Modern Weather Forecasts Are Stunningly Accurate (theatlantic.com) 153
An anonymous reader quotes a report from The Atlantic: Meteorologists have never gotten a shiny magazine cover or a brooding Aaron Sorkin film, and the weather-research hub of Norman, Oklahoma, is rarely mentioned in the same breath as Palo Alto. But over the past few decades, scientists have gotten significantly -- even staggeringly -- better at predicting the weather. How much better? "A modern five-day forecast is as accurate as a one-day forecast was in 1980," says a new paper, published last week in the journal Science. "Useful forecasts now reach nine to 10 days into the future." "Modern 72-hour predictions of hurricane tracks are more accurate than 24-hour forecasts were 40 years ago," the authors write. The federal government now predicts storm surge, stream level, and the likelihood of drought. It has also gotten better at talking about its forecasts: As I wrote in 2017, the National Weather Service has dropped professional jargon in favor of clear, direct, and everyday language. "Everybody's improving, and they're improving a lot," says Richard Alley, an author of the paper and a geoscientist at Penn State.
Understanding months-long events like El Niño, for instance, has allowed meteorologists to go beyond the seven-day forecast. Alley, the Penn State professor, says that he is awed by the new models. Well-studied features of Earth's climate -- like the temperate Gulf Stream in the Atlantic Ocean -- emerge in computer models, even though developers have written code that only mimics basic physics. We are now surrounded by the products of these miraculous models. In 2009, a back-of-the-envelope study estimated that U.S. adults check the weather forecast about 300 billion times per year. Perhaps in all that checking we have forgotten how strange the forecast is, how almost supernatural it is that people can describe the weather before it happens. More than 1,000 years ago, the Spanish archbishop Agobard of Lyon argued that no witch could control the weather because only God could understand it. "Man does not know the paths of the clouds, nor their perfect knowledges," he wrote. He cited the Book of Job for authority, which asks: "Dost thou know when God caused the light of his cloud to shine? Dost thou know the balancings of the clouds ?"
Understanding months-long events like El Niño, for instance, has allowed meteorologists to go beyond the seven-day forecast. Alley, the Penn State professor, says that he is awed by the new models. Well-studied features of Earth's climate -- like the temperate Gulf Stream in the Atlantic Ocean -- emerge in computer models, even though developers have written code that only mimics basic physics. We are now surrounded by the products of these miraculous models. In 2009, a back-of-the-envelope study estimated that U.S. adults check the weather forecast about 300 billion times per year. Perhaps in all that checking we have forgotten how strange the forecast is, how almost supernatural it is that people can describe the weather before it happens. More than 1,000 years ago, the Spanish archbishop Agobard of Lyon argued that no witch could control the weather because only God could understand it. "Man does not know the paths of the clouds, nor their perfect knowledges," he wrote. He cited the Book of Job for authority, which asks: "Dost thou know when God caused the light of his cloud to shine? Dost thou know the balancings of the clouds ?"
Excuse me, but "stunningly accurate"? No. (Score:3, Interesting)
"Stunningly accurate" would be a whole different thing, like the forecast being able to tell me "rain will start to fall at my location from 10:34h to 11:27h tomorrow" - and that they very much still cannot.
Re:Excuse me, but "stunningly accurate"? No. (Score:4, Informative)
You must be reading the wrong forecasts. Hourly windows for the next day are pretty accurate these days
Re: (Score:1)
Re: (Score:2)
Re: Excuse me, but "stunningly accurate"? No. (Score:1)
Not where I live. The forecasts generally get things within a few hours, but its not great even then. We were supposed to have 3 to 4 inches of snow last week and we barely got a dusting.
Re: (Score:2)
For the next 12-24 hours, they are generally accurate. Beyond that - they suck. Bounce on to Weather Underground or some other site and look at the 10 day forecast and every day you'll see it change for more than 1-2 days out.
And? We are talking about a 5 day forecast, not a 10 day forecast. We are also talking about matching the 1980 for the 1 day forecast which back then used to change in the mornings too.
Re: Excuse me, but "stunningly accurate"? No. (Score:1, Informative)
Not accurate at all. In fact, go to several different weather websites and look at the *current* temperature. Now check a local thermometer. All different! Can't predict the future if you can't get the present right.
Re: Excuse me, but "stunningly accurate"? No. (Score:1)
Last week in Denver area, snow forecasts ranged from nothing, up to 2 inches, lasting no more than 3 hours. What we got was at least 4 inches, and some places 10, and it lasted 6 hours.
Ya, it's better today than it was 35 years ago. Mostly because we have a lot more, and a lot better, satellite imagery, and processing power is cheap enough to crunch the data faster.
Re: (Score:2)
One of the problems is that whilst weather forecasts are done using ensembles, the reported result is normally just the median or mean result, not the range. So the forecast might have been, if you looked at the raw output, encompassing the weather seen, but the weather seen might have been a relative outlier. But media outlets don't tend to say 'likely 0 to 10 inches of snow, most likely an average of 4 inches, lasting from 0 to 6 hours, most likely 3'. The other issue is 4 to 10 inches might be an area-ad
Re: (Score:2)
Re: (Score:2)
Stereotyping people really is pretty stupid.
Re: (Score:2)
It's not a matter of where you live not knowing how to drive in winter weather. It's a matter you are either an asshole that has no respect for both the conditions and the other people on the road, or completely scared to death. And a
Re:Excuse me, but "stunningly accurate"? No. (Score:5, Interesting)
That depends a lot on where you are. Depending on local geography and prevailing wind patterns, the difficulty of predicting the weather varies hugely. I used to live somewhere that was usually at the intersection of three large weather systems, two coming from the sea and one from land. The actual weather depended on the interaction of the three and so it was pretty common for the forecast for the current day to be wildly inaccurate (as in, predicting sleet in the afternoon on days that turned out to have clear skies and warmish temperatures, or vice versa). Now I live somewhere where most weather systems roll straight over us. Most of the time, you can predict the weather by clanging at a satellite map - assume clouds will follow their current paths all day and you're pretty accurate, do a little bit of curve fitting and you're very accurate.
My biggest complaint about weather forecasts is that they never report their error margins. The weather is a chaotic system, but that's fairly well understood. The Met Office in the UK runs several different models and then picks one result. If all of the models predict the same thing, it's pretty likely. If all of them are predicting different things, then it would be nice to have that information presented so I can see if 'sunny today' means 'we're pretty sure it will be sunny today' or if it means 'it's either going to be sunny or piss it down with rain, depending on what happens when these two fronts collide. Slightly more likely to be sunny, but don't bet on it...'
Just look at this comparison of current forecasts (Score:2)
As you can see, only 3 days into the future the predicted temperatures vary by 4C and the predicted precipitation (tab labeled "Niederschlag")looks almost randomly different between the models.
This is all but "stunningly accurate".
Re: (Score:2)
You must be reading the wrong forecasts. Hourly windows for the next day are pretty accurate these days
So accurate that they need to keep changing them. I'm not sure where the Atlantic got it's data from, but they need to come here to the UK. Met managed to predict there would be snow last week... but couldn't say which day. Weather is easy to predict in temperate climates where the patterns do not change dramatically.
Blame those pesky emacs users (Score:1)
With their C-x M-c M-butterfly [xkcd.com].
Re:Excuse me, but "stunningly accurate"? No. (Score:5, Informative)
Re:Excuse me, but "stunningly accurate"? No. (Score:5, Informative)
Possibly.
On the other hand, the forecast for this past Tuesday (where I live, of course), for each of the five days before that Tuesday, were different. And none of them came especially close (not even the one from Monday night) to what we actually had Tuesday.
Re: (Score:2)
Normally when people look at a specific instance of something in a specific location to try and debunk a general comment the response is "Weather is not climate". Actually I'm at a loss what to say now.
Re:Excuse me, but "stunningly accurate"? No. (Score:4, Insightful)
From what I observe, the weather report is stunningly inaccurate, to the point that I don't even consult it any more except for major trends, and I don't take seriously the idea that something will or won't happen on any given day. I live near Mendocino, CA, and the weather report is all but worthless. Rain pretty much always starts a day earlier or later than is claimed for this location. It was terrible when I lived slightly further inland, in Kelseyville, as well. We had a rise on the property and if I wanted to know what the weather would be even that same day, I had dramatically better results just going up on the little hillock (which, as an aside, is a nascent volcano) and looking in the direction the wind was coming from.
It's actually worse than that, because most of the time they don't even know what the weather is doing RIGHT NOW. They say it's raining, it isn't. They say it's clear, it's raining.
Maybe you get better results inland, when they have all that fancy high-resolution doppler radar to play with, but out here on the northern portion of the left coast the weather report is worthless. Wear layers, and if it even conceivably might rain, bring one that's waterproof.
Re: Excuse me, but "stunningly accurate"? No. (Score:1)
Getting back to your first pint. It really is about the quality of forecast specifically in relation to what the average person really cares about. The average person cares about their journey, whether it is in a minivan or a sports car and they want to travel worry free. It is a huge plus when the weather turns out to be beautiful but is never necessary. It is always about real lives of regular people. Now take that idea and drop some weather stations in the correct locations with all that in mind.
Re: (Score:3)
Now take that idea and drop some weather stations in the correct locations with all that in mind.
Yes, this is something that drives me nuts. What's it cost to build a weather station with a cellular module in it to send back data occasionally? Why don't we have literally millions more small weather stations sprinkled all over the nation? What year is it?
Re: (Score:2)
It's 2019.
Re: Excuse me, but "stunningly accurate"? No. (Score:4, Informative)
I'm guessing they have this already. Weather Bug (.com) has cameras everywhere for example; so I'd expect they have sensors too (which are out of our sight, out of mind).
They do not. Most of these companies have zero of their own weather stations. They depend on sensors at airports, and home weather stations people have tied into their services.
Re: (Score:2)
Re: (Score:1)
What year is it?
1812... They're burnin' down da house!
Imagine, if you will, a boiling mud pit. Can you predict when and where the next bubble will pop? Random events like that are entirely impossible to predict, but with the cold fronts and hurricanes they're spot on. And when you use the NWS you get just the facts without the hype.
Re: (Score:2)
Re: (Score:2)
Re:Excuse me, but "stunningly accurate"? No. (Score:4, Insightful)
There are a lot of microclimates around Mendocino. Maybe the weather report just needs to be more granular?
There's a ten-mile stretch of Hwy 101 down here where you can go from dense fog and cool to blazing blue skies and warm and then back again twice.
Re: (Score:2)
There are a lot of microclimates around Mendocino. Maybe the weather report just needs to be more granular?
There are a lot of microclimates everywhere that's not flat AF. But I'm talking not just about small showers, but about the reporting on weather systems that cover the entire region. They don't know when or even if they will arrive until hours before it happens.
Re:A good Matlab replacement, not the next big thi (Score:2)
The issue you may be having is that the forecast is accurate but the news aggregator that is presenting it to you doesn't understand it and misrepresents it. In particular the 'x% probability of rain' element is widely misunderstood. '40% chance of rain' means 40% chance of raining at any point during the given period within the catchment area, and not, as popularly interpreted, including by news papers, as '40% of it raining at the moment the given time period specified in all locations in the catchment ar
Re: (Score:2)
Re: (Score:2)
For completeness, it might mean that the chance of it raining in any given hour of the time period, including the one at the end of the period, might be as low as, say, 5%, even though rain on that day at some point is fairly likely. It's relatively counterintuitive for most people, even when newspapers report it correctly, and they might want to consider changing the system.
E.g. if any hour rain has a totally independent chance of 5%, the chance of rain during 24 hours is 1-(0.95)^24, or 70%. Of course, if
They USED to do it that way... (Score:4, Interesting)
"Cold front will arrive next Tuesday morning and lows will be in the teens and highs in the 30s by Thursday afternoon."
That's funny that you put it that way, because that's how they USED (~ 1980s) to predict the temperature: in the lower 70s, mid-20s, etc. Nowadays, all the TV stations here in Chicago give impossibly exact numbers: 32 for the high, 14 for the low, etc.
Re: (Score:2)
I find that for up to 48 hours they are accurate. And you need to look at the hourly forecast. After that, predicted weather tends to lag by up to 24 hours after you get beyond 5 days, either the storm comes within 24 hours early or 24 hours late.
Also, a lot fo people rely on the daily summary. For example, if they see a sun with clouds they think that it's going to be nice all day, but if you dig into the hourly, you find that a band of thunder showers are predicated for about an hour in the mid afterno
Re: (Score:2)
Re: (Score:1)
"Stunningly accurate" would be a whole different thing, like the forecast being able to tell me "rain will start to fall at my location from 10:34h to 11:27h tomorrow" - and that they very much still cannot.
I don't know. Weather Underground does exactly that. Sure, it's not even close to 100% accurate but quite often it's correct within an hour even several days to a month out. I'd say it more accurate than not.
Way better than the old days anyway.
Re: (Score:1, Informative)
And that's why you have no clue about climate change either; you're incapable of telling the difference between your own, highly local, experiences and larger trends.
Forecasts are getting better and better, that doesn't mean that they aren't "fuzzy around the edges". For instance, there might very well be certain locations where the model isn't entirely accurate or reliable; this however doesn't mean the forecast is inaccurate, it means you're an edge-case.
Re: (Score:2)
...and people STILL complain.
Re: (Score:1)
Not sure what's wrong with your weather forecasts, but where I live we use an app that will tell you to the minute when it will start to rain and when it will stop.
People use it to decide if they should leave now or wait 15 minutes before cycling home.
So yeah, maybe you need to see why your local weather people aren't doing this.
Re: (Score:2)
How cloudy? When? I can tolerate a good bit and be fine, but when matters. If off-peak hours, don't care.
I may care if it's all sun in the AM and then dark - going to batteries too soon, or
Re: (Score:2)
One thing to keep in mind when looking at the spectacular new "Geocolor" imagery from NOAA's latest satellites -- it's far less "real" than it appears to be at casual first glance.
The cloud imagery is real & better than anything we've ever had, but contrary to appearances, GOES-17 is NOT a full-color webcam in space. The beautiful blue oceans, lush green terrain, brown deserts, majestic white snow-capped mountains, and glowing city street lights are all computer-generated images with the cloud data over
Re: (Score:2)
Here, where I manage a rather large solar photovoltaic system for a decent sized campus, an IR pic is often plenty good enough if it's timely, and I really don't care what the ground looks like on their pic anyway - I can always look out the window and get a look with fantastic resolution and color rendition after all...
Re: (Score:2)
0% might be correct. They talk about percentages it's a combination of "probability" and "area" (and maybe length?). So if there are few enough weather stations around you, and the rain is well localized enough, then the completely accurate probability might be 0.4%. Rounded down of course. Although I think they tend to round to the fives, so it could even be 2.4%
Really, the problem is they call it "probability" but it doesn't mean "the probability of any precipitation in the area", it means "the probab
Look at pollen as well (Score:2)
Re: (Score:2)
"Stunningly accurate" would be a whole different thing, like the forecast being able to tell me "rain will start to fall at my location from 10:34h to 11:27h tomorrow"
Oh? Interesting comment. Personally I have found the exact GPS location based astronomy apps which provide cloud cover information, including a breakdown of the type and height of clouds, precipitation type, precipitation chance, possible max precipitation level, as well as wind speed, direction, relative humidity, and temperature all with an accuracy of +/-minutes to be "stunningly accurate".
Re: (Score:2)
Personally I have found the exact GPS location based astronomy apps which provide cloud cover information, including a breakdown of the type and height of clouds, precipitation type, precipitation chance, possible max precipitation level, as well as wind speed, direction, relative humidity, and temperature all with an accuracy of +/-minutes to be "stunningly accurate".
Are you saying astronomers secretly produce better forecasts than meteorologists? Then let us know their secret.
Re: (Score:2)
No I'm saying you're picking either a crappy source to get your weather forecast, or the source of your weather forecast is seriously deficient in your local area.
"Stunningly accurate" is definitely how I describe modern weather forecasts.
Re: Excuse me, but "stunningly accurate"? No. (Score:2)
"stunningly accurate" is the typical lingvo of Science and Nature. They must to live up to expectation of being leading general science journals.
Basically it's journalism. That's what they became. Good journalism of the past, like BBC or WaPo, became tabloidal. I do not even want to think what tabloids have become.
yr.no for the win! (Score:5, Interesting)
Here in Norway we have learned to depend upon https://yr.no/ [yr.no] which provides both short-term (2+ days) and long-term forecasts:
When the short-term forecast states that it will be 0.5 to 0.8 mm rain (or snow equivalent) between 10:00 and 11:00 tomorrow, and that it will clear up starting at 13:00, this is very likely to be correct. If it isn't exactly right it is usually because the changes happen a little bit before or after the maximum likelihood prediction.
The presentation of the weather data is so good that many people in our neighboring countries have started to use YR instead of their local weather service.
Terje
YR.no user here ... (Score:2)
I am not in countries neighbouring Norway. In fact, I am an ocean away, in Canada, but I do use YR.no.
First, it powers the XFCE Weather Plugin. Some of us still use Linux for the desktop (remember that?), unlike many on Slashdot.
Second, the web site loads fast, and does not use AJAX and other abominable practices (as opposed to Canada's TWN and Wunderground).
Third, the UI is simple and to the point.
Fourth, they have a API that can be used from many applications (one is the XFCE Weather plugin), including Ho
Re: (Score:2)
Fifth, they provide a cloud forecast, which is useful for astronomy.
Interesting, I didn't know this service. For astronomy forecasts I cross-check a couple of services, but my favourite is 7timer.info [7timer.info], which, apart from cloud cover, gives you transparency and does an astro-seeing (atmospheric turbulence) forecast. In fact, its only issue was its unreliable server, which is why I donated a reliable server to the project, and made a free iOS client (Xasteria [ecuadors.net]).
Re: (Score:2)
That 7timer.info looks good. Adding it to my bookmarks.
There is also CalSky.com, in their Meteo section. They have seeing.
ClearOutside.com shows cloud, but not seeing.
We have ClearDarkSky.com, which does seeing too, but it is probably North America only.
Correlations Should Be Published (Score:5, Insightful)
I wish every weather service published a graph that showed the progress of the correlation between their 1 to 7 day forecasts and what actually happened, somewhat like the graph [sciencemag.org] for hurricane tracks in the referenced article. Published confidence levels would also help to know how locked-in a prediction was.
My experience has been that forecasts a day or two ahead are amazingly accurate, but that you can't rely on forecasts a week out for scheduling an important event.
Re: (Score:2)
Re: (Score:2)
Too many ideas... not enough time.
Re: (Score:2)
Location granularity (Score:5, Informative)
A big part of this is that forecasts and current conditions have vastly smaller location granularity than in the past. In 1980, for a given state, you'd be lucky to obtain specific forecasts for maybe 10-15 large cities in the entire state (less for smaller states). I'm sure most of you have seen weather where it rained at your house, but just a few blocks down the street they didn't get rain at all. When your forecast granularity is representative of hundreds of square miles, then of course you can never be very accurate for that entire area.
Now the forecast is latitude and longitude based, and the precision is vastly finer. That alone increases the accuracy tremendously. Weather forecasts now are also down to "minutely" (as in hourly or daily) time spans. Again, same thing. When your forecast broke the entire day into "night" and "day" periods, you can never be very accurate. Most weather apps now forecast what will happen in the next hour down to the minute ("Light rain will begin in around 12 minutes"). It's easy to be accurate when you can forecast such a small time into the future.
There are many reasons weather is more accurate now, everything from the lead time (if your forecast has to be in to the newspaper before 5 AM so it can meet the press deadline, then you're accuracy will be reduced compared to a forecast calculated the moment it is asked for), to the technology that allows people to ask for and view data when they want it for a very specific area.
Re:Location granularity (Score:5, Interesting)
More than just lead time, increased processing power and bandwidth allows weather agencies to run the models more frequently. 10 years ago in Canada, the model only ran every 12 hours. Now they're routinely running it every 6. So short term model-based forecasts are using fresher observation data, which makes a huge difference in prediction quality since forecasts react quicker to unexpected changes. There's now less of that "they said we'd only get 2cm of snow when I went to bed, and I woke up to 10cm".
Another change that happened a while back (~25 years) is they stopped letting meteorologists mess with the longer range forecasts. They found that in terms of quality, the probability of a human improving on the model beyond 2-3 days was only 0.5 (i.e. half the time, they'd make the forecast better, and half the time worse), and the models have only gotten better since then. So they've focused human intervention on the short range high impact stuff (0-18 hours, mostly) and left the longer term predictions to the computers.
Re:Location granularity -ahem data is better too. (Score:2)
Re: (Score:3)
Not surprising (Score:2)
We have a lot more data and we have a lot better understanding of the correlation betweeen past and present conditions around the world and future contitions at any given spot than we did 40 years ago.
Yes, there has been a huge improvement.
No, I am not surprised.
Except for when their not? (Score:2)
Example: last week when the Twin Cities was predicted to get 8-12 inches of snow and got 3-4.
Re: Except for when their not? (Score:1)
They predicted substantial snow, you got it. It was valid decisional information.
Also, the prediction wasnâ(TM)t for your house in the Twin Cities, but a prediction for the aggregate area. Thereâ(TM)s often wide variation within a geograpical area. In the Washington Metro area I live in, for example, the prediction was 1-2â of snow. Many neighborhoods within the region got that much, some got less, some (where I live) got substantially more. My âhood typically gets double what the region
Watch TPT (Score:2)
On an aggregate we've come a long way. Not news... The more detailed the harder it is. You can improve a skill overall but make imperceptible improvements at components of it or even go backwards at them.
The models are are more detailed now and they get results faster so they can do better now than in the past; as well as constantly update with new data but they do not yet model all the local impacts or report it to specific locations. 100% accurate will always be wrong for people physically located betwee
Re: (Score:2)
Will I need an umbrella tomorrow? I'm going to be in either New York or Denver.
As we demand forecasts become more *accurate* and *temporally* precise, we also need them to become more *geographically precise*. If you want to know whether you'll get rained on at exactly 10AM tomorrow morning, you have to use the forecast for the exact place you will be, not some place twenty miles away. If you look at high precision radar map it could be pouring in one place and completely dry ten miles away. As you inte
Nonsense. (Score:2)
I traveled with my family for a week in Paris last October. The forecast basically said it will be rainy all week and changed to sunny all week (which turned out to be correct) just one day ahead of our arrival. Not exactly what I'd call "stunningly accurate".
Why Not 11 Day Forecasts? (Score:2)
So the longer range forecasts are more accurate than ever. Why cut off at 10 days, then, and not 11? Sure, 10 is a very human number, but there's no scientific reason to assume it's a natural cutoff in reliability. Why not use the current methods for a 10-day forecast and extend it out to 100 days. Then study the accuracy and see if there's a sudden drop-off, and cut off the published forecasts just short of that. It would also be interesting if they would publish a confidence rating on the forecasts,
Bollocks (Score:2)
I am stunned of how wrong they get it. Sometimes they can't even agree with the current data, like "we forecast a maximum of 22C for today, current temperature 23.5". I'm always thinking of the first part from "Two Man in a Boat" when they get it (once again) spectacularly wrong.
Lots of hearsay and beliefs (Score:3)
I see lots of people claiming they know better than the study but it doesnâ(TM)t seem to be based on any rigorous research.
Re:Lots of hearsay and beliefs (Score:5, Insightful)
People just have to complain. It's their nature. And they use the "Smoking kills you!" "That is not true because my granfather smoked like a chimney and he died at 92" argument where N=1.
Re: (Score:2)
People just have to complain. It's their nature. And they use the "Smoking kills you!" "That is not true because my granfather smoked like a chimney and he died at 92" argument where N=1.
He smoked, he died. Ergo, smoking kills you.
I also recall an ad saying "90% of smokers die". It might be a good idea to start smoking now, in case you're in the lucky 10% that will live forever.
Re: (Score:2)
Didn't Pratchett make a joke like that in one of the Discworld books?
Re: (Score:2)
I see lots of people claiming they know better than the study but it doesnâ(TM)t seem to be based on any rigorous research.
There's a lot of confirmation bias at work. People who enjoy grumbling about the inaccuracy of weather forecasts regularly notice every deviation but mostly ignore the correct predictions, and therefore see proof that forecasts are extremely inaccurate.
Anecdotal experience is basically useless in something like this. You'll see whatever you want to. To know anything real, you need to actually record forecasts and variance for a good period of time, and then systematically analyze the results.
Most of it is CPU power, not model improvement (Score:3)
The issue with complex modelling has traditionally been in computational power vs usefulness. With weather modelling, you can generate a very accurate prediction a week ahead. Problem with it was in 1980, you needed so much computational power, you would get this "forecast" finally calculated a few years after it was relevant. Computational power to do it in time frame that was useful was simply not there.
Today it is there, so meteorologists can get calculations for days ahead done in time frame which is useful, i.e. before the events they're trying to predict occur, rather than long after.
Models improved too, but most of that improvement has actually been "add even more detail to do something with all the increasing computational power".
Re: (Score:2)
It hasn't been "vastly improved". The primary "improvement" comes from the simple fact that model has been made more accurate by introducing more heavy calculations on error margins. This is not improvement - it's a trade-off. You trade the efficiency of the model for accuracy of the model.
And in modern world with supercomputers we have today, it's a worthwhile trade-off. Which doesn't make it any less of a trade-off rather than a straight up improvement. If you went to 1980s computational power, current mo
Snowzilla predicted 8 days out (Score:3)
In 2016, in the mid-Atlantic region, we had a mild winter. There was one snow event only - but that event dumped about 33-36 inches in the Baltimore/DC metro region. That event was informally called Snowzilla, and it was predicted to the tee, 8 days out. That was probably the most amazing forecast achievement I've seen.
Of course, the meteorologists still screw it up, sometimes fantastically. But other times, they knock it out of the park.
Define stunningly accurate (Score:2, Interesting)
Weather forecasting hasn't got better, at all.
My Gran, rest her soul, was always far better at weather than any bloody meteorologist and that over her life of 95 years. Was she perfect? No, but a lot more use than any bloody meteorologist.
Add computers and satellites and they suddenly think they're all bloody geniuses.I'd actually say they're worse or the same. Probably get better results reading entrails.
"Stunningly accurate" (Score:2)
At 3am the morning before, forecasters calling for an inch of snow.
Later that same day: A foot of snow.
Or what about the snow on no forecast that suddenly arrives one day.
Forecasts are better to be sure but to call them "stunningly accurate" is laughable.
Most common forecast: (Score:1)
"Late night and early morning low clouds and fog"
Anybody wanna tell me where you hear that every day?
However I concur with the article. Weather, especially hurricane forecasting is top notch. It would be a real shame if anything happened to it on the whim of some crazy politician.
Strange (Score:2)
Here in Montreal I have the feeling the weather predictions were far more accurate in the 1990s and 2000s. Lately they have been consistently wrong summer and winter. They keep forecasting major snowfalls and nothing happens.
Slightly offtopic. (Score:2)
More than 1,000 years ago, the Spanish archbishop Agobard of Lyon argued that no witch could control the weather because only God could understand it.
Some people complain Harry Potter is evil because of "magic". Same complaint for other shows.
Yet they go home, wave their hands, and suddenly night turns to day (hit the light power switch.)
Their house is somehow cool in the Dog Days of summer. (AC)
They open a door, pull out cold food that was cut/killed weeks ago, and it's still good. They put it in a different box, and 2 minutes later it's hot. (fridge, microwave.)
They handle a small box and suddenly demon voices imitating their friends are respond
Compute power (Score:4, Interesting)
I spoke to a guy at the UK met office about this in the 90s and he explained how they were basically compute limited. They run a number of sims with randomized perturbations at the start and see which outcomes are the most common across perturbations. They were using all their Crays full whack and that's what determined and limited the accuracy of the results. 30 years later, compute power is somewhat cheaper and my desktop is faster than one of those Crays.
Re: (Score:1)
So they still run their Crays at full whack. Your desktop may be faster that Crays from 30 years ago but probably not the Crays UKMET are using today:
https://www.hpcwire.com/2018/09/26/uk-met-office-deploys-cray-ai-analytics-to-enhance-forecasting/
Re: (Score:2)
That was my point. Didn't you read it?
Re: (Score:2)
That was my point. Didn't you read it?
Welcome to Slashdot...
Not really (Score:2)
My wife and I spend a lot of weekends hiking. I watch weather forecasts to determine plans (including camping plans) days or even a week in advance. Is it going to be sunny or heavy rain? Modern weather forecasts are better than flipping a coin, but not by much. I usually end up looking at weather radar a day in advance (for day trips) and make my own forecast.