Seven-day hurricane warnings on the horizon

[wxman57]

The article says in about a decade or more. My opinion is that if you can't make an almost perfect 24 hr forecast, what would make you think that you can make an accurate 168 hr forecast?

That depends on what you mean by "accurate". Even 24 hour forecasts are off about 80 miles on average (67% of time). At 5 days it's closer to 400 miles. The point is not to get a track perfect 7 days out, it's about risk assessment. Some clients just cannot tolerate much risk at all.

But if a 7-day forecast can identify a potential threat, say, to anywhere in the NW Gulf then they may delay a risky operation and wait for another weather window. This kind of forecast is reserved for the very few who need that little bit of extra info about a disturbance/TC that could pose a threat in the long term.

I agree, a 7-day track for most people (like the general public) is useless due to the large range of error.

[Scorpion]

I like the idea, however I'm afraid it will set off a trend of news stations constantly harping these 7 day tracks and scaring the public.

[Brent]

No, no, and NO! The 5-day cone is bad enough.

[DanKellFla]

This looks like beuracratic B.S. to me. Somebody had a meeting and got an 'action item' to make some 'stretch goals' so the 'paradigm' of forecasting changes. And, along the way, a request for a larger budget is made in order to meet the goals.

[wobblehead]

Some storms like the cat 5's last year you could put out a 10 day track and pretty much be right on. Other storms your lucky if you can get a 1 day track right.

My vote is to improve forecast of 3 areas: intensity, intensity, and last but not least intensity.

[Cryomaniac]

Some storms you'd be lucky to get a 6 hour forecast right.

[Category 5]

:uarrow: Some storms you'd be lucky to get a 6 hour forecast right.

Exactly. A 7 day forecast just has PR nightmare written all over it.

[Derek Ortt]

until model NCEP stops depicting cat 5 hurricanes like Felix as weak, open waves, model NCEP should not really be used as TP... much less a 7 day forecast

[Tampa Bay Hurricane]

Technology is not good enough yet for a 7 day forecast. The five
day error is still hundreds of miles.

[wxman57]

Technology is not good enough yet for a 7 day forecast. The five
day error is still hundreds of miles.

I would say that the statement that "Technology is not good enough yet for a 7 day forecast." depends upon how the forecast will be used. If you're talking about using a 7-day forecast to call for evacuations along the coast, then I'd agree. Distribution of such a long-range forecast to the general public would do more harm than good. But let's say we're talking about a company considering towing a multi-billion dollar deepwater platform from the shipyard at Ingleside, TX to the deepwater areas off the Louisiana coast during hurricane season. A 7-day forecast can be good enough to warn them of a potential problem in the northwest Gulf over the next 7-10 days, the time it would take to safely tow the platform out and get it securely moored. A 5-day forecast won't cut it for such operations, they need a 7-10 day window. This means 7-day forecasts not only for named systems (TD/TS/H) but also for tropical disturbances. It's a real challenge, but any information that we can give them can help them to make that multi-billion dollar decision.

Of course, these forecasts are not seen by the general public, or even the vast majority of our clients. These forecasts are reserved for the handfull of highly risk-averse clients who understand what we're providing and are willing to accept the risks that such forecasts are not high-confidence.

[Ed Mahmoud]

Technology is not good enough yet for a 7 day forecast. The five
day error is still hundreds of miles.

I would say that the statement that "Technology is not good enough yet for a 7 day forecast." depends upon how the forecast will be used. If you're talking about using a 7-day forecast to call for evacuations along the coast, then I'd agree. Distribution of such a long-range forecast to the general public would do more harm than good. But let's say we're talking about a company considering towing a multi-billion dollar deepwater platform from the shipyard at Ingleside, TX to the deepwater areas off the Louisiana coast during hurricane season. A 7-day forecast can be good enough to warn them of a potential problem in the northwest Gulf over the next 7-10 days, the time it would take to safely tow the platform out and get it securely moored. A 5-day forecast won't cut it for such operations, they need a 7-10 day window. This means 7-day forecasts not only for named systems (TD/TS/H) but also for tropical disturbances. It's a real challenge, but any information that we can give them can help them to make that multi-billion dollar decision.

Of course, these forecasts are not seen by the general public, or even the vast majority of our clients. These forecasts are reserved for the handfull of highly risk-averse clients who understand what we're providing and are willing to accept the risks that such forecasts are not high-confidence.

If anybody but a private forecasting company that only shares the forecasts with paying clients does this, ie, if NHC does this, than I'd expect the fun and games here over 144 hour CMC forecasts will be small potatoes, as NHC would have to publically provide the data in some way to end users.

[Tampa Bay Hurricane]

Good point Wxman. 7-10 day forecasts in general are pretty good
for forecasting weather conditions, but for specific evacuations it is
not as great.

[Ptarmigan]

It would be nice to see 7 day forecasts. However, anything can change in 7 days. Let alone we cannot accurately predict the weather 7 days in advance.

[RL3AO]

Nothing will help public confidence in NHC forecasts like having over half of the Gulf coast in the cone for a day 7 forecast.

[wxman57]

Nothing will help public confidence in NHC forecasts like having over half of the Gulf coast in the cone for a day 7 forecast.

Extrapolating the 5-day cone out to 7 days would yield a 7-day circle about 1200 miles across. That's about from Brownsville, TX to Charleston, SC - a bit larger than half the Gulf coast.

Again, such forecasts would be useless for the general public. Their value lies in risk management for certain very expensive special projects which need such lead times.