Potential Hermine Impacts/Obs Part 2

[RU848789]

 

Forecasting Hermine’s trajectory was exceptionally challenging.  All things considered, I believe the NHC did a good job considering the degree of challenge involved.

First, no strong block was present to steer Hermine to landfall in the Middle Atlantic region.

Second, an extended period during which steering currents would weaken dramatically, leaving the storm’s track subject to the changing synoptic features occurred. When steering currents are weak, it’s very difficult to be sure where the storm will track with any great degree of precision.

While a trough capture could have brought Hermine to landfall, no such capture was reliably modeled (a few runs in a few cycles featured that possibility, but the bulk of the modeling did not support it).

The largest factor that precluded a clear landfall scenario with a rare WNW, W, WSW trajectory was the absence of Atlantic blocking. The Arctic Oscillation was as follows during the past five days:

September 2: +0.566
September 3: +0.301
September 4: +0.061
September 5: +0.298
September 6: +1.099

Two recent cases of storms taking such an unusual track involved a later February 2010 snowstorm that brought 20.9” snow to New York City and Sandy that made landfall along the New Jersey coast. Both cases featured very strong blocking. Even as wave lengths are shorter in the summer/fall, they exceptional blocking precluded Sandy’s escape out to sea.

February 25-27, 2010 Snowstorm:

From NWS Mount Holly:

An area of low pressure developed off the Carolina coast late Wednesday night February 24th and then strengthened as it tracked northward to near Long Island, New York by Thursday evening. As low pressure aloft deepened over the Mid Atlantic coast Thursday night into Friday, the surface low retrograded and moved westward into northern New Jersey and southern New York.

http://www.weather.gov/phi/02252010wss

http://cstar.cestm.albany.edu/nrow/NROWXII/Hanes.ppt.pptx (See slide 15 for the track)

Arctic Oscillation:
February 22: -4.403 (1st lowest for the date)
February 23: -4.591 (1st lowest for the date)
February 24: -4.211 (1st lowest for the date)
February 25: -3.818 (3rd lowest for the date)
February 26: -3.230 (7th lowest for the date)
February 27: -3.118 (8th lowest for the date)

Sandy: October 29-30, 2012:

http://www.nhc.noaa.gov/data/tcr/AL182012_Sandy.pdf (See p.127 for the track)

Arctic Oscillation:
October 25: -2.912 (1st lowest for the date)
October 26: -2.393 (2nd lowest for the date)
October 27: -2.058 (2nd lowest for the date)
October 28: -1.792 (5th lowest for the date)
October 29: -1.718 (4th lowest for the date)
October 30: -2.075 (2nd lowest for the date)

Don - always love your posts, but I have to disagree a bit with the high marks for the NHC.  Thought they did great from original formation in the Gulf through FL landfall and the ride up the coast to NC, but thought they did pretty poorly after that.  I get that meteorological models, while deterministic, in theory, suffer from propagating errors as one moves forward with each numerical time increment, due to their inherently chaotic nature of the partial differential equations governing the physics and chemistry of the atmosphere, but I think if one looked at 24-hour track forecasts, especially, I'm pretty sure they'd be much more accurate for most tropical systems than they were for Hermine (would be interested to know if that statement is true - I don't have the data).  

For example, the predicted track for Saturday and actual outcome on Saturday into Sunday were pretty far apart, as the track was supposed to be fairly close to the DE/NJ coasts, followed by a stall and a retrograde, all of which would've made for major impacts for DE/NJ coastal sections, especially south of LBI (as I was quite aware, as we came down to Ocean City, NJ Saturday morning for the week and were quite concerned).  The reality was that the system steamed far further east than the NHC thought (if I recall the Euro did nail this outcome) it would, leading to far less impacts from Hermine.  

This was the most important part of the forecast and in my recollection, being that far off for a 24-hour track forecast is pretty unusual.  After that, the NHC has also missed the much more pronounced retrogression by ~200 miles over the last 2 days - again, we're not talking about a 200 mile error 4 days out but 1-2 days out.  Fortunately, the storm was far enough out and weak enough for it not to matter much, but error-wise for a short term forecast, it was pretty significant.  Curious what you think about this part of the forecast.

[RU848789]

 

wetaher shouldn't be so personal.  there will always be two camps in every storm.  someone will be wrong.  that isn't what it's about.  if it were a contest, we would have a section for that (and we have).

i agree with Rjay requiring bacon strips (and anyone else) to provide empirical support when you make claims/assertions.  Otherwise, we are all entitled to our opinion of how a storm will play out.  if you have to constantly remind everyone how right you were, it might be a flaw in your character too.  don't be the kanye west of a weather forum.  i've never seen bluewave start a sentence in the vein of  "like i've been saying for days" and yet he is still probably recognized as the best poster in this sub-forum by many.  substance over ego.

Fair points, thanks.  I don't think I've ever gotten in a confrontation like this on this board, before, so hoping it's not a general character flaw, lol.  I do, though, get a little wound up when people make what I think are outlandish claims without any evidence.  Sorry for cluttering up the board with an argument.  

[Sophisticated Skeptic]

 

He was right in the sense it came a lot further west than everyone thought it would. 

thanks.

the only part that truly shocked me (and turned into a learning lesson) was how quickly Hermine could weaken over relatively warm waters.   It practically went "poof" without even hitting land.  I've never seen that happen before.  Usually it's because of land interaction or much cooler waters.  Plus, with the ship reports, it seemed stronger than it really was.  You learn something new with every storm.

 

[nycsnow]

Windiest it's been with constant drizzle 

[Morris]

 

thanks.

the only part that truly shocked me (and turned into a learning lesson) was how quickly Hermine could weaken over relatively warm waters.   It practically went "poof" without even hitting land.  I've never seen that happen before.  Usually it's because of land interaction or much cooler waters.  Plus, with the ship reports, it seemed stronger than it really was.  You learn something new with every storm.

 

It was modeled well. You were a "bitter clinger."

[Sophisticated Skeptic]

 

It was modeled well. You were a "bitter clinger."

 

Indeed.  ..was so adamant on the Track , but completely threw out the intensity forecasts.  

felt like the models had no grasp on the system intensity wise, since they were so horrible with the track as is...and the whole time forecasting the storm about 10 mb's weaker than it actually was.  For like 3 whole forecast days, they (models) were forecasting Hermine around 10 mb's weaker than it was..   Pressure wise.

[BxWeatherwatcher]

it's quite breezy and some steady light rain falling now. as we are finally getting some liquid outta this storm,it only took 2 days of course but hermine is still around lol..

[USCG RS]

"This is the way the [system] ends, not with a bang, but a whimper."

Sent from my SAMSUNG-SM-N920A using Tapatalk

[BxWeatherwatcher]

 

"This is the way the [system] ends, not with a bang, but a whimper."

Sent from my SAMSUNG-SM-N920A using Tapatalk
 

Well technically it's still going,its raining quite nicely here in the bronx currently.right under a band and our first real rain from this system,a bit late obviously..but its dying a slow death..

[USCG RS]

Well technically it's still going,its raining quite nicely here in the bronx currently.right under a band and our first real rain from this system,a bit late obviously..but its dying a slow death..

She's still trying to get her act together tbh, but I figured it was a fitting quote

Sent from my SAMSUNG-SM-N920A using Tapatalk

[bluewave]

 

 

The Euro was off by more then 500 miles with Hermine's current location.

Performed the worst by far.

Rgem was by far the best as it never waivered, even when every other model did on Saturday. And it had this solution from Friday.

 

The RGEM was just as bad as the other models. Like the Euro it had Hermine deepening similar to the Euro and coming back 

off the Delmarva Sunday with higher impact conditions at coast instead of the wide swing east Saturday. It was also overdone

with the rainfall. The move back west yesterday wasn't that significant since it was over cooler waters and cutoff form 

it's moisture source to the SE. So the closest approach to the coast was while weakening and not bombing out like

the models were showing later last week.

 

\

 

[uncle W]

 

Forecasting Hermine’s trajectory was exceptionally challenging.  All things considered, I believe the NHC did a good job considering the degree of challenge involved.

First, no strong block was present to steer Hermine to landfall in the Middle Atlantic region.

Second, an extended period during which steering currents would weaken dramatically, leaving the storm’s track subject to the changing synoptic features occurred. When steering currents are weak, it’s very difficult to be sure where the storm will track with any great degree of precision.

While a trough capture could have brought Hermine to landfall, no such capture was reliably modeled (a few runs in a few cycles featured that possibility, but the bulk of the modeling did not support it).

The largest factor that precluded a clear landfall scenario with a rare WNW, W, WSW trajectory was the absence of Atlantic blocking. The Arctic Oscillation was as follows during the past five days:

September 2: +0.566
September 3: +0.301
September 4: +0.061
September 5: +0.298
September 6: +1.099

Two recent cases of storms taking such an unusual track involved a later February 2010 snowstorm that brought 20.9” snow to New York City and Sandy that made landfall along the New Jersey coast. Both cases featured very strong blocking. Even as wave lengths are shorter in the summer/fall, they exceptional blocking precluded Sandy’s escape out to sea.

February 25-27, 2010 Snowstorm:

From NWS Mount Holly:

An area of low pressure developed off the Carolina coast late Wednesday night February 24th and then strengthened as it tracked northward to near Long Island, New York by Thursday evening. As low pressure aloft deepened over the Mid Atlantic coast Thursday night into Friday, the surface low retrograded and moved westward into northern New Jersey and southern New York.

http://www.weather.gov/phi/02252010wss

http://cstar.cestm.albany.edu/nrow/NROWXII/Hanes.ppt.pptx (See slide 15 for the track)

Arctic Oscillation:
February 22: -4.403 (1st lowest for the date)
February 23: -4.591 (1st lowest for the date)
February 24: -4.211 (1st lowest for the date)
February 25: -3.818 (3rd lowest for the date)
February 26: -3.230 (7th lowest for the date)
February 27: -3.118 (8th lowest for the date)

Sandy: October 29-30, 2012:

http://www.nhc.noaa.gov/data/tcr/AL182012_Sandy.pdf (See p.127 for the track)

Arctic Oscillation:
October 25: -2.912 (1st lowest for the date)
October 26: -2.393 (2nd lowest for the date)
October 27: -2.058 (2nd lowest for the date)
October 28: -1.792 (5th lowest for the date)
October 29: -1.718 (4th lowest for the date)
October 30: -2.075 (2nd lowest for the date)

I mentioned this before the storm hit on another sight and was shot down by some mets on that board...I figured a high latitude block was needed like Sandy had...I didn't think the storm would meander or loop which it finally did...So the forecast wasn't a complete bust...

[bluewave]

 

I mentioned this before the storm hit on another sight and was shot down by some mets on that board...I figured a high latitude block was needed like Sandy had...I didn't think the storm would meander or loop which it finally did...So the forecast wasn't a complete bust...

We are really lucky that this was an error further east than the models were showing on Saturday and Sunday when it mattered than the other way around.

Could you imagine if all the models showed a harmless OTS track and the NHC had no tropical storm warnings up. You would have had people away

at the shore for their vacations and the NHC would have had to put up a short fused TS warning when the storm turned back to the coast defying the models.

 

I think the last time a system came much closer to the coast with a high impact storm was the January 2000 modeling error. All the models had

a OTS track and the east got a surprise snowstorm instead.

[donsutherland1]

 

Don - always love your posts, but I have to disagree a bit with the high marks for the NHC.  Thought they did great from original formation in the Gulf through FL landfall and the ride up the coast to NC, but thought they did pretty poorly after that.  I get that meteorological models, while deterministic, in theory, suffer from propagating errors as one moves forward with each numerical time increment, due to their inherently chaotic nature of the partial differential equations governing the physics and chemistry of the atmosphere, but I think if one looked at 24-hour track forecasts, especially, I'm pretty sure they'd be much more accurate for most tropical systems than they were for Hermine (would be interested to know if that statement is true - I don't have the data).  

For example, the predicted track for Saturday and actual outcome on Saturday into Sunday were pretty far apart, as the track was supposed to be fairly close to the DE/NJ coasts, followed by a stall and a retrograde, all of which would've made for major impacts for DE/NJ coastal sections, especially south of LBI (as I was quite aware, as we came down to Ocean City, NJ Saturday morning for the week and were quite concerned).  The reality was that the system steamed far further east than the NHC thought (if I recall the Euro did nail this outcome) it would, leading to far less impacts from Hermine.  

This was the most important part of the forecast and in my recollection, being that far off for a 24-hour track forecast is pretty unusual.  After that, the NHC has also missed the much more pronounced retrogression by ~200 miles over the last 2 days - again, we're not talking about a 200 mile error 4 days out but 1-2 days out.  Fortunately, the storm was far enough out and weak enough for it not to matter much, but error-wise for a short term forecast, it was pretty significant.  Curious what you think about this part of the forecast.

You raise valid points. I agree that the NHC had some difficulties, but felt that it did good, not great, considering the level of challenge involved. Track errors were larger than usual, but there were big questions concerning the timing when the steering currents would weaken dramatically and Hermine would meander/drift westward. The models provided little consensus and run-to-run continuity was quite poor. Perhaps I'm being a little too generous, though.

[allgame830]

 

You raise valid points. I agree that the NHC had some difficulties, but felt that it did good, not great, considering the level of challenge involved. Track errors were larger than usual, but there were big questions concerning the timing when the steering currents would weaken dramatically and Hermine would meander/drift westward. The models provided little consensus and run-to-run continuity was quite poor. Perhaps I'm being a little too generous, though.

Honestly, now more than ever we really are at the mercy of these models which seem to have been really poor lately... not sure what the cause of that is... But I really when I started following weather about 15 years ago things seemed be a little different... We had the ETA which now has been replaced by the NAM which honestly is probably our worst model we have out there.  It has scored some wins but very few.  I think the GFS has to have a higher score than the NAM at this point.  

Things to have seemed to have changed and maybe a met or a person with a little more knowledge on that subject can give some insight.  

[donsutherland1]

 

Honestly, now more than ever we really are at the mercy of these models which seem to have been really poor lately... not sure what the cause of that is... But I really when I started following weather about 15 years ago things seemed be a little different... We had the ETA which now has been replaced by the NAM which honestly is probably our worst model we have out there.  It has scored some wins but very few.  I think the GFS has to have a higher score than the NAM at this point.  

Things to have seemed to have changed and maybe a met or a person with a little more knowledge on that subject can give some insight.  

It's possible that some of the issue concerns perceptions from changing public expectations. The public is now looking at longer time frames and at a level of greater detail. Those perceptions may have been shaped by increasingly local/hyperlocal forecasts.

At the same time, there have been issues that have arisen with the modeling e.g., continued resolution increases and changes in various parameters have led to uneven performance changes (improvement overall, but not in all regions) In general, the models are still performing better than those of even a few years ago.

Perhaps because the models have become so good relative where they once were, some forecasters may well have become dependent on the guidance to an extent that their own synoptic skill has eroded. I believe there was a paper that raised that issue in an analysis of a missed forecast on a high-impact winter storm. Nevertheless, in the larger scheme of things, today's forecasts are, on average, improved over where they were in the past.

The National Hurricane Center has made a lot of progress in improving its forecasts. Differences arise on a tropical cyclone to tropical cyclone basis and Hermine won't be among its best forecasts. Below is some information on the National Hurricane Center's performance.

NHC Forecasting Errors:
http://www.nhc.noaa.gov/verification/verify4.shtml

NHC Annnual Verification Reports:
http://www.nhc.noaa.gov/verification/verify3.shtml

[RU848789]

 

You raise valid points. I agree that the NHC had some difficulties, but felt that it did good, not great, considering the level of challenge involved. Track errors were larger than usual, but there were big questions concerning the timing when the steering currents would weaken dramatically and Hermine would meander/drift westward. The models provided little consensus and run-to-run continuity was quite poor. Perhaps I'm being a little too generous, though.

Fair points about the challenging aspects of the forecast, with regard to steering currents, the capture timing, the transition to post-tropical, etc.  Unfortunately, although understandably, the vast majority of people only look at the outcome, not the underlying complexities of the forecast.  

[donsutherland1]

 

I mentioned this before the storm hit on another sight and was shot down by some mets on that board...I figured a high latitude block was needed like Sandy had...I didn't think the storm would meander or loop which it finally did...So the forecast wasn't a complete bust...

IMO, your observation was a good one. Although wave lengths are shorter now, a very strong Atlantic block could have shaped the overall synoptic environment. A weaker one would have had little impact. The effect of Atlantic blocking becomes increasingly significant as we advance deeper into the fall (meteorological fall) on account of the growing wave lengths.

The two cases I provided, one of which was Sandy, illustrated how a large-scale, exceptional blocking regime can impact a storm's track. Even as it occurred earlier in the season than Sandy, the 1938 hurricane's path also was shaped by the combination of a retrograding block and deepening trough that led to its turn to the coast and landfall. The September 1903 hurricane that made New Jersey landfall also occurred during a period of strengthening Atlantic blocking. The cases of such storms are uncommon, but one can't automatically dismiss the role of Atlantic blocking in influencing their overall track. Even as a range of variables is involved, such blocking might well make the difference in a marginal situation.

Finally, I thought that the idea of taking a trimmed model consensus, arrived at by throwing out the outliers, worked out reasonably well. It wasn't perfect, but it avoided the error of assuming Mid-Atlantic landfall and the other error of assuming a harmless escape out to sea.

[Juliancolton]

It's of little consolation for all the forecasting headaches, but Hermine's remnant circulation is reasonably attractive this morning:

 

 

 

[ag3]

 

The RGEM was just as bad as the other models. Like the Euro it had Hermine deepening similar to the Euro and coming back 

off the Delmarva Sunday with higher impact conditions at coast instead of the wide swing east Saturday. It was also overdone

with the rainfall. The move back west yesterday wasn't that significant since it was over cooler waters and cutoff form 

it's moisture source to the SE. So the closest approach to the coast was while weakening and not bombing out like

the models were showing later last week.

 

\

 

 

JFK had .17" and ISP had .50"+ of rain yesterday.

Euro was the worst model BY FAR out of the entire suite. Had the wrong track on every run and way off on precip, even on yesterday's 12z run.

[bluewave]

 

 

JFK had .17" and ISP had .50"+ of rain yesterday.

Euro was the worst model BY FAR out of the entire suite. Had the wrong track on every run and way off on precip, even on yesterday's 12z run.

To the Euro's credit, it got away from the earlier bad forecast when it shifted east 12z Saturday while the RGEM was still riding the strong tucked in solution near the coast.

RGEM 12z Sat forecast

 

[ag3]

 

To the Euro's credit, it got away from the earlier bad forecast when it shifted east 12z Saturday while the RGEM was still riding the strong tucked in solution near the coast.

RGEM 12z Sat forecast

 

 

Euro was 700 miles too east and you are giving it credit? Rgem was 15-20mbs too strong on Saturday but the overall track was acceptable.

[bluewave]

 

 

Euro was 700 miles too east and you are giving it credit? Rgem was 15-20mbs too strong on Saturday but the overall track was acceptable.

The only part of the forecast which mattered was the storm going well east on Saturday missing the strong phase and rapid deepening near the coast.

That's the only reason the earlier guidance was calling for a 2-4 or 3-5 potential surge event at the coast and much stronger winds further west

than verified. A weakening storm coming west too late off shore to matter wasn't a big deal in the big scheme of things.

[Brian5671]

 

 

Euro was 700 miles too east and you are giving it credit? Rgem was 15-20mbs too strong on Saturday but the overall track was acceptable.

being off 15-20 mb 48 hrs out is pretty horrid.

[Morris]

Time to lock this thread.

[LongBeachSurfFreak]

Rainy and windy In wantagh right now. This things gotta be one of the longest tropical impacts we have ever had here 

[bluewave]

13 minutes ago, LongBeachSurfFreak said:

Rainy and windy In wantagh right now. This things gotta be one of the longest tropical impacts we have ever had here 

This probably sets the record for the most persistent wind blown drizzle in a tropical to post tropical storm around here.

[RU848789]

2 hours ago, Morris said:

Time to lock this thread.

Why would you lock the thread?  Probably the best discussion in the 3 week histor of Hermine is going on now...

[wthrmn654]

Lock it. If you wanna start another topic maybe about what went wrong about the storm etc, that seems like a better idea.imo

[RU848789]

3 hours ago, bluewave said:

The only part of the forecast which mattered was the storm going well east on Saturday missing the strong phase and rapid deepening near the coast.

That's the only reason the earlier guidance was calling for a 2-4 or 3-5 potential surge event at the coast and much stronger winds further west

than verified. A weakening storm coming west too late off shore to matter wasn't a big deal in the big scheme of things.

Agree 110%.  All the models had the storm heading from FL to the NC coast, but how the storm evolved after that would have a huge impact on the DE/NJ/NY coasts, so IMO nailing the unexpected long duration eastward move by Hermine on Sat/Sun, as the Euro did, was the most important part of the entire storm's forecast.  Once the storm was that far east, it was almost certainly never going to have the kinds of impacts it would've had if it had stayed close to the coast.