ATL: MATTHEW - Post-Tropical - Discussion

[HurricaneBelle]

The NHC needs to seriously consider dumping their current method for creating the cone in favor of one that considers variability and uncertainty in modeling.

The current method isn't broken. It's been on point so far. The variability and uncertainty is already accounted for. Additionally, NHC has other products that provide more information than the cone. For example, while Florida is only incredibly barely in the cone, much of the FL East coast has a 20%-40% chance of TS winds. This method of delivering information works.

Here's the problem. Before the cone, they just showed the projected track as a line. So the uninformed treated the line as gospel and claimed to be surprised when the hurricane moved somewhere else. So now they've implemented the cone, and everyone treats the cone as gospel. I can't believe how many times people have done the "we're in the cone/we're not in the cone" dance when we're talking about track error at 5 days when only a few miles are at stake. Alyono has a good sticky about the cone that more people should read, as it has nothing to do with potential storm impacts over a given area.

But yeah, as I've mentioned before, and as Evil Jeremy mentions here, the real thing to watch are the wind speed probabilities which are broken down into low-end TS force winds, high-end TS force winds, and hurricane-force winds. These are more revealing than a cone that has nothing to do with forecast uncertainty and everything to do with historic track error. Why these are always forgotten around here I don't know - they never get posted in the advisories thread, and hardly anyone discusses them.

[Alyono]

Seeing some reports than more than 8 inches of rain has already fallen in the DR. May have mudslides ongoing there now

[Raebie]

Don't know, but our local met is saying he's not buying the GFS because it's too fast. And if I'm reading his take right, he thinks the track is too far west. Trying to get clarification.

so the tv met is only buying the euro..fair enough but you know better to not too hang one model at this point..this is the biggest euro run in a very long time for the united states..less than an hour to go...a major hurricane with a shot to hit the usa in heavily populated areas..havent seen this in a very long time

He's actually a very good met. But he seems to be overly biased to the Euro right now.

[Evil Jeremy]

The NHC needs to seriously consider dumping their current method for creating the cone in favor of one that considers variability and uncertainty in modeling.

The current method isn't broken. It's been on point so far. The variability and uncertainty is already accounted for. Additionally, NHC has other products that provide more information than the cone. For example, while Florida is only incredibly barely in the cone, much of the FL East coast has a 20%-40% chance of TS winds. This method of delivering information works.

Here's the problem. Before the cone, they just showed the projected track as a line. So the uniformed treated the line as gospel and claimed to be surprised when the hurricane moved somewhere else. So now they've implemented the cone, and everyone treats the cone as gospel. I can't believe how many times people have done the "we're in the cone/we're not in the cone" dance when we're talking about track error at 5 days when only a few miles are at stake. Aloyno has a good sticky about the cone that more people should read, as it has nothing to do with potential storm impacts over a given area.

But yeah, as I've mentioned before, and as Evil Jeremy mentions here, the real thing to watch are the wind speed probabilities which are broken down into low-end TS force winds, high-end TS force winds, and hurricane-force winds. These are more revealing than a cone that has nothing to do with forecast uncertainty and everything to do with historic track error. Why these are always forgotten around here I don't know - they never get posted in the advisories thread, and hardly anyone discusses them.

This is why I'm personally against expanding the cone beyond 5 days. For the past half decade there has been talk of a 7 day cone, and I believe it's being tested internally. That said, I don't think it's necessary. Actually, I think it would make things worse. We can't nail down a 5 day forecast, a 7 day forecast would only add more media/hype fuel, and seeing wild swings that far out would lower the confidence (non-weather watching) people have in the cone overall.

[psyclone]

I mention the wind probs all the time. it is one of the most valuable items in the advisories. The cone is the most misunderstood weather products out there. and it's not difficult to understand. It's baffling to me. Even more incredible is the fact that TV mets and other weather professionals routinely screw it up/misinterpret it. it's head against the wall material.

[sponger]

NAVGAM now showing into SFL enroute to Tampa!

[abajan]

The SSHWS is based on knots. 113 knots starts category 4 which is really 115 knots because they use 5 knot increments. 115 knots = 132 mph which rounds down to 130 mph.

In the past, there was an issue rounding a 115 knot wind to mph that made a cat 4 (knots) look like a cat 3 (mph). The scale was adjusted to fix this issue.

Here's a document for the benefit of those wanting more info about the change.

[WilmingtonSandbar]

Landfall near Myrtle Beach as depicted in the GFS is a nightmare scenario for SC. Getting traffic out from there is not easy, and only one road supports lane reversal if needed.

On the plus side, this is a slow mover that should allow decent advance-notice (at least for long-lead preparations), and it's October, not August, so the beach-going crowds should be smaller.

Also on the plus side, it is the lull between the summer tourist season and the Canadian snow bird season, so there normally would be fewer people in town beyond the locals.

[OntarioEggplant]

The NHC needs to seriously consider dumping their current method for creating the cone in favor of one that considers variability and uncertainty in modeling.

The current method isn't broken. It's been on point so far. The variability and uncertainty is already accounted for. Additionally, NHC has other products that provide more information than the cone. For example, while Florida is only incredibly barely in the cone, much of the FL East coast has a 20%-40% chance of TS winds. This method of delivering information works.

It is not accounted for. The cone is purely based on NHC error only. The public does not care about the other products, frankly, they only pay attention to the cone.

[Michele B]

The current method isn't broken. It's been on point so far. The variability and uncertainty is already accounted for. Additionally, NHC has other products that provide more information than the cone. For example, while Florida is only incredibly barely in the cone, much of the FL East coast has a 20%-40% chance of TS winds. This method of delivering information works.

Here's the problem. Before the cone, they just showed the projected track as a line. So the uniformed treated the line as gospel and claimed to be surprised when the hurricane moved somewhere else. So now they've implemented the cone, and everyone treats the cone as gospel. I can't believe how many times people have done the "we're in the cone/we're not in the cone" dance when we're talking about track error at 5 days when only a few miles are at stake. Aloyno has a good sticky about the cone that more people should read, as it has nothing to do with potential storm impacts over a given area.

But yeah, as I've mentioned before, and as Evil Jeremy mentions here, the real thing to watch are the wind speed probabilities which are broken down into low-end TS force winds, high-end TS force winds, and hurricane-force winds. These are more revealing than a cone that has nothing to do with forecast uncertainty and everything to do with historic track error. Why these are always forgotten around here I don't know - they never get posted in the advisories thread, and hardly anyone discusses them.

This is why I'm personally against expanding the cone beyond 5 days. For the past half decade there has been talk of a 7 day cone, and I believe it's being tested internally. That said, I don't think it's necessary. Actually, I think it would make things worse. We can't nail down a 5 day forecast, a 7 day forecast would only add more media/hype fuel, and seeing wild swings that far out would lower the confidence (non-weather watching) people have in the cone overall.

I would be in favor of a further cone (7 days vs 5 days) ONLY IF CONFIDENCE IS VERY HIGH of a certain outcome. Sometimes, there's nothing that's going to change a storm's trajectory (think Andrew, '92).

But when the steering currents are weak, or being overtaken by other currents, then I think they (who's "they" again??) owe it to the public to NOT confuse people, or lull them into a false sense of security by saying, oh", you're area is out of danger already." I believe I've heard that about a FL landfall about 8 times already since Friday!

[WPBWeather]

NAVGAM now showing into SFL enroute to Tampa!

Now that is what I call a MAX west shift. Hard to believe this thing's potential movements.

[SouthDadeFish]

Seeing some reports than more than 8 inches of rain has already fallen in the DR. May have mudslides ongoing there now

I saw that as well. I am very worried about the news that will emerge from Hispaniola over the next few days. The slow progress of the storm combined with its large moisture envelope has catastrophic potential.

[OntarioEggplant]

The NHC needs to seriously consider dumping their current method for creating the cone in favor of one that considers variability and uncertainty in modeling.

so its up to the forecaster to draw the cone? the cone will move as the track shifts accordingly, i like the cone as there is no randomness or opinion...if the private forecasters including media want to step out and make a forecast than they can..look, there is plenty of time to get ready in florida, everybody is warned before june 1 in hurricane areas to have supplies and a plan..if you dont then you should

No. By accounting for the variability in models, it shouldn't shift much overall, it just makes a larger cone. The shifts would be very slight unless there was a massive shift in all models with good agreement.

[ThetaE]

Interesting... Is it me, or has Matthew finally begun the long-awaited NNE turn?


Wobble Watching!

Well, maybe "turn" is a generous word. More like a lean.

[KWT]

Seeing some reports than more than 8 inches of rain has already fallen in the DR. May have mudslides ongoing there now

The event has barely started for them as well, got a solid 24hrs of really heavy rain for that neck of the woods to come yet...cannot see anything but mudslides and death from this.

[Hurricaneman]

Seeing some reports than more than 8 inches of rain has already fallen in the DR. May have mudslides ongoing there now

I saw that as well. I am very worried about the news that will emerge from Hispaniola over the next few days. The slow progress of the storm combined with its large moisture envelope has catastrophic potential.

and not just there, portions of the east coast of the US could see considerable damage depending on track

[BirdyCin]

Brevard County EOC beginning the distribution of sandbags:
http://apps.brevardcounty.us/PublicInfo ... px?ID=2937

[Raebie]

Can you say HUGO II ?

A little north, but yes. That is the southern Grand Strand area, and evac will be a mess.

Local WX folks are not sold on the trend west. The term finger ridge was brought up on a recent call and was dismissed. When will the next Euro run start?

In a little less than an hour.

And there aren't many roads out of Myrtle or Wilmington. Basically Columbia, Charlotte, and Raleigh the only directions with a few national highways. ILM doesn't do I-40 lane reversal anymore as a plan for mass-evac. Would be nuts.

Well the good news is that UNC Wilmington is out on fall break next Thursday and Friday. Tons of students will be heading out for home mid-week. Daughter is about to get instructions to join them.

[Evil Jeremy]

The NHC needs to seriously consider dumping their current method for creating the cone in favor of one that considers variability and uncertainty in modeling.

so its up to the forecaster to draw the cone? the cone will move as the track shifts accordingly, i like the cone as there is no randomness or opinion...if the private forecasters including media want to step out and make a forecast than they can..look, there is plenty of time to get ready in florida, everybody is warned before june 1 in hurricane areas to have supplies and a plan..if you dont then you should

No. By accounting for the variability in models, it shouldn't shift much overall, it just makes a larger cone. The shifts would be very slight unless there was a massive shift in all models with good agreement.

So what you want is a larger cone basically? The size of the cone is determined by statistical error.

[KWT]

Alot of mets seems to be banking on the trends that is recently happening not really reflecting reality, risky to go down that route for sure!

Does seem like an east hint to the motion, still near due north but a touch east, like 5 degrees or something.