Atlantic Tropical Action 2013

[icebreaker5221]

Well maybe something forming here in the next week or two.

 

Superstorm posted this yesterday on the previous page.  But yes, encouraging signs, particularly with respect to the MJO. 

 

 

I would strongly disagree with that statement.  Sandy required several very anomolous events in close proximity to one another in order to produce what it did.  In fact, the 500mb pattern we saw (let alone the anomolous nature of it) can only occur at the mid latitudes as we get later into the fall.  The 1938 hurricane, while stronger than Sandy, is less anomolous than Sandy with respect to getting certain weather features to align for a potential repeat.   

 

Both storms were similar in some ways, however. 1938 wound up taking a left hook in northern New England and southern Quebec as it merged with a large cut off over the Apps. 

 

Both Sandy and 1938 exhibited strong fujiwhara with a deep/anomalous cut off to the west. 

 

I would guess the return time on both storms is probably similar. 

 

I think we have a better chance of seeing another Sandy in the next 50 years then another 38. 38 was the ultimate aligning of events.

 

Not sure you can assign a return period with a sample size of one, but I would agree with CTRain and expect an equivalent storm to either would have roughly equal returns. 

 

There was a study that determined Sandy to be something like a 1 in 700 year event, but that was based upon the landfall location, intensity and trajectory, not the phasing of the mid-latitude jet with a tropical system that took place.  Something similar to Sandy that moves more parallel to the coast rather than perpendicular to the coast is I'm sure more frequent than a 1 in 700 year event. 

 

 

I could be misinterpreting, but I think the Euro ensembles use both perturbed initial conditions *and* physics models, from reading link above.

 

 

The ECMWF ensemble uses 4 different methods to perturb their members:

- Singular vector perturbations, which are optimized for the fastest-growing linear modes through the first 48 h

- An ensemble of data assimilations, which assimilates perturbed observations to better account for observation uncertainty

- A stochastic kinetic energy backscatter scheme (SKEB), which accounts for the uncertainty in sub-gridscale processes

- And lastly, a stochastically perturbed parametrization tendencies scheme (SPPT), which accounts for uncertainty in the parameterized processes.  This is probably the one you were talking about, but it's not the same as perturbing the physics.  They're actually only perturbing the tendencies (u, v, T, q, etc) that come out of the physics package without changing the physics itself.  

[Roger Smith]

Regardless of subjective opinions voiced earlier, the fact of the matter is that tropical storm activity is both more frequent and more intense when the Moon is close to its northern declination maximum. My research data show almost twice as many active storms, landfalls and intense hurricanes in the 25% of time between the Moon's northward transit (which in statistical terms is often close to full moon in autumn) and northern declination maximum. Hazel (1954), Audrey (1957), Frederic (1979), Hugo (1989), Andrew (1992), Isobel (2003), Charley (2004) and Katrina, Rita, Wilma in 2005 all made landfalls very close to the northern declination max. Sandy was heading for the NJ coast at full moon three days before a northern declination max last year. This is not meant to be a complete report, I just recalled these facts and checked them today against research notes. Even the 1823 storm was close to this northern declination max (it occurred on Aug 31, 1823 with a new moon on Sept 4).

 

Take note that August northern max events will fall generally 4-5 days before new moon then by September about halfway from full to new moon, by mid-October about 5-6 days after full moon, by late October 3-4 days after full moon. Camille (1969) does not follow the pattern above as it formed closer to an August new moon and had its most active phase a few days before the southern declination maximum. I don't claim that this research perspective covers all cases, just that it shows a statistically significant peak.

 

 I've noted in my research that there's also a weak secondary effect in the seven days leading up to the southern declination max and I'm looking for some data sets to examine the situation for other theatres in both northern and southern hemisphere to see whether the same processes can be observed there. Now, to balance this, it needs to be said that the data also support other possible external energy sources unrelated to the Moon and the scattering of all data indicates that the Moon's declination cannot be the only driver of the cycle, so please be clear on what I actually believe here, which is that lunar declination is an important but not exclusive source of energy for storm formation. The same can be said for land-based deep cyclonic storms in winter and spring. My working hypothesis is that storm formation in the deep tropics is likely unrelated to lunar energy cycles but storm activity in the transitional stages from NW or W moving to N or NE moving intense phase in the lower mid-latitudes is more governed by lunar energy and begins to converge on a model of similar cause and effect with ordinary lows, hence the growing chance of hybrid activity in the range of 30-40 deg latitude where the two separate types of storm systems begin to interact.

 

My comments about October 20-23 factor in that climatology then favours a high chance of early hybridization and that by mid-October the land-based cyclonic storms are more likely to be forming in closer proximity to any tropical systems that may pass 35 N. Comments about this forecast being spinnable are unfair as there is no comment available to support the charge that I am planning to "spin" any sort of result. Clearly I am talking about a window of opportunity for an intense storm and if there's something more pedestrian in the same general region I will say in advance that my minimum criteria for validation of this potential would be wind gusts in excess of 80 mph and a central pressure below 950 mb, so that will create plenty of spin anyway.

 

As to my forecast of mid-September being a very safe bet, probably true. I wasn't trying to pull a fast one there, just saying that I am expecting this quiet interval to last a while longer and then fade out in mid-September. I don't imagine anyone would be too surprised by this although an even safer bet would be to say plenty of action by first ten days of September and that I'm not too sure would verify but it's not the main focus of my research-driven comments.

[phil882]

The enhanced low-level flow from the MJO wave spectra is coming... the recent activity in the WPAC and CPAC, and future activity in the EPAC should lead to a relatively active September in the Atlantic basin.

 

http://www.atmos.albany.edu/student/macritch/showhoriz.php

[Drz1111]

Do we have a "quack" tag?  We need a "quack" tag.

Regardless of subjective opinions voiced earlier, the fact of the matter is that tropical storm activity is both more frequent and more intense when the Moon is close to its northern declination maximum. My research data show almost twice as many active storms, landfalls and intense hurricanes in the 25% of time between the Moon's northward transit (which in statistical terms is often close to full moon in autumn) and northern declination maximum. Hazel (1954), Audrey (1957), Frederic (1979), Hugo (1989), Andrew (1992), Isobel (2003), Charley (2004) and Katrina, Rita, Wilma in 2005 all made landfalls very close to the northern declination max. Sandy was heading for the NJ coast at full moon three days before a northern declination max last year. This is not meant to be a complete report, I just recalled these facts and checked them today against research notes. Even the 1823 storm was close to this northern declination max (it occurred on Aug 31, 1823 with a new moon on Sept 4).

 

Take note that August northern max events will fall generally 4-5 days before new moon then by September about halfway from full to new moon, by mid-October about 5-6 days after full moon, by late October 3-4 days after full moon. Camille (1969) does not follow the pattern above as it formed closer to an August new moon and had its most active phase a few days before the southern declination maximum. I don't claim that this research perspective covers all cases, just that it shows a statistically significant peak.

 

 I've noted in my research that there's also a weak secondary effect in the seven days leading up to the southern declination max and I'm looking for some data sets to examine the situation for other theatres in both northern and southern hemisphere to see whether the same processes can be observed there. Now, to balance this, it needs to be said that the data also support other possible external energy sources unrelated to the Moon and the scattering of all data indicates that the Moon's declination cannot be the only driver of the cycle, so please be clear on what I actually believe here, which is that lunar declination is an important but not exclusive source of energy for storm formation. The same can be said for land-based deep cyclonic storms in winter and spring. My working hypothesis is that storm formation in the deep tropics is likely unrelated to lunar energy cycles but storm activity in the transitional stages from NW or W moving to N or NE moving intense phase in the lower mid-latitudes is more governed by lunar energy and begins to converge on a model of similar cause and effect with ordinary lows, hence the growing chance of hybrid activity in the range of 30-40 deg latitude where the two separate types of storm systems begin to interact.

 

My comments about October 20-23 factor in that climatology then favours a high chance of early hybridization and that by mid-October the land-based cyclonic storms are more likely to be forming in closer proximity to any tropical systems that may pass 35 N. Comments about this forecast being spinnable are unfair as there is no comment available to support the charge that I am planning to "spin" any sort of result. Clearly I am talking about a window of opportunity for an intense storm and if there's something more pedestrian in the same general region I will say in advance that my minimum criteria for validation of this potential would be wind gusts in excess of 80 mph and a central pressure below 950 mb, so that will create plenty of spin anyway.

 

As to my forecast of mid-September being a very safe bet, probably true. I wasn't trying to pull a fast one there, just saying that I am expecting this quiet interval to last a while longer and then fade out in mid-September. I don't imagine anyone would be too surprised by this although an even safer bet would be to say plenty of action by first ten days of September and that I'm not too sure would verify but it's not the main focus of my research-driven comments.

[Ed Lizard]

About the Moon, and landfalls near a maxima. 

 

Since most of the examples sited were long trackers, they developed prior to any lunar maxes, and in the case of Isabel, at least (just going by memory) reached maximum intensity several days before landfall, and wasn't near max intensity at landfall.

 

I can see a scientific basis for saying Moon phase will have an affect on storm surge, by what manner Moon phase affects storm formation or intensity, if there is a scientific basis, or at least a theory, I'd like to know it.

 

BTW, if you create a graph of all hurricane landfalls by Moon phase, or hurricanes at Cat 3 or above intensity by Moon phase, or genesis by Moon phase, well, if I can see it pictorially, even if I can't understand it, I might be more prone to accept it.

[LakeEffectKing]

^{This is all I need to see.....}

 

[Ed Lizard]

From 31 July...

The CFSv2 weeklies during the mid to late August time frame are suggesting a bit of an uptick in precip across the Caribbean and Gulf. I suspect this fit with the arrival of the MJO/Kelvin Wave that others have alluded to heading E across the Pacific.

 

 

I don't have complete faith in the CFS weeklies, and the ESRL reforecast GEFS anomalies from 0Z seem to suggest the above normal precipitation will be mainly confined to the Pacific for next two weeks, with no clear signal at all for a TC near the Atlantic side of North America. (BoC gets a little rainy, but nothing one can hang one's hat on)  Euro looks quiet through the end of the month.

 

 

I'm sure things will pick up with the climatological peak approaching, but the GFS shows little of interest through Labor Day, and while the GEFS start developing quite a few storms at the end of the month, almost all look like fish well into the first week of September.

 

September and October will have to average two named storms a week for many of the preseason forecasts I saw to verify.

 

On the plus side, sort of, early September 500 mb heights on the spaghetti look like, well, spaghetti.  East Coast troughs.  Troughs back into Kansas and Missouri.  And a few East Coast ridges.  All hope is not lost.  I just don't see much exciting the rest of August.

 

 

 

Edit to add- not picking on any particular admin, just common late July sentiment seemed to be that things would really start cooking mid-August  Edit 2- I feel stupid, I had a gut of a slow August for the contest, but tempered my monthly pessimism with general board optimism.  1/0/0 or 2/0/0 will win the month...

[Ed Lizard]

Mike's Weather Page ( http://spaghettimodels.com/ ) shows a 30 day European (who knew?) valid September 19th that has what looks like a 929 mb hurricane slamming an unusual place, the Georgia/Florida border.

 

I wonder if I pay for the AmWx models (I used to PayPal donate to RaleighWx models) if a 30 day Euro would be part of the package.

^^  Yes, I realize.

 

This post is only semi-serious.

 

 

But, I think maybe I know where the long lead time AccuWx forecasts come from (from a Facebook response to my posting a Facebook image)  AccuWx forecast for JAX is for 'very windy with heavy rain'.  Or maybe AccuWx uses lunar phases.

 

http://www.accuweather.com/en/us/jacksonville-fl/32202/daily-weather-forecast/347935?day=29

 

Coincidence?

[bluewave]

From 31 July...

 

 

I don't have complete faith in the CFS weeklies, and the ESRL reforecast GEFS anomalies from 0Z seem to suggest the above normal precipitation will be mainly confined to the Pacific for next two weeks, with no clear signal at all for a TC near the Atlantic side of North America. (BoC gets a little rainy, but nothing one can hang one's hat on)  Euro looks quiet through the end of the month.

 

 

I'm sure things will pick up with the climatological peak approaching, but the GFS shows little of interest through Labor Day, and while the GEFS start developing quite a few storms at the end of the month, almost all look like fish well into the first week of September.

 

September and October will have to average two named storms a week for many of the preseason forecasts I saw to verify.

 

On the plus side, sort of, early September 500 mb heights on the spaghetti look like, well, spaghetti.  East Coast troughs.  Troughs back into Kansas and Missouri.  And a few East Coast ridges.  All hope is not lost.  I just don't see much exciting the rest of August.

 

 

 

Edit to add- not picking on any particular admin, just common late July sentiment seemed to be that things would really start cooking mid-August  Edit 2- I feel stupid, I had a gut of a slow August for the contest, but tempered my monthly pessimism with general board optimism.  1/0/0 or 2/0/0 will win the month...

 

It will be interesting to see when we finally see our first hurricane of the season this year. Since we should have to wait until

September, I put together the latest first hurricane dates of the satellite era that I could find.

So the record is about a month later than the August 10th average date for the first hurricane.

 

2002....9-11

1984....9-10

2001....9-09

1967....9-03

1988....9-02

[icebreaker5221]

I don't have complete faith in the CFS weeklies, and the ESRL reforecast GEFS anomalies from 0Z seem to suggest the above normal precipitation will be mainly confined to the Pacific for next two weeks, with no clear signal at all for a TC near the Atlantic side of North America. (BoC gets a little rainy, but nothing one can hang one's hat on)  Euro looks quiet through the end of the month.

 

 

Yes, but now it's not just the CFS.  Every major ensemble now shows increasing MJO activity to varying extents. 

 

I really wouldn't put too much stock in GEFS spaghetti plots for easterly waves beyond 7 days. 

[Roger Smith]

TBH, the connection between weather and lunar declination is so self-evident that I am often quite surprised that other people interested in the weather don't easily see it. I suppose very few of those people are also interested in astronomy. Perhaps there is a fear of being labelled an astrologer, but my perspective is simply that of a physical scientist and I don't have any interest in astrology. I will try to work up a graph of major hurricane positions at this declination maximum but it will take a few days as I am rather bogged down with another project that needs to be finished soon.

 

I don't think I'm a "quack" and I find that rather ironic at this stage of the failing political stunt known as climate change.

[Ed Lizard]

About the Moon, and landfalls near a maxima. 

 

Since most of the examples sited were long trackers, they developed prior to any lunar maxes, and in the case of Isabel, at least (just going by memory) reached maximum intensity several days before landfall, and wasn't near max intensity at landfall.

 

I can see a scientific basis for saying Moon phase will have an affect on storm surge, by what manner Moon phase affects storm formation or intensity, if there is a scientific basis, or at least a theory, I'd like to know it.

 

BTW, if you create a graph of all hurricane landfalls by Moon phase, or hurricanes at Cat 3 or above intensity by Moon phase, or genesis by Moon phase, well, if I can see it pictorially, even if I can't understand it, I might be more prone to accept it.

 

 

I don't think I was nailing anybody to any crosses, or even discussing climate change.  I just asked either for a theory about why lunar cycles would affect tropical cyclone activity, and graphs that showed a correlation between something significant about TCs (Cat 3 or above TCs in moon phase quadrants or octants, or hurricane landfalls as a function of moon phase, or maybe RI episodes as a function of lunar phase).

 

I don't think that was all that terrible to ask for.  I'm more open to the idea that our Moon, via some kind of gravitational affect, can affect sensible weather over, say, the alignment of Jupiter's moons producing an energy field that enhances tornado outbreaks in Minnesota or what not.

[Mallow]

TBH, the connection between weather and lunar declination is so self-evident that I am often quite surprised that other people interested in the weather don't easily see it. I suppose very few of those people are also interested in astronomy. Perhaps there is a fear of being labelled an astrologer, but my perspective is simply that of a physical scientist and I don't have any interest in astrology. I will try to work up a graph of major hurricane positions at this declination maximum but it will take a few days as I am rather bogged down with another project that needs to be finished soon.

 

I don't think I'm a "quack" and I find that rather ironic at this stage of the failing political stunt known as climate change.

 

You can't just show "a graph of major hurricane positions at this declination maximum" and expect that to settle it. You have to show that there is a statistically significant difference between the "graph" and the one during non-declination-maximums. And then you have to explain in a consistent manner the physical connection between the lunar declension and storm locations. Otherwise, you're not doing science. And yes, until you do that, I'm going to call you a "quack" for espousing such a theory. Why? Because I have never seen a reasonable physical mechanism to explain any such connection, and any time I ask for one, the answer is usually vague and NOT useful, or just plain silly.

 

And re: your statement on climate change... really dude? Take it to the CC forum, please.

[Srain]

 

 

And re: your statement on climate change... really dude? Take it to the CC forum, please.

[andyhb]

You can't just show "a graph of major hurricane positions at this declination maximum" and expect that to settle it. You have to show that there is a statistically significant difference between the "graph" and the one during non-declination-maximums. And then you have to explain in a consistent manner the physical connection between the lunar declension and storm locations. Otherwise, you're not doing science. And yes, until you do that, I'm going to call you a "quack" for espousing such a theory. Why? Because I have never seen a reasonable physical mechanism to explain any such connection, and any time I ask for one, the answer is usually vague and NOT useful, or just plain silly.

 

And re: your statement on climate change... really dude? Take it to the CC forum, please.

 

Or he just doesn't give an answer, as with some of the other theories he has devised.

[bluewave]

The 12z Euro run is more of the same with all the waves coming off the African Coast hitting 

the wall over the Tropical Atlantic.

 

 

[andyhb]

The 12z Euro run is more of the same with all the waves coming off the African Coast hitting 

the wall over the Tropical Atlantic.

ts_al_tat_THDV.gif

 

I'm going to need a new record player stylus after this year...

[GaWx]

 Similar to what I did for July and then August, I just did an analysis of the number of TC geneses since 1995 in September for each MJO phase and came up with these overall % chances for genesis per day:

 

Phase 1: 24% (Ivan of 2004)

Phase 2: 17% (Ike of 2008 and Jeanne of 2004)

Phase 3: 6%

Phase 4: 16%

Phase 5: 8% (Humberto of 2007)

Phase 6: 6%

Phase 7: 14%

Phase 8: 15%

 

In Circle: 21%

(Rita of 2005, Isabel of 2003, Lili of 2002, Floyd of 1999, Georges of 1998, Opal of 1995)

 

ALL        16%

 

- So, phase 1 has had the highest daily rate of genesis in Sep. since 1995 (24% vs. only 14% in Aug.). So, with just three days of phase 1, genesis is actually more likely than no genesis.

- Being within the circle in Sep. is actually more conducive than average with 21% (vs. 16% overall) unlike in August when it was only 10%/lower than Aug. average.

- Surprisingly, phase 3 had only a 6% daily genesis rate, tied for the bottom in Sep. Compare this rate to a whopping 38% in August, the top rate by a wide margin!

- All but one of the CONUS H hits and all of the major CONUS H hits from Sep. geneses since 1995 were from geneses during phase 1, phase 2 , or within the circle.

 

 

 # Days in each phase in Sep vs. Aug. since 1995:

 

Phase 1: 29 vs. 63

Phase 2: 60 vs. 88

Phase 3: 35 vs. 32

Phase 4: 63 vs. 15

Phase 5: 79 vs. 34

Phase 6: 18 vs. 36

Phase 7:   7 vs. 14

Phase 8: 26 vs. 22

In Circle: 223 vs. 254

 

- Phases 1, 2, 6, and 7 many fewer days in Sep. vs. Aug.

- Phases 4 and 5 many more days in Sep. vs. Aug.

[HurricaneJosh]

Very well said

 

Thanks.  

 

The weather is in hibernation mode. Low tornado year, tropics sleeping, not very hot in the east. A tropical system would be appreciated. 

Yeah.  

 

I think we have a better chance of seeing another Sandy in the next 50 years then another 38. 38 was the ultimate aligning of events.

 

Agreed.

 

I would strongly disagree with that statement.  Sandy required several very anomolous events in close proximity to one another in order to produce what it did.  In fact, the 500mb pattern we saw (let alone the anomolous nature of it) can only occur at the mid latitudes as we get later into the fall.  The 1938 hurricane, while stronger than Sandy, is less anomolous than Sandy with respect to getting certain weather features to align for a potential repeat.   

 

Disagreed.  Sandy was big-- 1938 was nuclear.  Sandy feels bigger because it's fresher and more recent.  But I hesitate to even post this, because I don't want to get into an endless back-and-forth about this topic-- we already did it in 20 other threads.  We just won't agree on it.

[Coach McGuirk]

I'm okay that the Atlantic Hurricane season is a snooze this year. 

[HurricaneJosh]

P.S.  The 1938 hurricane was a Cat 5 E of the Bahamas that shot straight, due N at spectacular speeds to reach Long Island as a very strong Cat 3.  That is mega-anomalous.  The entire basin goes years and years without even seeing one Cat 5 (even in the deep tropics), let alone one shooting toward the USA's biggest population centers at top speed.

[Sunny and Warm]

Disagreed.  Sandy was big-- 1938 was nuclear.  Sandy feels bigger because it's fresher and more recent.  But I hesitate to even post this, because I don't want to get into an endless back-and-forth about this topic-- we already did it in 20 other threads.  We just won't agree on it.

 

Josh,

 

I believe you misconstrued today's discussion.  We weren't talking about comparative strength, but which storm would have a longer return period.

[Coach McGuirk]

P.S.  The 1938 hurricane was a Cat 5 E of the Bahamas that shot straight, due N at spectacular speeds to reach Long Island as a very strong Cat 3.  That is mega-anomalous.  The entire basin goes years and years without even seeing one Cat 5 (even in the deep tropics), let alone one shooting toward the USA's biggest population centers at top speed.

Like with a lot of stuff that happened in the past, it would be a non stop story in the media today.  Then they would have to say, was it global warming? to extend the 24/7 media coverage. 

[Sunny and Warm]

P.S.  The 1938 hurricane was a Cat 5 E of the Bahamas that shot straight, due N at spectacular speeds to reach Long Island as a very strong Cat 3.  That is mega-anomalous.  The entire basin goes years and years without even seeing one Cat 5 (even in the deep tropics), let alone one shooting toward the USA's biggest population centers at top speed.

Steve probably had it right.  The return periods are roughly the same(250-500 years), and none of us are likely to be around to see a repeat of either.  Let's just let it go.

[HurricaneJosh]

Josh,

 

I believe you misconstrued today's discussion.  We weren't talking about comparative strength, but which storm would have a longer return period.

 

Nope, I understood that we were talking return period.  There's no question or debate Re: which was stronger.

[PSUBlizzicane2007]

Steve probably had it right. The return periods are roughly the same(250-500 years), and none of us are likely to be around to see a repeat of either. Let's just let it go.

The fun part about that is that although we likely won't see a repeat of those in particular, there are many other 250-500 year events that we will get to see...you just have to remember that.

[Drz1111]

P.S.  The 1938 hurricane was a Cat 5 E of the Bahamas that shot straight, due N at spectacular speeds to reach Long Island as a very strong Cat 3.  That is mega-anomalous.  The entire basin goes years and years without even seeing one Cat 5 (even in the deep tropics), let alone one shooting toward the USA's biggest population centers at top speed.

Yeah, but '38 ended up missing most of them. Oh no, Providence!

[mob1]

Yeah, but '38 ended up missing most of them. Oh no, Providence!

What!?!? That hurricane hit one of the most populated areas in the US, there's more to the east coast than NYC.

[andyhb]

Not this discussion again.

[GaWx]

 IF the intensity of the MJO being forecasted by the ensemble mean of both the GFS and the Euro for late August/early Sep. in phases 1/2 were to closely verify, it would be the strongest in those phases for that time of year since 1979, which was the only other year since MJO records started with a comparably strong MJO in phases 1/2 during late August through mid Sep.:

 

http://www.bom.gov.au/climate/mjo/graphics/archive/MJO_phase/197907.phase.90days.gif

 

 What happened during that period of 1979? Hurricanes David and Frederic formed. These became cat.4-5's at their strongest and both hit the US as a H. Since 1995, phases 1 and 2 have produced during Aug. and Sep. 7 of the 12 US H hitters and 3 of the 5 US major H hitters of those forming when the MJO wasn't within the circle. Translation: imo there's significantly elevated potential (vs. climo) for a genesis during the last few days of this month into early Sep. that would later result in a significant and quite possibly major US H hit. I've already forecasted a major H hit on the US this season based on a combo of cool August analogs (though it isn't turning out as cool as earlier expected) and neutral negative to weak Nina ENSO. Whereas I'm not outrightly predicting this major US hitter to form in late Aug. to early Sep., the forecasted MJO seems to suggest that this period may very well be the time that it occurs. We'll see.