AccuWeather says it has created its own hurricane scale, but why?
- Thread starter JournalBot
- Start date
Data and money are the real issue. IBM has done the same thing with The Weather Company and its wholesale decimation of Weather Underground, killing off the public API with no replacement being the latest example of their terribleness.
AccuWeather wants free data and then be able to license their (perhaps) better scale in order to generate revenue. While I don't begrudge a company trying to make money, this path seems exceptionally crappy because of the real need to have clear and accurate forecasts in the hands of individuals.
The scientific competition between forecasting models is fine; a commercial competition to see who can get the most eyeballs (like HD-DVD and BluRay) when it comes to something as critical as hurricanes is not good.
AccuWeather wants free data and then be able to license their (perhaps) better scale in order to generate revenue. While I don't begrudge a company trying to make money, this path seems exceptionally crappy because of the real need to have clear and accurate forecasts in the hands of individuals.
The scientific competition between forecasting models is fine; a commercial competition to see who can get the most eyeballs (like HD-DVD and BluRay) when it comes to something as critical as hurricanes is not good.
Upvote
191
(192
/
-1)
Myers being put in charge of NOAA is a scary thought. As much as we are headed to a world where only those who can pay can access quality news/reporting, it should terrify everyone to their soul to have critical weather alerts privatized.
It is rent-seeking almost at its purest form and it should be stopped at all costs.
It is rent-seeking almost at its purest form and it should be stopped at all costs.
Upvote
161
(163
/
-2)
Like all Republican efforts, socialize the costs, privatize the profits (aka I hate the government until it can make me money). Also, the "AccuWeather RealImpact Scale for Hurricanes" honestly sounds like something out of Idiocracy, it makes me vomit a little in my mouth every time I read about it.
Upvote
134
(148
/
-14)
1-5 doesn't give enough granularity on the severity of a storm anymore.
Either go 1-10 or give multiple 1-5 ratings i.e. Hurricane Todd is a Category 2 for wind, a category 4 for storm surge and a category 4 for size.
Either go 1-10 or give multiple 1-5 ratings i.e. Hurricane Todd is a Category 2 for wind, a category 4 for storm surge and a category 4 for size.
Upvote
101
(106
/
-5)
Given that Saffir-Simpson is accused of being one-dimensional, I'm sure that RealImpact will gain it's edge using imaginary numbers.
Cough.
Cough.
Upvote
59
(62
/
-3)
This is a troubling trend.
I'm concerned that part of the incessant hammering by some interested parties in the US about taxes being too high and government too big is actually a concerted effort to privatize profits and socialize the costs of data and other services that rightfully belong to the public.
Eric had mentioned Michael Lewis' The Coming Storm, but I'd also mention another one of his books, The Fifth Risk, where this trend in privatization of weather data is also covered.
I'm concerned that part of the incessant hammering by some interested parties in the US about taxes being too high and government too big is actually a concerted effort to privatize profits and socialize the costs of data and other services that rightfully belong to the public.
Eric had mentioned Michael Lewis' The Coming Storm, but I'd also mention another one of his books, The Fifth Risk, where this trend in privatization of weather data is also covered.
Upvote
82
(85
/
-3)
If you want to understand why this is terrible, read "The Fifth Risk" by Michael Lewis. It provides many painful details of Barry Myers' machinations to undermine government weather reporting. Accuweather is a private company which has a conflict of interest with free and open data. Having this person in charge of NOAA defies comprehension.
Link:
https://www.npr.org/2018/10/02/65256390 ... interested
Link:
https://www.npr.org/2018/10/02/65256390 ... interested
Upvote
105
(107
/
-2)
I'm really torn on this. On one hand, the SSHWS is very flawed, and cases like Ike happen more often than we'd like them to. On the other, though, the current scale is simple and easy to understand, something NOAA is always looking for in forecast products. The lack of transparency from AccuWeather regarding their scale is worrying to me also, especially given NOAA is the world's only source of free-to-access global model data (the GFS).
Edit for clarity, another edit for what I actually mean in the last sentence.
Edit for clarity, another edit for what I actually mean in the last sentence.
Upvote
8
(12
/
-4)
that's fucking hilarious
input (feels + public weather data) -> [algorithm] -> output ($$$ to myers family)
Upvote
60
(62
/
-2)
I wish there was a way to inflict punitive damages against entities so mind boggling stupid/corrupt/corrupt.
AccuWeather wouldn't be the first company I punitive damaged out of existence, but it would definitely be gone before the end of the first day.
AccuWeather wouldn't be the first company I punitive damaged out of existence, but it would definitely be gone before the end of the first day.
Upvote
20
(23
/
-3)
I must push back on this notion to an exceptional degree.
Adding more numbers to your scale does not necessarily impute additional accuracy or precision. Just as likely, you are implying you have more confidence in the forecast than is actually appropriate. This is not only wrong, it is also very dangerous. You don't want to give the public the impression that you have an overly firm view of the ceiling of a storm's potential -- this was, after all, the core failing of the current scale during Harvey.
Recall the absurdity of late 1990s computer games being reviewed and ranked a 93 or 94 on a 100-point scale. Was there meaningful difference between a score of 93 or a score of 94? Of course not.
Upvote
71
(72
/
-1)
What use is a government of the people if certain people cannot steal from it? Getting all of your data free from the public trough, and then being the only person to be allowed to broadcast analyses of it, sounds exactly like the "for the people" that Lincoln was referring to, but maybe it is perishing from this earth. That the last slime bucket to lead this charge was arch hypocrite Santorum can come as a surprise to few, but the irony of this coming out of the home state of Gettysburg is icing on the cake.
Upvote
19
(21
/
-2)
AccuWeather's apprehension to release the algorithms should be a deal killer instantly. The entire reason it's using the "1 to 5" scale is to literally introduce confusion to both the government and public, as changing the current system would cost more money to implement.
NOAA should provide the data necessary for any weather agency to build models from, but to allow any company to pull the data and declare a new scale (without approval) shouldn't even be a topic of discussion.
I have no problem if AccuWeather has a model in addition to the current (regardless of state), but to replace it completely is stupid beyond words.
Also, call me cynical we won't see this in our future:
"AccuWeather's RealImpact Scale is brought to you by Bristol-Meyers, when seasonal allergies increase due to the rise of pollen in the air..."
*shivers*
NOAA should provide the data necessary for any weather agency to build models from, but to allow any company to pull the data and declare a new scale (without approval) shouldn't even be a topic of discussion.
I have no problem if AccuWeather has a model in addition to the current (regardless of state), but to replace it completely is stupid beyond words.
Also, call me cynical we won't see this in our future:
"AccuWeather's RealImpact Scale is brought to you by Bristol-Meyers, when seasonal allergies increase due to the rise of pollen in the air..."
*shivers*
Upvote
48
(49
/
-1)
Saffir-Simpson is perfectly fine as long as people know the area in which they live. When I lived on the Atlantic coast of Florida, my rule of thumb was the one referenced in the article - Cat 2 or lower wasn't significant, Cat 3 or above was. That's because the coastal shelf in my area limited storm surge, so wind speed was the biggest risk. Where I am now, the area's prone to flash flooding, so any storm with high rainfall would be time for evacuation, because the higher risk is flooding rather than wind speed.
As far as the AccuWeather announcement, without access to their decision-making system (and preferably their algorithms), it's worthless at best and a potentially deceptive risk to life at worst. Disaster planning is not the place for proprietary black boxes.
As far as the AccuWeather announcement, without access to their decision-making system (and preferably their algorithms), it's worthless at best and a potentially deceptive risk to life at worst. Disaster planning is not the place for proprietary black boxes.
Upvote
26
(27
/
-1)
It's really no different than big pharma using and patenting public research on drugs.
"Overall, we find that direct government funding is more important in the development of "priority-review" drugs-sometimes described as the most innovative new drugs-than it is for "standard-review" drugs. Government funding has played an indirect role-for example, by funding basic underlying research that is built on in the drug discovery process-in almost half of the drugs approved and in almost two-thirds of priority-review drugs."
https://www.healthaffairs.org/doi/10.13 ... .2009.0917
Upvote
20
(20
/
0)
This is exactly what I was going to post, no transparency...no acceptance or adoption.
Upvote
24
(24
/
0)
I understand you, but the video game review problem only becomes a thing if the people issuing the warning refuse to use the whole scale in which case we are right back where we started.
I prefer the multiple ratings, 1 for wind/surge/etc, cause it gives a better idea of the storm as a whole, but then I suspect you have people who claim that 3 numbers is too many for the common folk to understand. (This is not a shot at you by the way)
I prefer the multiple ratings, 1 for wind/surge/etc, cause it gives a better idea of the storm as a whole, but then I suspect you have people who claim that 3 numbers is too many for the common folk to understand. (This is not a shot at you by the way)
Upvote
6
(6
/
0)
As someone who lives on the gulf coast....this is equally worthless. A simple number cannot tell people how dangerous a storm is.
Instead, they should focus on a scale that has multiple vectors and imparts as much of the raw data as possible while still being simple to digest.
For instance, it could be something like <wind-speed>/<storm-surge>/<rainfall>. So a 2/5/5 would be quite dangerous for those on the beach and those near rivers, but those not in flood plain, and not subject to storm surge would be able to ride it out with good preparation. One number is just insufficient and nothing's going to change that, because the severity of a storm depends greatly on the topography of the land around you.
They should also continue to make the rainfall & flooding predictions more and more prevalent as city planners/builders never learn from mistakes, and flooding downstream from cities gets worse with every storm that hits.
Instead, they should focus on a scale that has multiple vectors and imparts as much of the raw data as possible while still being simple to digest.
For instance, it could be something like <wind-speed>/<storm-surge>/<rainfall>. So a 2/5/5 would be quite dangerous for those on the beach and those near rivers, but those not in flood plain, and not subject to storm surge would be able to ride it out with good preparation. One number is just insufficient and nothing's going to change that, because the severity of a storm depends greatly on the topography of the land around you.
They should also continue to make the rainfall & flooding predictions more and more prevalent as city planners/builders never learn from mistakes, and flooding downstream from cities gets worse with every storm that hits.
Upvote
16
(16
/
0)
Also worth noting that Accuweather supports climate change denial, even though in recent years they've gained a more subtle approach.
Upvote
29
(32
/
-3)
This is just going to cause confusion. Only reason AccuWeather is doing this is for data and money as others have suggested.
Upvote
19
(19
/
0)
Edited for what I actually meant by that, apologies. I'm in a hurry this morning.
Upvote
4
(4
/
0)
There seems to be no end to these corrupt or at least venally self-interested republican nominees. I continue to be stunned that the entire republican political class just goes along with it. Thank God the NASA leader doesn't seem to be doing that.
Upvote
1
(5
/
-4)
Side note: Unless I missed an article somewhere the prediction models / data have been put on hold because of the gov shutdown. I'd like to get ARS's take on how that impacts....well everything as we depend a lot of those models.
Here's an article from 2013. And I can't imagine the need has decreased since then.
https://weather.com/science/news/govern ... r-20131010
It has to be worse as we are now in the longest shutdown in US history.
Here's an article from 2013. And I can't imagine the need has decreased since then.
https://weather.com/science/news/govern ... r-20131010
It has to be worse as we are now in the longest shutdown in US history.
Upvote
8
(8
/
0)
Public safety needs to be a transparent government service, not an opaque, private, money-making enterprise. No good can come from this.
Upvote
26
(27
/
-1)
Statistical
Ars Legatus Legionis
#OhFuckNo
Public safety and disclosure should not be based on proprietary for profit algorithms. Eric is being diplomatic. No doubt it is going to be overhyped and that will desensitize people and they will avoid evacuation orders or even taking basic precautions because the storm is "only" a 3 on the Accuweather hype meter and the last two "3s" were duds.
Upvote
26
(28
/
-2)
This is the thing. It's not that they came up with their own scale. It's that they won't provide the algorithm upon which the scale is based. Without that, the scale number is essentially meaningless.
A wink-wink, nudge-nudge, "Trust us" mentality isn't going to cut it.
Upvote
38
(38
/
0)
"but why?"
Barry Lee Myers is a whackjob and a climate denier, just for starters.
Trump wanted him to head NOAA - that should tell you everything you need to know about Myers.
Don't give AccuWeather your clicks - Weather Underground is much better anyways.
Barry Lee Myers is a whackjob and a climate denier, just for starters.
Trump wanted him to head NOAA - that should tell you everything you need to know about Myers.
Don't give AccuWeather your clicks - Weather Underground is much better anyways.
Upvote
35
(36
/
-1)
Saffir-Simpson Hurricane is a very general threat estimate for a region. Today our ability to pinpoint with greater accuracy high threat areas based on multiple factors such as rainfall, flood potential, storm surge vulnerability, etc. are such that a regional scaled system is inadequate. A regional assessment simply does not have the granularity necessary for people to know if they must evacuate or not. Evacuating everyone en mass regardless of whether or not they would be affected by a storm creates logistical problems that complicates that same evacuation and subsequent recovery.
Upvote
4
(4
/
0)
Yeah, with all the talk for the past several years on the various impacts beyond the classing 1D scale, I'd think that a multi-dimensional scale would be far more useful weighing impact since the public has been primed. It would require some attention to the local conditions that newer residents might not quite have fully learned, but that's something that local authorities and media can help interpret.
Upvote
6
(6
/
0)
As terrible as weather.com has turned in recent years, AccuWeather is worse.
Upvote
24
(24
/
0)
https://www.wunderground.com/ works pretty well. It's goddamn bloated though.
Upvote
1
(3
/
-2)
Brought to you by the same company who uses their *patented* ‘RealFeel’ value as an alternative to the wind chill equation. Curious how they even managed to get a patent on this?
Make the equation patent free and a public good, otherwise it provides little value beyond marketing. If I can’t use the scale freely, then it is just a waste of time, since it limits our ability to talk about something that has a universal definition. It also diminishes its ability to be used in a scientific context.
Edit: typing fixes
Make the equation patent free and a public good, otherwise it provides little value beyond marketing. If I can’t use the scale freely, then it is just a waste of time, since it limits our ability to talk about something that has a universal definition. It also diminishes its ability to be used in a scientific context.
Edit: typing fixes
Upvote
18
(21
/
-3)