Seven day cone - Good or bad idea?

[cycloneye]

There is talk about a seven day cone implementation. Discuss the pros and cons in the thread about this idea.

[HURAKAN]

I think it will be a great idea once we can have the same error spread in the 5-day cone as we have today in the 3-day cone. When that happens, then we're ready for a week cone. Right now it will be crazy but I can see it in the future.

[somethingfunny]

What would be the point?

[FireRat]

No-Way-Jose!

7 Day cone would hype more than the mid-range models already do. I'd rather give an orangutan a gun and have him shoot at the map of the Atlantic Basin, precisely the Caribbean/S.E. Atlantic coast. He'll likely miss the target altogether meaning no storm exists or land the bullet hundreds, if not thousands of miles away.

We need to sharpen the 5 day cone more before considering 7 day cones. We can't even see the strength of a storm 5 days out with much success, that in 7 days there might not even be a storm.

[Aquawind]

I agree that with the current margin of error I see no need to include a couple more days of even greater uncertainty. I think it defeats the idea of getting better at disrupting the fewest people possible outside of the actual storm impact. I am all for pushing forcasters to improve their skills but, the effort and time required will only hurt public perception of quality and preparation. Think of the size the 7day cone would cover..wayyy many more people than needed if 5 days is enough to prepare imo.

[abajan]

I agree that with the current margin of error I see no need to include a couple more days of even greater uncertainty. I think it defeats the idea of getting better at disrupting the fewest people possible outside of the actual storm impact. I am all for pushing forcasters to improve their skills but, the effort and time required will only hurt public perception of quality and preparation. Think of the size the 7day cone would cover..wayyy many more people than needed if 5 days is enough to prepare imo.

Yep. I totally agree. Perhaps forecasting skills will be good enough (say) ten years from now that the radius of uncertainty seven days out will be no wider than currently obtains for five days. I think the plan is for seven day cones to be made public in two to three years time but that’s too soon IMO.

[psyclone]

bad idea. the 5 day cone of death provides plenty torture.

[Hurricanehink]

I think if they're already doing it, why not show it to the public? The NWS already does a seven day forecast. Is someone going to freak out if they see there is a chance of rain in 7 days? No, of course not. They see it's there and that it's so far away that it could easily change. Nevertheless, the option is still good to have.

I feel it's the same with tc warning. In the grand scheme of things, there is little difference between five and seven days. I believe the public knows the uncertainty, and if not, such uncertainty should be emphasized. Speaking from personal experience, when Earl was forecast to be off the coast in five days, my co-workers took note of it and they started making preliminary plans. Nothing significant, just "oh, if Earl brings us rain in five days, I'll do this instead". I don't see why seven days would be any different.

[JonathanBelles]

Yes, it is true. The NHC is running 7-day forecasts in house with every advisory package. When I was at the NHC in June, they said it would probably be 2-3 years before the public sees it.

[somethingfunny]

I think if they're already doing it, why not show it to the public? The NWS already does a seven day forecast. Is someone going to freak out if they see there is a chance of rain in 7 days? No, of course not. They see it's there and that it's so far away that it could easily change. Nevertheless, the option is still good to have.

I feel it's the same with tc warning. In the grand scheme of things, there is little difference between five and seven days. I believe the public knows the uncertainty, and if not, such uncertainty should be emphasized. Speaking from personal experience, when Earl was forecast to be off the coast in five days, my co-workers took note of it and they started making preliminary plans. Nothing significant, just "oh, if Earl brings us rain in five days, I'll do this instead". I don't see why seven days would be any different.

Yes but this is exactly what the existing 7 day forecasts from the local NWS office are meant for. If NJ is on the edge of Earl's 7 day cone then the NWS can give you a 20% chance of rain without raising hurricane alarms.

[vbhoutex]

I agree that with the current margin of error I see no need to include a couple more days of even greater uncertainty. I think it defeats the idea of getting better at disrupting the fewest people possible outside of the actual storm impact. I am all for pushing forcasters to improve their skills but, the effort and time required will only hurt public perception of quality and preparation. Think of the size the 7day cone would cover..wayyy many more people than needed if 5 days is enough to prepare imo.

Yep. I totally agree. Perhaps forecasting skills will be good enough (say) ten years from now that the radius of uncertainty seven days out will be no wider than currently obtains for five days. I think the plan is for seven day cones to be made public in two to three years time but that’s too soon IMO.

I wouldn't say that is too soon, but it may be. Mostly, at least to me, it depends if NHC has managed to bring down the margin of error significantly enough to make a cone that is realistic in size that will not "scare" more people than necessary. With the advances seeming to continue somewhat it may be possible, but I sure would like to see the intensity forecasts improve along with it. We all know the track forecasts have improved significantly over the last 10 years.

[RL3AO]

Not yet. Still five years away from it being a good idea in my opinion.

Even if they got better with the track forecasts, when they usually are every year, intensity forecasts for day 7? Big problems. If they did track only for days 6-7, fine. If you have a weak storm in September approaching the islands and you have it near the Yucatan straight in 7 days, what do you do with intensity? That little trick of forecasting 95kts the first advisory and waiting to see if it takes off doesn't work for days 6-7. Do you forecast a 125kt storm near Cuba when its just a 35kt storm over the Atlantic? Creates more problems then it solves IMO.

[wxman57]

When Bill Read brought up the 7-day track forecasts at the AMS meeting in Tucson last spring, he turned to me and asked if I had any comments about the potential value to our clients. I told him that we have a number of quite weather-sensitive clients operating in the deepwater areas off TX/LA (or at least we did until the moratorium). These clients had "T-times" (total preparation time for getting the rigs secure and evacuation) of up to 7 days. So they have to take action greater than 7 days in advance of a potential impact. That's why we've been producing 7-day tracks on active storms AND disturbances (for a few super-sensitive clients) for the last 2-3 years. So there is most definitely a need for a 7-day track, just as there was a need for a 5-day track ten years ago (US Navy).

One of the goals of an ongoing hurricane project is to reduce track error by half in the next 10 years. That would mean that the error circle at day 7 could be less than the 650 mile wide circle currently at the end of the 5-day track. Initially, though, the 7-day error cone (circle) would be in the 850-950 mile diameter range - considerably larger than the present 5-day circle. That's big enough to take up most of the Gulf of Mexico. I don't think that most businesses or local governments would actually take any significant actions 7 days prior to impact, but I think they'd be curious if that hurricane projected to be near western Cuba at day 5 would then be heading to Texas or Florida.

Realistically, it may be around 2014-2015 before any 7-day track goes public. It won't really matter to me much either way, as we'll likely go with a 7-day track for all clients next year or the year after. I'd be curious to see the NHC's thoughts on the 7-day track for active storms, though.

[Cryomaniac]

Wxman57, do you think that intensity forecasting will improve enough by 2014 to make 7 day forecasts reliable in terms of track and intensity? We've all seen complaints and examples of intensity forecasts being vastly wrong (storms forecast to be hurricanes dissipating, and to a lesser extent vice versa).

[wxman57]

Wxman57, do you think that intensity forecasting will improve enough by 2014 to make 7 day forecasts reliable in terms of track and intensity? We've all seen complaints and examples of intensity forecasts being vastly wrong (storms forecast to be hurricanes dissipating, and to a lesser extent vice versa).

I doubt that we'll see any significant improvements in intensity forecasts in the next 5 years, maybe 10 years. We won't have much additional data by 2014-2015. The great HWRF has not proven even as accurate as the GFDL. No new models on the horizon. The next generation of GOES satellites (GOES-R) with remote sensing capabilities keeps getting pushed back later and later (2016 now?). So there's no reason to think we'll make any significant improvements in intensity forecasts for many years to come.

[LarryWx]

I'm totally against the 7 day projections going public due to the inherent huge uncertainties/big cone. However, IF they are going to be extended to 7 days, maybe they can somehow leave off the intensity, which is horribly unreliable.

[somethingfunny]

What would be the point?

http://i54.tinypic.com/2jdgu4x.jpg

So uh, Richard is just barely inside this cone five days later.

What's your point? Yes, the 7-day error will be greater than the 5-day error. Some people, not the general public, MUST take action 7 or more days before possible impact.

[mitchell]

I think the 7 day cone is not a good idea. The five day cone is frequently several hundred miles wide and i would think for many storms a 7 day cone would include too much real estate to be useful. It will unnecessarily create hype for areas included in, and then removed from, the cone. There is already enough cynicism when the storm passes hundreds of miles away from areas that were once in the cone. The "cry wolf" effect is real.

Initially, though, the 7-day error cone (circle) would be in the 850-950 mile diameter range - considerably larger than the present 5-day circle. That's big enough to take up most of the Gulf of Mexico. I don't think that most businesses or local governments would actually take any significant actions 7 days prior to impact.

Exactly...

Some people, not the general public, MUST take action 7 or more days before possible impact.

True...and more elaborate actions may have to begin 8, 10, or 14 days before a hurricane hits to avoid consequences, but that doesnt necessarily mean that 8, 10 or 14 day forecast cones can be of much help if they are simply inaccurate.