NHC's Tropical Weather Outlook: How Accurate are its Predictions?

By: Dr. Jeff Masters , 2:46 PM GMT on July 20, 2013

Share this Blog
73
+

Every Internet-savvy tropical weather enthusiast is familiar with the National Hurricane Center's Tropical Weather Outlook (TWO), which details potential threat areas that might become a tropical cyclone. (Tropical depressions, tropical storms, and hurricanes are all tropical cyclones.) The Tropical Weather Outlook is issued four times per day during hurricane season, and beginning in 2010, NHC began issuing 48-hour forecasts of the probability that specific threat areas identified in the TWO could develop into a tropical cyclone. Their Graphical Tropical Weather Outlook now color-codes each threat area depending upon how likely development is expected to be. A yellow circle is drawn for 0%, 10%, and 20% chances; orange for 30%, 40% and 50% chances, and red for 60% and higher odds. For example, the Saturday morning, July 20 TWO gave a 0% chance of development for an area of disturbed weather near the Florida Panhandle (Figure 1.)


Figure 1. NHC's Tropical Weather Outlook for 8am EDT July 20, 2013, showed an area of disturbed weather over the Gulf of Mexico being given a 0% chance of development in 48 hours. In 2012, 8% of all disturbances being given a 0% chance of development actually did develop.

How accurate is NHC's Tropical Weather Outlook?
So the big question is, how good are these forecasts? When NHC gives a 30% chance that an "Invest" will become of tropical cyclone, does this happen 30% of the time? Well, according to the 2012 National Hurricane Center Forecast Verification Report, NHC should have drawn fewer yellow circles and more orange circles and red circles during 2012, as there was a tendency to under-predict when a threat area might develop. For example (Figure 2), for the 46 forecasts where a 30% chance of development was given, 50% of the threat areas actually developed. Every disturbance that was given a 70% and higher chance of development ended up developing. This under-prediction tendency in 2012 is in contrast to the results from 2011, when the genesis forecasts were closer to the mark. For example, the 59 forecasts for a 30% chance of development resulted in a 31% "hit" rate of the storm actually developing in 2011. So far in 2013, there seems to be a tendency to under-predict again. For example, 48 hours before Tropical Storm Barry developed, NHC was carrying just a 10% chance of development.




Figure 2. NHC did predictions on new formation of a tropical depression or tropical storm (cyclogenesis) beginning in 2010. The forecasts are expressed in the Tropical Weather Outlook in 10% probability increments, and in terms of categories (“low”, “medium”, or “high”) for a tropical cyclone forming within a 48-hour period. These genesis forecasts had a low (under-forecast) bias in the Atlantic basin during 2012 (top). For example, for cases where a 30% chance of formation was given, the actual percentage of storms that formed was 50%. However, there the 2011 forecasts showed no systematic bias, and were closer to the mark (bottom.)

Expansion of the Tropical Weather Outlook to 5 Days
By August of the 2013 hurricane season, NHC is planning to begin including information about a system’s potential for development during the following five-day period. This will supplement the 48-hour probabilistic formation potential already provided in the Tropical Weather Outlook. NHC is currently developing a corresponding five-day genesis potential graphic that might also be available in 2013.

Quiet in the Atlantic
There are no other tropical cyclone threat areas in the Atlantic to discuss today. Saturday morning's 06Z run of the GFS model predicted that a tropical wave expected to move off the coast of Africa on Monday will develop later in the week. None of the other reliable models develop this wave, though the unreliable Canadian (CMC) model also suggests that the wave could develop.

Have a great weekend, everyone, and I'll be back Monday with a new post.

Jeff Masters

Reader Comments

Comments will take a few seconds to appear.

Post Your Comments

Please sign in to post comments.

Sign In or Register Sign In or Register

Not only will you be able to leave comments on this blog, but you'll also have the ability to upload and share your photos in our Wunder Photos section.

Display: 0, 50, 100, 200 Sort: Newest First - Order Posted

Viewing: 542 - 492

Page: 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54Blog Index

Member Since: July 3, 2005 Posts: 437 Comments: 134812
Member Since: July 3, 2005 Posts: 437 Comments: 134812
Member Since: July 3, 2005 Posts: 437 Comments: 134812
Quoting 521. pottery:
Wow!
It's coming.
It's HUGE.
We'er DOOMED....

..sorry. I get a little excited when I see a cloud, these days.
Me too..Especially when two merge into one......its terrifically powerful stuff.
Member Since: Posts: Comments:
Quoting 534. TropicalAnalystwx13:
People put entirely too much stock into the ECMWF when looking a storm that has a chance to develop. In fact, that's the models biggest issue in the tropics. If you can find more than one run where the ECMWF developed Andrea, or at least one where it developed either Chantal, Alvin (EPAC), Barbara (EPAC), and Dalila (if I remember correctly), I'd be glad to see it. The fact that the model does not develop the wave over Africa does not mean it won't develop. The GFS is a reliable model too.
It has been consistent to.I'd say at least a T.S.
Member Since: Posts: Comments:
GFS at 156 hours..wont be a fast moving storm............
Member Since: Posts: Comments:
Quoting nigel20:

Lol! We're having extremely dry conditions in Kingston. Two of the reservoirs in Kingston are below 60% of capacity. Most of Jamaica is having above normal rainfall with the exception on eastern Jamaica.

Yeah, thats pretty strange for this time of year.
Hope the dams fill up before dry season.
Member Since: Posts: Comments:
Quoting 505. Grothar:


But, we've never been able to figure out where BaltimoreBrian and JerseyCityGal are from.


haha. Still though, Brian could live right on the water, or 'uphill' and inland. Some of those neighborhoods on the water are awesome, but they are right on the water. Never been to Jersey City (just drove by/through), so don't really have a feel for it, as it pertains to surge, etc. I know Battery Park got hit, but that's about all I know.

And Brian seems to keep close tabs on NYC - his moniker could be a red herring...
Member Since: Posts: Comments:
People put entirely too much stock into the ECMWF when looking a storm that has a chance to develop. In fact, that's the models biggest issue in the tropics. If you can find more than one run where the ECMWF developed Andrea, or at least one where it developed either Chantal, Alvin (EPAC), Barbara (EPAC), and Dalila (if I remember correctly), I'd be glad to see it. The fact that the model does not develop the wave over Africa does not mean the system cannot. The GFS is a reliable model too, and it gains even more credence with the wave being over West Africa and still showing development consistently.
Member Since: Posts: Comments:
GFS wont give up...192 hours..
Member Since: Posts: Comments:
Quoting pottery:
Wow!
It's coming.
It's HUGE.
We'er DOOMED....

..sorry. I get a little excited when I see a cloud, these days.

Lol! We're having extremely dry conditions in Kingston. Two of the reservoirs in Kingston are below 60% of capacity. Most of Jamaica is having above normal rainfall with the exception on eastern Jamaica.
Member Since: Posts: Comments:
Not the best view, but we have an idea

Member Since: July 17, 2009 Posts: 72 Comments: 28763
Quoting sar2401:

Dr. Masters clearly laid out the problem I have with the NHC. They consistently underestimate AOI's in terms of the probability of development. The point of a 30% AOI is not that there's a medium chance something might form. It means there's a 3 out of 10 chance a tropical cyclone will form in the area of the circle, and a 7 out of 10 chance a tropical cyclone won't form. The NHC has been consistently under forecasting the probability that TC will develop. As Dr. Masters pointed out, when the NHC (in 2012) had a AOI at 30%, the actual chance of a TX forming was 50%. That's a large under forecast bias in estimating TC formation, which is why we've seen 20% and 30% AOI's jump to a TD or even a TC six hours later. This under forecast bias is even worse from 70% up, when 100% of all AOI's with a 70% or above probability became TC's.

Something changed from 2011, when this under forecast bias was well within a typical margin of error, in 2012 and, so far, in 2013. I don't know how AOI percentages are calculated, but I suspect a lead forecaster has a lot of influence on the number. Maybe a more conservative forecaster has influenced the AOI percentage, or something changed in the algorithms used to generate the percentage. Whatever the cause, it's a bad trend for people to see a yellow circle and thing "No big deal" when the actual verification shows that there should have been more orange and red circles and less yellow. I have been saying essentially the same thing this season and have been vilified for trying to make the NHC look "bad". I appreciate Dr. Masters dispassionate look at the issue and pointing out this is a weakness in the AOI system than needs to addressed.


..within 48 hrs.
Thats the caveat.
Member Since: Posts: Comments:
storms in the northern gulf swing to the atlantic Monday...
Member Since: Posts: Comments:
Interesting...

SYNOPSIS 2013072000

P12L
10N, 1W
700 hPa


ECMWF: The hovmoller signal is quite distinct. However, tracking a pouch is more difficult. For the first couple days, I track a tiny OW max (one of many over west Africa). Upon reaching the Atlantic, P12L becomes difficult to track, with little to no OW max or trough in the immediate vicinity. About a day later, a circulation is definitely spinning off the African coast, in roughly the area where P12L should be, but it never has a clear center, and the fact that P12L weakens so much during the middle portion of the forecast would make for a few too many "guess" points to get to this circulation. I stop tracking at the coast, but keep in mind that ECMWF does attempt to spin up something off the coast not long thereafter.

GFS: Easily tracked pouch the entire 120 hours. P11L, which is west of P12L, weakens as P12L approaches it. OW increases as P12L moves off of Africa.

UKMET: While ECMWF is only hinting at starting to agree with the bullish GFS, UKMET has jumped on the bandwagon. (Interesting since UKMET had no positions at all for P12 in yesterday's 00Z forecast.) UKMET does weaken P12L a little as it leaves Africa, as well as on Day 5 over the central Atlantic, but otherwise, UKMET depicts a distinct pouch all 120 hours.

NAVGEM:

HWRF-GEN:


ECMWF -9.0 v700 48h
GFS -9.1 v700 120h
UKMET -9.8 v700 120h
NAVGEM ---- ---- ---h
HWGEN ---- ---- ---h
Member Since: June 30, 2013 Posts: 12 Comments: 9439
This is from CWG while Sandy was making landfall

Those gusts have been quite high recently as well. National has reached at least 58 mph, which surpasses the criteria needed to verify a severe thunderstorm warning and indicates high potential for tree downing as well as other damage. Gusts between 60-80 mph remain possible. There’s no reason to venture out there, please stay safe!
Member Since: Posts: Comments:
Quoting VR46L:
I don't have have a problem with the NHC giving 10% chance of a system developing .all they are doing is recognizing a that a system has a 1 in 10 chance of development into a TC in the next 48 hours . If a system makes a TC in that 48 hr time frame they are not wrong but also if it does not develop again they are not wrong . all they are stating is there is some likehood of the AOI developing and that they are keeping their eye on it.

Dr. Masters clearly laid out the problem I have with the NHC. They consistently underestimate AOI's in terms of the probability of development. The point of a 30% AOI is not that there's a small chance something might form. It means there's a 3 out of 10 chance a tropical cyclone will form in the area of the circle, and a 7 out of 10 chance a tropical cyclone won't form. The NHC has been consistently under forecasting the probability that TC will develop. As Dr. Masters pointed out, when the NHC (in 2012) had a AOI at 30%, the actual chance of a TC forming was 50%. That'sa large under forecast bias in estimating TC formation, which is why we've seen 20% and 30% AOI's jump to a TD or even a TS six hours later. This under forecast bias is even worse from 70% up, when 100% of all AOI's with a 70% or above probability became TC's.

Something changed from 2011, when this under forecast bias was well within a typical margin of error, in 2012 and, so far, in 2013. I don't know how AOI percentages are calculated, but I suspect a lead forecaster has a lot of influence on the number. Maybe a more conservative forecaster has influenced the AOI percentage, or something changed in the algorithms used to generate the percentage. Whatever the cause, it's a bad trend for people to see a yellow circle and think "No big deal" when the actual verification shows that there should have been more orange and red circles and less yellow. I have been saying essentially the same thing this season and have been vilified for trying to make the NHC look "bad". I appreciate Dr. Masters dispassionate look at the issue and pointing out this is a weakness in the AOI system than needs to addressed.
Member Since: October 2, 2004 Posts: 0 Comments: 22222
Quoting Grothar:


That would be a good trivia question. What is the average time it takes a pouch to become a blob?


The Kangaroo knows......
Member Since: Posts: Comments:
Thunderstommike , at it again saying , it's going to be quiet for the next month ! Untrue if that TW coming of Africa , developes , which some of the models suggest , and other bloggers say it looks like , my guess we will have Dorian soon !
Member Since: Posts: Comments:
Quoting 499. Tropicsweatherpr:


Yes,is now pouch 12L.

Link


That would be a good trivia question. What is the average time it takes a pouch to become a blob?
Member Since: July 17, 2009 Posts: 72 Comments: 28763
I actually learned during Sandy that I live in a tidal flooding zone, the fear was that Sandy's center would come ashore in Delaware and thus push a monstrous surge up the Delaware River and cause significant flooding in town. Some areas were evacuated, which I had never seen before.
Member Since: Posts: Comments:
Wow!
It's coming.
It's HUGE.
We'er DOOMED....

..sorry. I get a little excited when I see a cloud, these days.
Member Since: Posts: Comments:
Dont let the sun being out fool ya..the storms ARE coming folks..
Member Since: Posts: Comments:
Quoting 512. Matt74:
ok. Thanks. So you're thinking the GFS is too quick on development ?
Actually it is not out of the realm of possibility for the system to emerge with a low pressure already, if there is enough spin to it.
Member Since: June 30, 2013 Posts: 12 Comments: 9439
A small surface trough located several hundred miles southeast of the Big Island of Hawaii has increased its thunderstorm intensity and coverage during the diurnal convective maximum. Satellite and scatterometer passes show no evidence of a closed surface circulation. The strong west-southwesterly shear as well as dry air will prevent any development of this disturbance. The system will move westward over the next several days and bring an increase in shower activity to my area by early next week.

Member Since: July 21, 2011 Posts: 83 Comments: 7167
Quoting 508. LAbonbon:


My experiences are split between storms I experienced while living in New England, and here in LA. For me, as I've never lived in a surge area, my fear has always been the darn trees. When they start cracking and snapping around me, I get verrry nervous.

During Isaac, I lost power (Entergy), but 2 houses down it's a local co-op (DEMCO). They never lost power, and they often stay up when we go down. But the kind neighbors offered their house as a place of respite from the heat for those of us poor souls saddled w/ Entergy. Gloria (in CT) was my longest outage, though, something like 2 weeks. Overall I've been lucky on that front. Even Gustav I was only down less than a week, while others here were down much longer.

Well for Isabel I lost power for two weeks and a huge tree fell down in my yard.How ever they fixed the transformer which at the time was 30 years old when Isabel came to visit.So we rarely lose power since it's up to date now.
Member Since: Posts: Comments:
This is the chart I will be paying attention to, to see later on if they attach a low pressure to the wave the GFS develops.

Member Since: June 30, 2013 Posts: 12 Comments: 9439
Member Since: Posts: Comments:
Intensification, eh? Not good:

Member Since: Posts: Comments:
.
Member Since: Posts: Comments:
Quoting 510. GTstormChaserCaleb:
I think it was for 2 or 3 runs, but that was more in the long range, according to this latest run of the GFS development may occur in the next 3 days. Although I am going with 4.
ok. Thanks. So you're thinking the GFS is too quick on development ?
Member Since: Posts: Comments:
Member Since: Posts: Comments:
Quoting 509. Matt74:
Yea but wasn't it also showing a storm last week or so for several runs?
I think it was for 2 or 3 runs, but that was more in the long range, according to this latest run of the GFS development may occur in the next 3 days. Although I am going with 4.
Member Since: June 30, 2013 Posts: 12 Comments: 9439
Quoting 487. GTstormChaserCaleb:
Here is the thing if it were the CMC showing this then I would say "Ghost," but the mere fact the GFS consistently has been showing a storm developing off the coast of Africa for days now leads me to believe we will have something noteworthy to track and that the GFS has more credibility than the CMC...If nothing develops I'll be the first to dive into some crow and admit my fault.
Yea but wasn't it also showing a storm last week or so for several runs?
Member Since: Posts: Comments:
Quoting 485. washingtonian115:
For Sandy I recorded a gust of 65mph at my house (Not sure about the air ports though).She brought down some trees and caused minimal widespread damage in the city.Irene I recorded a gust of 50mph R.N.A recorded 56 mph.Their was a few branches and trees down in my area after when she hit as well.Not to mention power outages for Sandy my lights went out for about an 1 hour.Irene they never went off.But the neighborhood up the street from me they had a entire black out.


My experiences are split between storms I experienced while living in New England, and here in LA. For me, as I've never lived in a surge area, my fear has always been the darn trees. When they start cracking and snapping around me, I get verrry nervous.

During Isaac, I lost power (Entergy), but 2 houses down it's a local co-op (DEMCO). They never lost power, and they often stay up when we go down. But the kind neighbors offered their house as a place of respite from the heat for those of us poor souls saddled w/ Entergy. Gloria (in CT) was my longest outage, though, something like 2 weeks. Overall I've been lucky on that front. Even Gustav I was only down less than a week, while others here were down much longer.

Member Since: Posts: Comments:
Does anyone know if the GFS predicted storm has an anti cyclone. And if so how do you know?
Member Since: Posts: Comments:
Quoting Patrap:
The power of Wave, Surge, Wind


Wow! These pics shows just how powerful a storm surge can be.

The Palisadoes (tombolo connecting Port Royal to the mainland) in Jamaica was constantly being batter by frequent storms even from indirect impact. Extensive work was done on the peninsula to prevent major damage from storm surge. It stood up well to hurricane Sandy while it was incomplete, but it's yet to face a major test. You can read more on my blog.
Member Since: Posts: Comments:
Quoting 477. LAbonbon:


I was just wondering, given people's reactions to potential and past storms. I am slowly (very very slowly) starting to get a handle on where some of the blog's 'regulars' are from.

W/ Sandy & Irene, how much impact to you in the city?


But, we've never been able to figure out where BaltimoreBrian and JerseyCityGal are from.
Member Since: July 17, 2009 Posts: 72 Comments: 28763
There we go...FIM-9 132 hrs.

Member Since: June 30, 2013 Posts: 12 Comments: 9439
Quoting 501. Sfloridacat5:


SPC's not showing anything coming up this week.

Maybe you need to look at the new ecmwf model and the low track and 850 mb wind.In the summer there is copious amounts of moisture...so that formula spells trouble.Go to instantweathermaps.
Member Since: July 2, 2013 Posts: 0 Comments: 1576
My Hood,in 2005, 1 mile west on the Dry Side, 17th St Canal Breach in this image.

thumb.4325
Member Since: July 3, 2005 Posts: 437 Comments: 134812
Quoting 490. Camille33:
Ecmwf shows a significant tornado outbreak evolving forma colorado low ....over the midd misouri valley spread into the tennesse valley and then into the northest. This is highly unusual to see this kind of a set-up in July.This is reminiscent of a noreaster set-up with a semi 50/50 low.


SPC's not showing anything coming up this week.
Member Since: Posts: Comments:
Quoting 496. wunderkidcayman:


Hey guys
I'm guessing the wave that GFS is forecasting to develop is the one that is near 10N 05E correct me if am wrong

That's what I think, the one in front of it looks like it would come off too far north to develop because of the cooler SSTs and drier air.
Member Since: February 11, 2012 Posts: 88 Comments: 8446
Quoting 496. wunderkidcayman:


Hey guys
I'm guessing the wave that GFS is forecasting to develop is the one that is near 10N 05E correct me if am wrong


Yes,is now pouch 12L.

Link
Member Since: April 29, 2009 Posts: 75 Comments: 15728
12z Experimental FIM:

Member Since: June 30, 2013 Posts: 12 Comments: 9439
KATRINA BASE MAP INDEX

Click on thumbnails in the above link


Member Since: July 3, 2005 Posts: 437 Comments: 134812
Quoting 469. hydrus:


Hey guys
I'm guessing the wave that GFS is forecasting to develop is the one that is near 10N 05E correct me if am wrong
Member Since: Posts: Comments:


MESOSCALE DISCUSSION 1448
NWS STORM PREDICTION CENTER NORMAN OK
0206 PM CDT SAT JUL 20 2013

AREAS AFFECTED...ERN WY...SW SD...WRN NEB

CONCERNING...SEVERE POTENTIAL...WATCH POSSIBLE

VALID 201906Z - 202130Z

PROBABILITY OF WATCH ISSUANCE...60 PERCENT

SUMMARY...THUNDERSTORMS WILL INCREASE ACROSS ERN WY AND MOVE INTO SW
SD AND WRN NEB THIS AFTERNOON AND EVENING. DAMAGING WINDS AND LARGE
HAIL WILL BE THE MAIN THREATS WITH THESE STORMS AS THEY TRACK TO THE
E/SE. TRENDS WILL BE MONITORED FOR POSSIBLE SEVERE WATCH ISSUANCE IN
THE NEXT 1-2 HOURS.

DISCUSSION...18Z MESOANALYSIS INDICATED A WEAKENING CAP OVER THE MCD
AREA. THIS APPEARS TO BE REASONABLE AS CUMULUS HAS BEGUN TO DEVELOP
ACROSS THE REGION. AS TEMPERATURES CONTINUE TO WARM TO AROUND 90
DEGREES IN THE NEXT 1-2 HOURS...INHIBITION SHOULD CONTINUE TO
DECREASE...AND THUNDERSTORM COVERAGE WILL INCREASE BY 21Z. MODERATE
INSTABILITY AND EFFECTIVE SHEAR VALUES WILL BE ADEQUATE TO SUSTAIN
ORGANIZED STORM CLUSTERS/SUPERCELLS. INVERTED-V TYPE SOUNDINGS
SUGGEST A DAMAGING WIND THREAT WILL BE POSSIBLE AND STEEP MIDLEVEL
LAPSE RATES EXCEEDING 7.5 DEG C PER KM WILL PROMOTE LARGE HAIL.

BASED ON SFC ANALYSIS...THERE APPEAR TO BE AT LEAST A FEW NW-SE
ORIENTED BOUNDARIES LYING ACROSS THE REGION FROM EARLIER CONVECTION.
STORMS SHOULD GENERALLY MOVE IN AN E-SE DIRECTION ALONG THE ERN
PERIPHERY OF INVERTED SFC TROUGH. TRENDS WILL CONTINUE TO BE
MONITORED FOR POSSIBLE SEVERE WW ISSUANCE IN THE NEXT 1-2 HOURS.

..LEITMAN/EDWARDS.. 07/20/2013


ATTN...WFO...LBF...UNR...BOU...CYS...BYZ...

LAT...LON 40940341 41360461 41840542 42230579 42620578 44530564
44900499 45040408 44970346 44540263 43960195 42950102
41480048 40780105 40710267 40940341
Member Since: Posts: Comments:
Quoting MAweatherboy1:

Hey Nigel. We've been baking for days up here, I hit 94 at my house southwest of Boston today, and now have some strong thunderstorms moving in.

Hey MA. Wow! The record in Jamaica is just 98F with an average temp of about 90F, so I know that 94F would be really uncomfortable.
Member Since: Posts: Comments:
high res. east coast still vis image



Member Since: Posts: Comments:
The disturbance given a 0% chance dissipated.
Member Since: Posts: Comments:

Viewing: 542 - 492

Page: 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54Blog Index

Top of Page

About

Dr. Masters co-founded wunderground in 1995. He flew with the NOAA Hurricane Hunters from 1986-1990. Co-blogging with him: Bob Henson, @bhensonweather

JeffMasters's Recent Photos

Gust front cloud, SE Michigan
Thunderstorm over Grand Teton
Double rainbow over Old Faithful
Rainbow in Riverside Geyser