Arthur T Knackerbracket has found the following story:
Last month, as much of the United States shivered in Arctic cold, weather models predicted a seemingly implausible surge of balmy, springlike warmth. A week later, that unlikely forecast came true—testimony to the remarkable march of such models. Since the 1980s, they’ve added a new day of predictive power with each new decade. Today, the best forecasts run out to 10 days with real skill, leading meteorologists to wonder just how much further they can push useful forecasts.
A new study suggests a humbling answer: another 4 or 5 days. In the regions of the world where most people live, the midlatitudes, “2 weeks is about right. It’s as close to be the ultimate limit as we can demonstrate,” says Fuqing Zhang, a meteorologist at Pennsylvania State University in State College who led the work, accepted for publication in the Journal of the Atmospheric Sciences.
Forecasters must contend with the atmosphere’s turbulent flows, which nest and build on each other as they create clouds, power storms, and push forward cold fronts. A tiny disruption in one layer of turbulence can quickly snowball, infecting the next with its error. A 1969 paper by Massachusetts Institute of Technology mathematician and meteorologist Edward Lorenz introduced this dynamic, later dubbed the “butterfly effect.” His research showed that two nearly identical atmospheric models can diverge widely after 2 weeks because of an initial disturbance as minute as a butterfly flapping its wings.
-- submitted from IRC
(Score: -1, Offtopic) by Anonymous Coward on Tuesday February 19 2019, @07:36AM
Ask the Hippy Dippy Weatherman. He's pretty far out...
(Score: 2, Informative) by Anonymous Coward on Tuesday February 19 2019, @10:38AM (12 children)
Just to pre-emt those inevitable comments saying "see, we can't predict weather for merely a month, so how could we predict climate over decades", let me make it immediately clear that climate is not the same as weather.
I can't tell if in Washington, Jan 10, 2020 will be colder or warmer than Jul 13, 2020. But I'm willing to bet that on average in the US, winter 2019/2020 will be colder than summer 2020.
(Score: 1, Insightful) by Anonymous Coward on Tuesday February 19 2019, @11:50AM (5 children)
So you are willing to bet on imaginary number like average temperature of an entire country that probably includes people experiencing a range from -20 C to +20 C at any given time? Average temperature of an entire country the size of the US doesn't mean anything.
Now, perhaps you actually want to be measuring something in Joules? Total energy content of the first 2 meters of atmosphere or something? That would be a real number.
(Score: 3, Touché) by khallow on Tuesday February 19 2019, @01:55PM (4 children)
What's imaginary about it? Integration is a well-known thing.
(Score: 1, Insightful) by Anonymous Coward on Tuesday February 19 2019, @04:38PM
You can plug any numbers into an equation, doesn't mean the results have physical meaning.
(Score: 0) by Anonymous Coward on Tuesday February 19 2019, @04:46PM (2 children)
If I measure the energy content of two regions A and B, I can add A + B to get "total energy". Can I do that for temperature?
(Score: 2) by fritsd on Tuesday February 19 2019, @05:09PM
Probably [wikipedia.org], but I don't remember how to do it.
Eventually the black body radiation of the two bodies will have warmed up the one and cooled down the other to equilibrium.
(Score: 1) by khallow on Tuesday February 19 2019, @07:01PM
That's not even wrong territory. The existence of addition doesn't tell you anything about whether the values in question are real or imaginary in an existence sense. But if we were to add average temperatures in a sensible way, the way to do it would be the average temperature of A times the area of A plus the average temperature of B times the area of B divided by the total area of A and B.
(Score: 2, Insightful) by Anonymous Coward on Tuesday February 19 2019, @01:20PM (1 child)
Weather and climate are not the same, but not completely different either.
The weather models predict that it is likely to rain and generally where and when. This means predicting averages over the next few days. Not exactly when and where for each drop.
The article points out the Butterfly effect where random variation limits how far in the future a prediction can go. Given initial knowledge of the state of the system the models can only predict so far in the future.
That limit aside, there seem to be restoring forces in the weather which pull the random variations back to statistically measurable long term weather patterns.
Perhaps these long term weather patterns are the climate.
If they are predictable remains to be seen. In the last 50 years, the weather predictors have seen numerous 10 day weather cycles in detail. They have used these to tune their models to their current state. Still, the models are far from perfect in their ability to say if it will rain or not.
Climate prediction seems a similar problem, only the data set includes more things which are just constants to the weather models. The pattern cycles lengths are much longer. (It's been quite a while since the last ice age, but not so long since the last El Nino.) Mankind hasn't been around to see these cycles and can only glimpse at a few clues from the last one. (For example ice cores.) Given this, there seems no reason to expect a anything like the finely tuned weather model for the climate.
That said, one can still try. It is possible to gather more and more of the current state of the system. This includes especially the sea flows. It seems likely that we might not be able to predict, but will be able to better know the current state. For a climate scientist to call this prediction seems dishonest and discredits their field. This probably limits any response that these folks are hoping to see.
(Score: 2) by HiThere on Tuesday February 19 2019, @06:14PM
Of course, the problem is that as the climate warms, poles melt, etc. we are moving into a area where we don't have the relevant data. The models make predictions, but we have no idea how accurate they will be when the climate is, say 2 degrees warmer. All we know is that it will be hotter, and that water will evaporate more quickly. Moving into a modified Venus scenario is not impossible...though doing so quickly is, since that depends on a lot of photo-dissociation. Still, a large increase of water vapor in the exosphere is not desirable by any means. Etc.
The really dire scenarios are believe to be less likely than, say, a large meteor strike setting of another "Siberian Traps" chain of volcanoes and cooling everything back down again, but that's a guess. And the intermediately unpleasant scenarios which are deemed more likely are also guesses. My real guess is that we'll either have an AI Singularity or a nuclear war before the poles finish melting...which will cause global warming to cease to be a problem in one way or another.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: -1, Troll) by Anonymous Coward on Tuesday February 19 2019, @01:39PM (3 children)
You didn't preempt jackshit. Climate, according to the 90's IPCC report itself, is a complex, dynamic, chaotic system And, quote, long term prediction of climate is impossible. Gee, something changed?
Complex systems are sensitive to initial conditions, and they are completely unpredictable.
Climate models were invented to study by experimentation, not to predict. They are pure futuristic bullshit.
Climate Change is bullshit.
Global Warming is a bullshit gibberish theory, based on only one predictable factor, CO2.
Anybody really interested should read Freeman Dyson's intro to a CO2 report here:
http://www.thegwpf.org/content/uploads/2015/10/benefits1.pdf [thegwpf.org]
Here, if you have an hour's time, is atmospheric scientist Richard Lindzen, calmy and dispassionately taking the bullshit theory apart:
https://www.youtube.com/watch?v=X2q9BT2LIUA [youtube.com]
And take your bullshit scientific consensus and shove it. Politics is consensus, science can be single individual being right with correct data e.g. Copernicus, Galileo, Einstein.
(Score: 4, Insightful) by fritsd on Tuesday February 19 2019, @05:12PM (2 children)
No, that's what we call "weather".
You're confusing weather with climate :-)
Do you own swimming clothes? If so, why?? It's freezing outside, you can't predict that it will become warm enough to swim in 6 months!
(Score: 2) by HiThere on Tuesday February 19 2019, @06:17PM
That's short term climate prediction. I'll agree that long term climate prediction is impossible. But long term is measured in at minimum thousands of years. Short term climate prediction is nearly as good as weather prediction, and in some ways better. (Summer is always warmer than winter, but noon isn't always warmer than midnight.)
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 0) by Anonymous Coward on Wednesday February 20 2019, @07:03AM
You really have no knowledge there, do you. You are talking seasons, not climate decades ahead. Way to prove your ignorance. You know of the famous/infamous Butterfly Effect. That's bloody weather, unpredictable, same with climate. Even more complex system. BTW, Comples, Chaotic are technical terms and if that what climate is it means it's inherently unpredictable, fucking perios. Same as earthquakes. There is not going to be any fucking long term future knowledge of climate unless you are psychic.
IPCC climate scenario oh we base it on:
prediction of future temps, prediction of future technology prediction of future populatiin numbers, prediction of future Dun activity, prediction of future cosmic rays, and we make readonable scenarios.
No, you fucking morons. You are bullshitting innocent people that you have knowledge that nobody can have, you dirty imbecilic liars.
(Score: 0) by Anonymous Coward on Tuesday February 19 2019, @11:54AM
Walking up to the top of a hill and pissing in the wind. The San Diego TV news weather girls are only there for the eye candy, and couldn't predict rain or shine even a day in advance.
(Score: 2) by PiMuNu on Tuesday February 19 2019, @01:17PM
From TFA:
> In comparison, the initial conditions for the EDA0.1 ensembles are perturbed with
> only 10% of the initial perturbations in the corresponding EDA ensembles centered
> at the control operational analysis of the IFS. With perturbation kinetic energy error only 1% of
> the current day state of the science analysis uncertainties, the EDA0.1 ensembles can be
> regarded as using nearly perfect initial conditions.
The level to which we can predict the weather depends on the quality of the measurement of the current weather. The authors assume a few percent uncertainty in the measurements; this determines how well one can make a prediction. No where do the authors justify this uncertainty. So the paper is just obvious. If one makes an imprecise measurement of a system, then applies some numerical extrapolation, then the extrapolation will be wrong. I get that the system is non-linear so the degree to which it is wrong grows, but that is pretty obvious too.
A more interesting paper might be to seek out fundamental uncertainties in the extrapolation, for example some stochastic process that is inherently unpredictable or beyond the scope of atmospheric physics - for example, solar or cosmic flux; biological or geothermal activity.
(Score: 0, Redundant) by Anonymous Coward on Tuesday February 19 2019, @02:22PM (2 children)
Forecasts longer than a few day ahead are shit for everything except temperature, and even that is often off by five or more degrees. I never trust forecasts longer than a couple days ahead, unless when I am in Miami where from May-November it's always 92F, humid, and with afternoon thunderstorms. Or when I am in a desert, or semi-arid prairie.
I don't think I've ever seen a local 5-day forecast where the 4th & 5th day have the weather they predicted on day one. On Monday they predict Sat-Sun will be hot with endless sunshine, sunshine, sunshine, then by Wednesday the weekend forecast is warm and partly cloudy, they on Friday they predict the weekend will now be cool and overcast is showers. Or the reverse is true: 5-days ahead is rain, rain, rain, they the weekend turns into sunshine.
Numerous articles have suddenly appeared in recent months about how much more accurate the forecasts are than 30 years ago, and these articles are just BS attempts at social engineering to manipulate the population into believing meteorologists can predict the weather any day any time anywhere. Why? Well every government now wants you to believe everything they say about climate change is true, and if they can "prove" that short-term forecasts are true, then maybe, just maybe more people will believe the predictions they make for 20, 50, 100 years ahead. Easier to carbon-tax the population then.
(Score: 0) by Anonymous Coward on Tuesday February 19 2019, @05:16PM (1 child)
You clearly haven't the slightest idea what you're talking about. Improvements in forecast are mostly due to improved numerical models, better input data and improved computers. You on the other hand are still an idiot.
(Score: 0) by Anonymous Coward on Wednesday February 20 2019, @06:56AM
He's right anf you are a fucking moron. Only an gofdn imbecile would trust a weather forecast more than a couple days ahead.
(Score: 2, Funny) by Anonymous Coward on Tuesday February 19 2019, @02:40PM
Seriously? Fuqing? Did nobody tell this poor bastard what that sounds like in English?
Yeah... My boss is Fuqing Zhang. He's a real ball buster.
(Score: 2) by kazzie on Tuesday February 19 2019, @06:02PM (1 child)
Editors have a New Answer...
(Score: 2) by bob_super on Wednesday February 20 2019, @06:13PM
I forecasted a slightly less stupid comment section.
It's amazing how vocal ignorant idiots get, must be some chaotic butterfly effect on dumb opinions.
(Score: 2) by arslan on Tuesday February 19 2019, @09:53PM
50% chance of rain - they're almost always right.