This study went back to see how 17 of the earliest climate models were at predicting the future, the future when they were published.
This study found that the higher predictions than the amount of actual temp. rise that has been measured since is about 95% the result of the original authors over estimating the amount of future greenhouse gases would be emitted. The model was not the problem, the problem was in the prediction of human behavior, of economic behavior. If the correct amount of greenhouse gases is entered as data and then the models {17 models} are run, then 14 of the models give us almost exactly the amount of warming as we have measured.
For those who are kind of dense,
what this means is that for 30 years those early models conform to the actual temps measured. This
increases our confidence that the models do, in fact, correctly predict what warming we will see in *
our* future if we can correctly predict the amount of greenhouse gases that will be emitted.
At a *
minimum*, this means that if we just stopped emitting them after next year, then those early models would correctly predict how much warming we wold see over the next 100 years. They do predict that temps would rise at least 1 deg. C and maybe 2 C. Many people and scientists are very worried that a 1-2 deg. C temp rise would trigger one of the known tipping points, and this would raise temps by a few more degrees C. When a tipping point will be triggered is very much unknown and probably impossible to predict before it happens. And it is certain that it can't be *proven* when a tipping point *will* be triggered before it happens.
https://www.scientificamerican.com/arti ... l-warming/The edits were for just 3 typos.
________________ ________________________________ __________________
I saw this article on another site. Below I post a reply I liked from that site's resulting thread.Sorry, I'm afraid to post a link to it on this site because I don't know the rules on such things. Nor do I know how to format it here correctly. The last quote runs til the bottom of this reply.
[quote="Samuel Vines"]
"Sir, a question!
So if climate models have changed a lot, and the very first ones are giving the values we observe now, and so do the ones that have changed a lot, that leaves us where?"
[/q]
[quote="Benkyo"]
"Just because a model becomes more accurate, that doesn't necessarily change the output."
[/q]
[quote="Samuel Vines"]
"I am well aware of that. Or, more precisely, that is it indeed, observably; [that it] didn't change the output is exactly what triggered my 'so where does that leave us' comment.
Take any subject. Usually, improving early and rough models leads to more accuracy (and finer detail). In this case it is leading to finer detail only, it was accurate to begin with. That is huge. It is far more than just a counter to people who state that Hansen 1988 was 50% off.
So you went back and found the disused model equally accurate.
Surely if you are a scientist, and on the one hand you state that 'climate models have changed a lot' and on the other hand you go back to the old ones and you find they provide the same observable output, that ought to trigger immediate follow-up questions in your mind.
If anything, I totally wouldn't have blamed early models to be inaccurate to a significant degree. Early models usually are and they become more accurate after feeding in related processes which were overlooked. Insight improves accuracy.
But we are not getting different results by improving the models. Just more detail.That's huge.
The models are simplified versions of utterly complex systems based on known chemistry/energy principles. They have two elements, looking back and looking forwards. They must be able to replicate the past if known amounts of driver compositions are fed into it. And they have predictive qualities which rely on assumptions about future changes in the composition of drivers. A reliable model, 50 years from now, a 100 years from now, being fed (by then) known data, ought to replicate the observed reality.
It doesn't surprise me that the early rough models provide the same output as the more sophisticated ones. Chemistry principles have not changed and the main drivers were already identified. That was what Hansen 1988 was modelling. And the effects of the Montreal protocol has been provided as a large part of Hansen 1988's overestimate long before this study above. Something tells me that Hansen himself did revisit his study a decade ago and concluded this very thing, but maybe my memory is no longer what it was.
The new model's sophistication, based on vastly increased insight in the complexity of the system is offering local detail that was impossible to extract from the rough early models. That is great for projecting locally, and studying local questions.
But that we are not 'improving' the accuracy of the core projection significantly at all is also suggesting that we might well be fiddling at the margins, as we simply are not discovering any process which is altering the base projection despite working with models which process insight a lot more insight in related processes.
So we now have a model which projects forcings that have been proven accurate for 3 decades. That's scientifically significant enough to take the model's further projections
very seriously. No more handwaving.
Also, we are not discovering model mitigation at all. Ouch. Nor, apparently, unforeseen acceleration. Yay!
That's my core takeaway.
So, if the core drivers really are 'it' and Hansen 1988 gave us already all we needed to know, and his predictive model has now been right for decades, where does that leave us?
It leaves us, as pointed out by the poster above, with a simple model that looks to be remarkable accurate in flagging up the horrors on the road ahead, and hasn't be shown to be inaccurate yet. It has been accurate for decades, seemingly. Hence,
this model leaves us without excuses for an obligation to act, knowing the scale of the disaster in our imminent future which this model predicts if current behaviour remains on the course it is. Including the points of no (quick) return.
Still, turning to the newer detailed models, they might provide answers on how to manipulate the outcome locally for maximum effect, although their local accuracy still has to be proven over time. They should suggest manipulation which can be tailored locally, all helping to accomplish the overall goals.
It also suggests we shouldn't be opening new layers of fossil fuels in a hurry, not until we have ways to mitigate those numbers being released into the system. Chuck in those numbers and nothing we are doing in the meantime will be enough to match our Paris' (already weak) ambitions. Given what we know, allowing companies to extract them is a crime against future generations. Nothing less.
Wir haben es gewusst."[/q]