Atlantic Tropical Action 2012 - Part II

[Superstorm93]

Definitely thinking that we transition from the constant CV fish storms into some more interesting WCARB/GOM systems as September winds down and October kicks up. Many of the "secret" ensembles have been showing pretty much what Srain mentioned, with "sleeper waves" making it much farther west than they have previously been able to over the past two months and finally developing in the more interesting parts of the basin.

October could produce a decent number of storms since we are in more of a "Modoki" Nino more than a classic one.

FNMOC 5-10 day Gen probs:

Just for kicks...

Current MJO forecast would more definitely support some increased October activity.

(Changed due to heckling)

[am19psu]

EWP is pretty useless fwiw

[wxmx]

EWP is pretty useless fwiw

The CLP of MJO

[HurricaneJosh]

I demand good news here. Post some... or else.

Just sayin'.

[Isopycnic]

I demand good news here. Post some... or else.

Just sayin'.

There's always the good chance of a late season slopgyre in the GOM to look forward to.

[Ed Lizard]

Does FNMOC mean NoGaps ensembles?

I am as down with the next guy with a 30% chance of a Carla Cradle storm headed for Tampico, but if that is based on the NoGaps, its glass 3/256th full type optimisim.

[Srain]

There's always the good chance of a late season slopgyre in the GOM to look forward to.

A repost from the Central/Western sub forum I made earlier...likely a 'rainer', but without a doubt the Western Basin is certainly not closed for business, IMO. Humberto (2007) comes to mind..

There still remain some differences between the overnight guidance, but the trend is looking a bit more favorable for rain chance increasing Thursday and extending into the weekend. The GFS and Euro are in some what better agreement regarding the potent short wave and mid latitude trough diving SSE into the Eastern US on Friday. A strong 500mb upper low develops over the Ohio Valley and the attending ‘cold front’ drops S into the Lone Star State. The potent short wave drops S into Texas and slowly meanders E providing lift in an unstable air mass with increasing moisture as the Gulf opens with a return flow. The fly in the ointment will be just where this boundary stalls. Another new wrinkle is a potential surface low developing in the Western Gulf that the Canadian, and to a weaker extent, the Euro are ‘sniffing’. The GFS does not show this feature, but does suggest lowering pressures in the Gulf, so that bears watching as well. All in all it does appear that some beneficial rains are ahead and we’ll need to monitor for any potential development in the Gulf. Stalled boundaries and lowering pressures in the Gulf always raise an eyebrow or two when we enter the month of September.

And then Jeff, one of our local Pro Mets follows up with this...

CMC and ECMWF try and develop a broad surface low over the central/western Gulf of Mexico late this weekend/early next week. This low could help drive the front further offshore. CMC drifts the low over the western Gulf while the ECMWF has the feature caught up in the trough and pulled toward Louisiana. GFS does not show this feature as much and is not nearly as well defined as the CMC and ECMWF.

[am19psu]

I'm on board with a rainer, though a record rapid spin up like Humberto seems like something at the top end of potential. Another Debby-like development seems more likely (though perhaps with less of an eastward bent).

[Srain]

I'm on board with a rainer, though a record rapid spin up like Humberto seems like something at the top end of potential. Another Debby-like development seems more likely (though perhaps with less of an eastward bent).

Yep. That is what I'm thinking as well. This frontal boundary is not as strong as the last in our part of the world and I tend to think a meandering low is the more likely solutiom as well.

[Ed Lizard]

I'm totally cool with a NW Gulf poorly organized TC, but assuming the Canadian, as it usually is, is too enthusiastic, and accepting the Euro, my backyard goes pretty dry.

(My philosophy, major hurricanes hitting Florida and winding up on YouTube are good, as are big hurricanes hitting near the Rio Grande, providng YouTube videos, and raining on my yard, and generic 45 knot storms that rain on my yard are cool, but 40 knot storms that don't rain on my yard, there is no point).

Sadly enough, since I can admit I do have the GEM ensembles bookmarked, the op Canadian isn't well supported by most ensemble members.

[turtlehurricane]

We're still getting storm after storm and it's a - MJO, + phase should be nuts.

[GaWx]

We're still getting storm after storm and it's a - MJO, + phase should be nuts.

I feel your statement is bold because I feel that this season is just one of many examples of past seasons going back to the 1970's that shows that the MJO indicator is at most a weak indicator as far as MDR Atlantic tropical activity timing is concerned. I've literally looked at a combo of past activity for each season and overlayed past MJO phases (1-8 as well as within the circle) over about a 35 year period and found very low correlations at best. I'm not anti-MJO by any means. However, using it to try to predict Atlantic activity, especially MDR activity, isn't one of the MJO's best uses imo. Even in the western basin in Oct/Nov, it unfortunately isn't all that much help per these same statistical studies I've done.

Edit: I have posted similar ideas about the weak correlation in the past. I'm not just Monday morning QBing.

[turtlehurricane]

I feel your statement is bold because I feel that this season is just one of many examples of past seasons going back to the 1970's that shows that the MJO indicator is at most a weak indicator as far as MDR Atlantic tropical activity timing is concerned. I've literally looked at a combo of past activity for each season and overlayed past MJO phases (1-8 as well as within the circle) over about a 35 year period and found very low correlations at best.

The MJO is very real from my experience, it's hard to capture it in TC numbers since it's very complex but when you follow the CPC charts it all matches up. A solid + MJO is way more favorable than a solid - MJO like right now.

[Mike.Ventrice]

We're still getting storm after storm and it's a - MJO, + phase should be nuts.

I feel your statement is bold because I feel that this season is just one of many examples of past seasons going back to the 1970's that shows that the MJO indicator is at most a weak indicator as far as MDR Atlantic tropical activity timing is concerned. I've literally looked at a combo of past activity for each season and overlayed past MJO phases (1-8 as well as within the circle) over about a 35 year period and found very low correlations at best. I'm not anti-MJO by any means. However, using it to try to predict Atlantic activity, especially MDR activity, isn't one of the MJO's best uses imo. Even in the western basin in Oct/Nov, it unfortunately isn't all that much help per these same statistical studies I've done.

Edit: I have posted similar ideas about the weak correlation in the past. I'm not just Monday morning QBing.

The MJO is very real from my experience, it's hard to capture it in TC numbers since it's very complex but when you follow the CPC charts it all matches up. A solid + MJO is way more favorable than a solid - MJO like right now.

Folks, tropical cyclones can still occur during unfavorable MJO phases. While predicting tropical cyclone activity at long-range time scales, the MJO can increase or reduce your probability of genesis. It should be noted that most of the hurricanes we've observed this year have occurred well north of 20N, therefore the overall impact of the MJO is reduced. With respect to the current phase of the MJO, the genesis of TD-14 is tied to a strong easterly wave. This easterly wave was initiated by the passage of the convectively active phase of a strong eastward propagating Kelvin wave on Sept. 6-7 over West Africa (which weakened in time in the filtered VP200 field and was superimposed with the convectively suppressed phase of the MJO). Therefore, one distinct feature alone (i.e. the MJO) cannot explain the spectrum of genesis in the main development region. But it can give you a sense of long-range predictability based on probabilistic forecast.

[am19psu]

Folks, tropical cyclones can still occur during unfavorable MJO phases. While predicting tropical cyclone activity at long-range time scales, the MJO can increase or reduce your probability of genesis. It should be noted that most of the hurricanes we've observed this year have occurred well north of 20N, therefore the overall impact of the MJO is reduced. With respect to the current phase of the MJO, the genesis of TD-14 is tied to a strong easterly wave. This easterly wave was initiated by the passage of the convectively active phase of a strong eastward propagating Kelvin wave on Sept. 6-7 over West Africa (which weakened in time in the filtered VP200 field and was superimposed with the convectively suppressed phase of the MJO). Therefore, one distinct feature alone (i.e. the MJO) cannot explain the spectrum of genesis in the main development region. But it can give you a sense of long-range predictability based on probabilistic forecast.

^^^^Everyone should be required to read this

[PSUBlizzicane2007]

Folks, tropical cyclones can still occur during unfavorable MJO phases. While predicting tropical cyclone activity at long-range time scales, the MJO can increase or reduce your probability of genesis. It should be noted that most of the hurricanes we've observed this year have occurred well north of 20N, therefore the overall impact of the MJO is reduced. With respect to the current phase of the MJO, the genesis of TD-14 is tied to a strong easterly wave. This easterly wave was initiated by the passage of the convectively active phase of a strong eastward propagating Kelvin wave on Sept. 6-7 over West Africa (which weakened in time in the filtered VP200 field and was superimposed with the convectively suppressed phase of the MJO). Therefore, one distinct feature alone (i.e. the MJO) cannot explain the spectrum of genesis in the main development region. But it can give you a sense of long-range predictability based on probabilistic forecast.

THANK YOU for this post. I've felt over the past month that some people have placed way too much emphasis on the MJO by saying things like "there won't be any tropical cyclones from x to y date"... despite models showing consistent TC formation.

[bluewave]

Folks, tropical cyclones can still occur during unfavorable MJO phases. While predicting tropical cyclone activity at long-range time scales, the MJO can increase or reduce your probability of genesis. It should be noted that most of the hurricanes we've observed this year have occurred well north of 20N, therefore the overall impact of the MJO is reduced. With respect to the current phase of the MJO, the genesis of TD-14 is tied to a strong easterly wave. This easterly wave was initiated by the passage of the convectively active phase of a strong eastward propagating Kelvin wave on Sept. 6-7 over West Africa (which weakened in time in the filtered VP200 field and was superimposed with the convectively suppressed phase of the MJO). Therefore, one distinct feature alone (i.e. the MJO) cannot explain the spectrum of genesis in the main development region. But it can give you a sense of long-range predictability based on probabilistic forecast.

That's a good point about north of 20N hurricanes. Years with this more northerly development and track have been

fairly common since 2006.

[Ed Lizard]

And 2003!

and 2001

Never seen that before, storms heading Northward after they develop, with some storms even developing North of 20ºN!

[bluewave]

And 2003!

and 2001

Never seen that before, storms heading Northward after they develop, with some storms even developing North of 20ºN!

2003 and 2001 had more major development south of 20N than the recent years I listed.

You had Iris, Isabel, and Fabian strengthening to a good degree while still south of 20N.

But 1997 and 2002 had weaker development action south of 20N between 1995-2005.

[Ed Lizard]

I was trying to point out many years have development from non-tropical origins, and that except for the strong cold ENSO years, where there have been some Southern runners, most storms eventually recurve well short of land. They move a little bit right of mean steering slightly because of beta, and eventually will encounter a weakness in the subtropical ridge and eventually get to the Westerlies.

The only thing vaguely notable I can see is 2011 and 2012 have favored quantity over quality.

[bluewave]

I was trying to point out many years have development from non-tropical origins, and that except for the strong cold ENSO years, where there have been some Southern runners, most storms eventually recurve well short of land. They move a little bit right of mean steering slightly because of beta, and eventually will encounter a weakness in the subtropical ridge and eventually get to the Westerlies.

The only thing vaguely notable I can see is 2011 and 2012 have favored quantity over quality.

Yeah, the ACE was higher last year than this year so far, but both years featured better development

north of 20N. Both years also produced overachieving cat 1 US hurricane landfalls with Irene and

Isaac along with drenching TS impacts of Lee and Debbie.

[Srain]

For what it's worth, the 00Z GFS is suggesting an area of disturbed weather spinning up in the Bay of Campeche near the 18th, +/- a couple of days and morving N to just off the SE TX/SW LA Coast as a TD or weak TS near the 20th before being shunted ENE.

[Mike.Ventrice]

I'd like to share an observation. Myself, as well as others, have been discussing the relative impact of the MJO on Atlantic tropical cyclone activity. We all know the climatological peak of Atlantic TC activity is September. After the genesis of Michael on September 3, we have only observed the genesis of one Atlantic tropical cyclone (as of yesterday, Nadine). Both the GFS and ECMWF forecasts suggest that the Atlantic (besides Nadine) will be quite through September 22. Therefore, it appears that even though we have had some long-lived tropical cyclones in the Atlantic for the first half of September, there really has been relatively low tropical cyclogenesis activity over the Basin, especially when compared to August.

Eight tropical cyclones developed in the month of August, with, so far, only two in September. According to an RMM-type plot that I created yesterday (below) using only MJO-filtered VP200 anomalies to construct the PCs, the active phase of the MJO was mostly in the favorable RMM Phases (8-3) during the month of August. September is characterized by RMM phases 3-5, with counterclockwise rotation into unfavorable TC genesis phases 6-7 in the forecast (linearly extrapolated by padding with zeros at the end of the filtered time-series, therefore dampening to lower amplitudes in the forecast is inevitable). RMM phases 5-7 is when the MJO suppresses convection and tropical cyclone activity over the Atlantic. Further, long-lived Atlantic Hurricanes are generally less likely to occur during this time.

I know we are currently in mid-September, but with the current forecasts from the NWP models indicating a lack of new genesis activity, I think discrediting the overall impact of the suppressed phase of the MJO on TC activity this year would be naiive. On the other side of the spectrum, to think tropical cyclones cannot develop in unfavorable MJO phases is also wrong. While tropical cyclones STILL develop in unfavorable MJO phases, they are less likely to do so. You also won't see trains of tropical cyclones to develop in unfavorable RMM phases. They are much more common when MJO convection initiates over the Indian Basin and propagates northward (phases 1-3).

That being said, we might still expect a future period of increased Atlantic genesis activity towards the end of September through the first half of October.

[am19psu]

I agree with everything, Mike. Super post. That basically described exactly how I use the MJO to forecast in the medium range. The only thing I'll add is that I'm bearish on another active period this season under the assumption that ENSO effects should start to increase as we head into autumn.

[Mike.Ventrice]

I agree with everything, Mike. Super post. That basically described exactly how I use the MJO to forecast in the medium range. The only thing I'll add is that I'm bearish on another active period this season under the assumption that ENSO effects should start to increase as we head into autumn.

This is a main variable in the equation. With respect to the current AMM state, the atmosphere "thinks" its in a La Nina base state, indicated by the low AMM since Spring.

However, the CFS v2 forecast on Kyle MacRitchie's page suggests that the atmosphere will begin to respond to ENSO, such that a strong oceanic Kelvin wave in the Pacific will couple to convection, and accelerate the transition to El Nino (Figure below). I'm not to keen on this solution, but it is a piece of information that shouldn't be ignored.

[Superstorm93]

I am but a n00b when compared to Mr.Ventrice, but our "Nino" (I don't know if you could even call it that any longer) seems to be more west-based or even a "Modoki" one. Some cool anomalies are starting to appear just east of where the basis of the above-normal anoms are located, and this has been the case for the past week or two. My thinking is that the atmosphere continues to stay in a neutral "state of mind" for the remainder of the seaosn, and we do not see an abrupt end to the season this time around as one would usually expect with a warm ENSO.

With that being said, I could see how the Atlantic manages to squeeze out another 3-5 cyclones by the time November rolls around (yielding a total of 17-19). I'd also speculate that we see maybe one or two more AEW developments before the wave train winds down later this month. As we head into the October period, that is when I think we see the typical Caribbean/Gulf development that could make the last gasps of this season a bit interesting. Just basing my thoughts on what I've experienced over the past 10 years, and any input from the red-taggers would be appreciated.

The Experimental ensembles GEN probs illustrate my thinking pretty well at this time (but much earlier than I thought)

NCEP Ensembles 5-10 day GEN probs:

CMC Ensembles 0-5 day GEN probs:

[andyhb]

Learned quite a bit in these last couple pages, thanks guys.

[Ed Lizard]

Modelling question - Canadian doesn't seem quite as TC happy as past years (not a scientific observation, just a feeling). I thought I saw something about some tweak to make it less TC happy.

I asusme the Canadian ensembles are run at lower resolution than the operational? Does this cause more grid scale feedback/spurious TCs than the operational?

And why don't the GFS ensembles have the same issue with lower resolution and spurious TCs? If there is an issue, it isn't as obvious as with the Canadian emsembles.

[turtlehurricane]

300+ undergrads at UW-Madison have to make a hurricane forecast every day for the rest of the season thanks to me

[Derecho!]

Modelling question - Canadian doesn't seem quite as TC happy as past years (not a scientific observation, just a feeling). I thought I saw something about some tweak to make it less TC happy.

I asusme the Canadian ensembles are run at lower resolution than the operational? Does this cause more grid scale feedback/spurious TCs than the operational?

And why don't the GFS ensembles have the same issue with lower resolution and spurious TCs? If there is an issue, it isn't as obvious as with the Canadian emsembles.

The struggle against model boguscanes was a long one (and yes, I believe there was a scientific paper that actually called them that - might want to google it up.) I'm not quite certain how they were removed, but the AVN/MRF was hideous regarding them about a decade ago.