is just awesome! Check out the final figure of this (hopefully) future manuscript:

 

http://www.atmos.albany.edu/student/ventrice/documents/VPM_Manuscript.pdf

 

Mike,

 I decided to check the MDR vs. non-MDR distribution for TC genesis days for 1995-2012 for 6/1-11/30 for each MJO phase and I divided by the # of days in each respective MJO phase. My def. of the MDR is the area at or below 20 N and at or E of 60 W. The phase with the lowest % of days with MDR genesis was phase 7, which had only a 1% daily MDR genesis rate. Then it rose gradually from phase 7 to phase 3, where it maxed at a 5% daily MDR genesis rate. Subsequently, it gradually fell back to the 1% daily MDR genesis rate of phase 7. The circle of death as well as overall % were both at 3%.

 

MDR TC daily genesis rate 6/1-11/30

 

Phase 1: 3% (Ivan from here)

Phase 2: 4% ***

Phase 3: 5% (Fran from here)

Phase 4: 4%

Phase 5: 3%

Phase 6: 3%

Phase 7: 1%

Phase 8: 2%

Circle:     3%

Overall:   3%

 

***US directly hit by 'cane from 1/3 of these phase 2 MDR formations (5 of 15): Bertha '96 (NC), Bonnie '98 (NC), Jeanne '04 (FL), Ike '08 (TX), Isaac '12 (LA). Compare that 1/3 US H hit rate with only 10% of the average of all of the other phases! So, not only does phase 2 have the 2nd highest MDR TC genesis rate, but also these phase 2 MDR TC's have been particularly impactful on the US.

----------------------------------------------------------------

 

Non-MDR TC daily genesis rate 6/1-11/30

 

Phase 1: 5%

Phase 2: 8% (Dennis '05 from here)

Phase 3: 5%

Phase 4: 7% (Wilma from here)

Phase 5: 4%

Phase 6: 4% (Charley from here)

Phase 7: 4%

Phase 8: 3%

Circle:      6% (Katrina, Rita, Opal from here)

Overall:   5%

---------------------------------------------------------------

 

Total daily genesis rate 6/1-11/30: tells me that the favorable phases of 2-4 are about twice as likely to yield a TC on any one day as the unfavorable phases of 7 and 8, all other things being equal. Also, keep in mind that these are chances for genesis per day. So, say, there are about six days' worth of phase 2-4 upcoming. There'd be about a 50-50 shot at genesis then vs. only about a 30% chance should phases 7-8 be projected for those same six days.

 

Phase 1: 8%

Phase 2: 13%

Phase 3: 10%

Phase 4: 11%

Phase 5: 7%

Phase 6: 7%

Phase 7: 5%

Phase 8: 6%

Circle:     9%

Overall:   9%

------------------------------------------------------------------------------------

 

Total # of days in each phase 6/1-11/30 1995-2012:

 

Phase 1: 347

Phase 2: 346

Phase 3: 192

Phase 4: 219

Phase 5: 271

Phase 6: 206

Phase 7: 123

Phase 8: 213

Circle:   1,377

Overall: 3,294

[Superstorm93]

I'll have a "State of The Basin" post sometime soon since we do have a Lame Lemon, but I just wanted to pass this along. 
 
I definitely think it's worth the time of whatever stragglers make it into this thread. 
 
Good breakdown from Mr. Levi Cowan

 

 

The gulf is a wreck. We won't get anything significant out of it. Potential depressions or weak TS's without significant prospects drive me nuts because it's hard to actually get the forecast right. The weather for most would be exactly the same anyway.

 

 

1. A SURFACE TROUGH PRODUCING AN AREA OF CLOUDINESS AND THUNDERSTORMS
OVER MUCH OF THE YUCATAN PENINSULA AND THE EASTERN BAY OF CAMPECHE
IS EXPECTED TO MOVE INTO THE SOUTHERN AND WESTERN GULF OF MEXICO
DURING THE NEXT SEVERAL DAYS. UPPER-LEVEL WINDS ARE CURRENTLY
UNFAVORABLE FOR DEVELOPMENT...BUT THEY COULD BECOME MARGINALLY
CONDUCIVE OVER THE NEXT FEW DAYS AS THE DISTURBANCE MOVES
WEST-NORTHWESTWARD TO NORTHWESTWARD AT ABOUT 10 MPH. THIS SYSTEM
HAS A LOW CHANCE...NEAR 10 PERCENT...OF BECOMING A TROPICAL
CYCLONE DURING THE NEXT 48 HOURS.

 

 

 

[Srain]

^^

Wave axis has crossed the Yucatan and is entering the Bay of Campeche. I'd keep an eye on the Western Gulf. While this will not be a wind threat, the retrograding trough and a washed out boundary near Tampico lends credence to what the NCEP ensembles have been suggesting. Keep an eye on a couple of short waves dropping SW to the W of the U/L over the Mississippi Valley. The BoC disturbance appears to be the main feature to monitor for any weak tropical development late this week into early next week.

[Scott747]

0z GFS has the makings of being a teaser.

[Mike.Ventrice]

Mike,

 I decided to check the MDR vs. non-MDR distribution for TC genesis days for 1995-2012 for 6/1-11/30 for each MJO phase and I divided by the # of days in each respective MJO phase. My def. of the MDR is the area at or below 20 N and at or E of 60 W. The phase with the lowest % of days with MDR genesis was phase 7, which had only a 1% daily MDR genesis rate. Then it rose gradually from phase 7 to phase 3, where it maxed at a 5% daily MDR genesis rate. Subsequently, it gradually fell back to the 1% daily MDR genesis rate of phase 7. The circle of death as well as overall % were both at 3%.

 

MDR TC daily genesis rate 6/1-11/30

 

Phase 1: 3% (Ivan from here)

Phase 2: 4% ***

Phase 3: 5% (Fran from here)

Phase 4: 4%

Phase 5: 3%

Phase 6: 3%

Phase 7: 1%

Phase 8: 2%

Circle:     3%

Overall:   3%

 

***US directly hit by 'cane from 1/3 of these phase 2 MDR formations (5 of 15): Bertha '96 (NC), Bonnie '98 (NC), Jeanne '04 (FL), Ike '08 (TX), Isaac '12 (LA). Compare that 1/3 US H hit rate with only 10% of the average of all of the other phases! So, not only does phase 2 have the 2nd highest MDR TC genesis rate, but also these phase 2 MDR TC's have been particularly impactful on the US.

----------------------------------------------------------------

 

Non-MDR TC daily genesis rate 6/1-11/30

 

Phase 1: 5%

Phase 2: 8% (Dennis '05 from here)

Phase 3: 5%

Phase 4: 7% (Wilma from here)

Phase 5: 4%

Phase 6: 4% (Charley from here)

Phase 7: 4%

Phase 8: 3%

Circle:      6% (Katrina, Rita, Opal from here)

Overall:   5%

---------------------------------------------------------------

 

Total daily genesis rate 6/1-11/30: tells me that the favorable phases of 2-4 are about twice as likely to yield a TC on any one day as the unfavorable phases of 7 and 8, all other things being equal. Also, keep in mind that these are chances for genesis per day. So, say, there are about six day's worth of phase 2-4 upcoming, there'd be about a 50-50 shot at genesis then vs. only about a 30% chance should phases 7-8 be projected for those same six days.

 

Phase 1: 8%

Phase 2: 13%

Phase 3: 10%

Phase 4: 11%

Phase 5: 7%

Phase 6: 7%

Phase 7: 5%

Phase 8: 6%

Circle:     9%

Overall:   9%

------------------------------------------------------------------------------------

 

Total # of days in each phase 6/1-11/30 1995-2012:

 

Phase 1: 347

Phase 2: 346

Phase 3: 192

Phase 4: 219

Phase 5: 271

Phase 6: 206

Phase 7: 123

Phase 8: 213

Circle:   1,377

Overall: 3,294

 

Thanks for this- very similar to my thoughts.. Are you looking depressions as a named TC or tropical storm strength? I believe in my previous study I was using depressions.

phases 7-8 I found to be the climotological mins: http://www.atmos.albany.edu/student/ventrice/documents/publications/Ventrice%20et%20al.%202011.pdf

 

After learning more about the standard RMM index, I have found that it does struggle during summer time accurately Identifying MJO events. It'd be interesting if you could replicate your results using the VPM index that I created (for a rainy day of course): http://www.atmos.albany.edu/student/ventrice/documents/VPM_dates.txt

[Mike.Ventrice]

^^

Wave axis has crossed the Yucatan and is entering the Bay of Campeche. I'd keep an eye on the Western Gulf. While this will not be a wind threat, the retrograding trough and a washed out boundary near Tampico lends credence to what the NCEP ensembles have been suggesting. Keep an eye on a couple of short waves dropping SW to the W of the U/L over the Mississippi Valley. The BoC disturbance appears to be the main feature to monitor for any weak tropical development late this week into early next week.

 

My 850 hPa relative vorticity and precipitation maps showing a nice bullseye of low-level rel. vorticity tracking into the GOM and landfalling over eastern TX, western LA on July 7.. http://www.atmos.albany.edu/student/ventrice/real_time/maps/u850/total/atlantic.html

 

Upper-level westerly shear looks to be strong though but maybe could help in the genesis process...

 

Euro op still keying on a disturbance that looks to be seeded from extratropical processes to track south of Florida and possibly spin up in the GOM on 7/10... 

 

Update: GFS suggesting a possible MDR storm in association to a developing easterly wave... I may be wishcasting but it is occurring in a time where MJO is pushing back over IO.

[Ed Lizard]

Not a 0Z GEFS individual member at/below 1004 mb, but some spread in the means.

[Srain]

06Z:

[GaWx]

Thanks for this- very similar to my thoughts.. Are you looking depressions as a named TC or tropical storm strength? I believe in my previous study I was using depressions.

phases 7-8 I found to be the climotological mins: http://www.atmos.albany.edu/student/ventrice/documents/publications/Ventrice%20et%20al.%202011.pdf

 

After learning more about the standard RMM index, I have found that it does struggle during summer time accurately Identifying MJO events. It'd be interesting if you could replicate your results using the VPM index that I created (for a rainy day of course): http://www.atmos.albany.edu/student/ventrice/documents/VPM_dates.txt

 

Mike,

 You're welcome. TD's that didn't become NS's were used whenever they were on the lists I used. These amounted to 26 TD's. So, essentially 260 NS's and 26 TD's were used. The inclusion or exclusion of these 26 TD's likely would have relatively little impact, regardless. Sometimes, the list I used excluded TD's.