SoylentNews is people
SoylentNews
A new study in Nature [Ed-Abstract only for non-subscribers, but see below.] predicts that climate warming will be 15% greater than previous high estimates have predicted. This new study suggests that humans need to reduce greenhouse gas emissions more than previously expected and more than the Paris Agreement calls for. This study was based on analyzing the earth's "energy budget" (absorption and re-emission of radiation) and inputting that into a number of different climate models.
Also covered in more detail in Phys.org and in the Guardian.
The researchers focused on comparing model projections and observations of the spatial and seasonal patterns of how energy flows from Earth to space. Interestingly, the models that best simulate the recent past of these energy exchanges between the planet and its surroundings tend to project greater-than-average warming in the future.
"Our results suggest that it doesn't make sense to dismiss the most-severe global warming projections based on the fact that climate models are imperfect in their simulation of the current climate," Brown said. "On the contrary, if anything, we are showing that model shortcomings can be used to dismiss the least-severe projections."
(Score: 5, Insightful) by bradley13 on Friday December 08 2017, @10:14AM (6 children)
Unfortunately, it is difficult to have much faith in any of these models. There have been literally hundreds of models proposed. All have been adapted until they could predict past climate changes. Their performance going forward has been basically random. So now they've picked a few of these models, and fiddled with them so that they can retroactively predict the recent past. Unfortunately, this still doesn't mean that they will do any better going forward.
"It makes sense that the models that do the best job at simulating today's observations might be the models with the most reliable predictions"
No, actually, it doesn't. At least, not if you know anything about modelling. Some mathematician once observed: Given enough variables to play with, you can create a function to fit any given data set. This says absolutely nothing about what happens when you go off the end of the known data - in fact, your function is likely to shoot off in some random direction.
I used to do this kind of stuff in a research group: not climate, but still fitting past data, and trying to make projections. Models get complicated quickly - you have plenty of variables to play with, and you can fit any past data. This says absolutely nothing about the accuracy of the model into the future. The only evidence of accuracy is provided by making and testing predictions.
If they want to prove that these models work, they must make near-term, falsifiable predictions. Taling about temperature in the year 2100 is completely useless.
Everyone is somebody else's weirdo.
(Score: 1, Interesting) by Anonymous Coward on Friday December 08 2017, @03:09PM (1 child)
Well, the fact that all of the models show a rising temperature probably means something.
(Score: 3, Interesting) by fyngyrz on Friday December 08 2017, @06:21PM
Yes, it means that all the models show a rising temperature.
When the environment's temperature shows an actual rise, that's one thing. When a model shows it, then fails to predict into the future... that's another.
The thing about science is that the method is that you form an idea, then you test the idea and attempt to falsify it. When you skip those latter steps, you're really operating on guesswork, not doing mainstream science. If, and it's a big if, time proves you right, then that's great - but it's still not really science. It's guesswork that turned out to be correct. If I say, "when I throw this die, it's going to come up seven", and then I throw it, and it does, I wasn't doing science. I was doing guesswork.
These models are all dice throws. Some very clever dice throws, I'll grant, but still - dice throws. Until the die lands, and all the forces that come into play have done so, we won't know what the number is, except that it will be a number on the die.
Likewise with climate: There's a great deal that can happen between here and the predicted outcomes (and there are many.) Transition to EVs. Technological ameliorations. Changes in vegetation growth and density and carbon sequestration. Volcanic activity. Petroleum pricing changes. Artificial meats (a consequence of that might be less methane generation.) Forcings (or lack thereof) in the atmosphere that were missed, or calculated wrong. Etc.
Predicting the future is a tough game to play and get right. We've seen - many times over - models get things wrong. The weather models get things wrong. The climate models get things wrong. The voting models get things wrong. Too many variables, models that are (far) too simplistic.
The obvious thing, I think, is to aim for backing off on emissions and pollution just because that seems most likely to leave the planet in a state closest to what we're used to; and we know we survive reasonably well in that state. If we're careless, or simply not cautious enough, we may cause problems we (well, our descendants) can't deal with. If you care about our descendants, the smart moves seem fairly clear in the light of changing the nature of our environment.
But are we smart enough? I read the news, and I have to say... I don't think we are.
(Score: 2, Informative) by khallow on Friday December 08 2017, @03:37PM (2 children)
No, it hasn't been random. There has been a consistent bias towards exaggerating the warming.
(Score: 0) by Anonymous Coward on Friday December 08 2017, @10:16PM (1 child)
MW67 underestimated [wordpress.com] the warming effect of carbon dioxide.
(Score: 0) by Anonymous Coward on Saturday December 09 2017, @07:25PM
The grand conclusion based on 3 outliers?
https://climategraphs.files.wordpress.com/2017/11/manabe_wetherald_1967_updated.png [wordpress.com]
Color me completely unsurprised. And cross out the "science" word from that "Climate Science" nameplate.
(Score: 0) by Anonymous Coward on Sunday December 10 2017, @10:22PM
I know. But what are you saying? They shouldn't be tuned to predict past climate? So they should predict what then? Magic climate? Unreal dreams? Or what?
Do you even know how empirical modelling works? You take some system, then you have some fitting parameters. Then you fit it. Like you know,
E = mgh
where m = mass, h = height, and g is the fitting parameter called gravity. It's a model. And it seems to predict things too. Would you rather that "g" wasn't what it is because you know, it's not constant on our planet either. So must be all wrong, right???