From Smithsonian Magazine
EVENT: “Paleoclimate: Digging into the Past to Chart our Future” streaming live tonight on our Facebook page. The discussion begins at 6:45 PM EST. (not live now, but will be at 6:45PM EST)
In this special program, NASA Goddard director and earth scientist Gavin Schmidt and Penn State climate scientist Richard Alley discuss what makes our paleoclimate models so robust—and how they can be used to confidently predict the future.
As human activities drive Earth’s rapidly changing climate, there is an urgent need to build better models that help us predict and prepare for our future. These models need robust data that stretch far back in time. Enter: the fossil record—a storehouse of climate evidence that paleontologists are getting better and better at deciphering. Join us for an evening with two renowned researchers—Richard Alley, Evan Pugh Professor of Geosciences and an Associate of the Earth and Environmental Systems Institute at Penn State, and Gavin Schmidt, Director of the NASA Goddard Institute for Space Studies—as they talk about their exciting work weaving together paleoclimate data and computer models to understand the future. Following their talks, Rachel Gross, a science editor at Smithsonian.com, will moderate a discussion and audience Q&A.
This program is part of the National Museum of Natural History’s Earth’s Temperature History Symposium* on March 29-31, 2018.
UPDATE: The archived interview is here
Super frustrating discussion of models that was completely expected.
Neither Richard nor Gavin seem to understand the absolutely fundamental difference between the GCMs and the models they quoted to do with bridge design and aircraft design and so on where those models are modelling instantaneous effects.
So for example bridge models look at immediate loads and stresses and those calculations are well known and work just fine. What those kinds of models dont do is take the loads and stresses at points in time and feedback those loads and stresses back into themselves to see how the bridge design evolves. Because it simply doesn’t evolve based on the history.
There is never any discussion of this and they simply cant afford to acknowledge it because then they need to address the propagation of error. And they know that they cant go down that path without every engineer on the planet immediately recognizing their limitations.
There are many engineers who post here, aren’t there?
” And they know that they cant go down that path without every engineer on the planet immediately recognizing their limitations.”
It seems to me that engineers would be well-advised to recognize their own limitations. Their experience with modeling, vast though it may be, does not enable the capacity to judge a very different kind of model. I know how incredible it seems that people can build models that represent climate realistically, and I also can imagine how important uncertainty is in engineering. It could take a veritable leap of faith to trust that the models really mean something, I suppose.
But engineering is not climate science. Would you trust a meteorologist, even a skeptic, to construct a safe and lasting highway bridge? I trust that the engineers will have done their jobs – even despite past bridge disasters (I live in Minneapolis) – because of their training and experience, and feel safe in my man-made environment because of it.
No. Not true.
When politicians chose “politically-connected” bridge design teams based on gender and political donations and political connections – as the Obama White House and the FL democrat representatives CELEBRATED with national honors and awards for the 14 million dollar 175 foot pedestrian walkway FIU Advanced Bridge Design – people die. Did their “models” work? Probably. (If their models were incorrect, the design team is liable for the deaths and the lead designer should be convicted and imprisoned. If her deign was built “wrong”, then the minority-owned construction firm owner should be convicted and imprisoned. If both were wrong – and that is what the forensics now indicate, both should be imprisoned.
But the politically-connected designers and builders only killed a few innocents. The self-called “scientists” and bureaucrats and politicians and academic teams who are using their CAGW models to push damaging fossil fuel restrictions and artificially high carbon prices for the benefit of themselves, their labs, their political power and their reputations (and their near-religious belief in their theory) ARE KILLING millions, condemning billions to squalid lives in the dirt and dark, dying needlessly of poor water, no sewage and no sanitation, bad food, and poor lives.
I cannot forgive them for their deliberate harm. For their false and misleading models and deliberate propaganda using their priest-garments of the white lab coat and “PhD incantations” of self-selected peer-review by partners and fellow operators.
Yes, innocent lives depend on a valid engineering model. And, too often today, the universities are sending out unqualified people (of all faiths, sexes, and races) that do NOT know the math, the curves, the equations, the metals and the concrete and the steel and the real physics of their material.
Politically correct? Certainly! But factually correct? No.
“Their experience with modeling, vast though it may be, does not enable the capacity to judge a very different kind of model.”
Well, I have an physics/mathematics background but I have worked extensively developing computer models in an engineering context. These were for logistics, a considerably simpler subject to forecast than climate, and my considered opinion is that models with more than two, or just possibly three, parameterized variables are utterly useless.
Kristi, you are so full of it
As an electrical engineer who learned control system theory back in the day, it was a piece of cake to understand how the climate system works, all the while the scientific high-rollers are off on this never-ending lark.
You just smeared one of the most intelligent and societally useful groups of people of all time.
Engineers have to understand things on first principles to make things that work.
Do you know the LASP TSI data is studied and finessed by scientists all round the world every day, but it was engineers and inventors who built the equipment? The lead scientist on LASP TSI decades ago said so.
Kristi,
This comment tells me you don’t understanding modeling or engineering.
Engineering is the application of science to real world design objectives. How do we design and build an airplane which can carry x amount of passengers and cargo x amount of miles, with a safety factor of 99.999% and do so as efficiently as possible? Is modeling used in this process? Yes. Do they do a good job of providing engineers with the information they need? Yes. Does this mean one can model anything and get results as good as the ones used for aircraft design? Not even close. Any model is only as good as the parameters it operates with and the data inputted to them. Whether you believe it or not, there is a huge difference between modeling based on well known parameters and modeling based on vague or generalized parameters which are essentially assumptions.
@RACookePE1978, couldn’t agree more. What’s next? Maybe they can have psychopaths design skyscrapers, so that they’ll feel “included.”
Kristi, do you understand modeling? Aerospace engineers use state-of-the-art supercomputer programs to design the newest commercial and military jet airplanes. The models aren’t perfect, and still ultimately require wind-tunnel and flight tests to confirm and modify the design, but are the most complex programs on the planet.
Now, many so-called climate scientists present their climate models as evidence of present and future CAWG. Really? The best/most complex models can just barely model airplanes, and then require real-world testing for finalization. How big is an airplane compared to the Earth? The extrapolation & hubris is astonishing — making such claims is like modeling a red-blood cell and saying they can now predict how a human being will function.
I hope you get my meaning……
Bob and timg56 – I’m so sorry! You completely, totally misunderstood me!!! I wasn’t smearing engineers at all! I’m just trying to understand how people’s background influence their perspective on climate and modeling. Mine does, why shouldn’t others’? No shame in that!
timg56’s comment illustrated it nicely:
“Engineering is the application of science to real world design objectives. … Any model is only as good as the parameters it operates with and the data inputted to them. Whether you believe it or not, there is a huge difference between modeling based on well known parameters and modeling based on vague or generalized parameters which are essentially assumptions.”
Believe me, I understand what you mean, and that’s my whole point! You have a different perspective on modeling. That’s not wrong at all, it’s essential to your profession. My point is that the parameters are not essentially assumptions. Your comment that “it was a piece of cake to understand how the climate system works” suggests overconfidence and simplification, since no one completely knows how climate works. It is extremely complex. The biotic interactions and feedbacks are poorly understood. The models are not intended to represent reality – they are models – but they are based solidly in reality. I may be completely, totally wrong, but I suspect you might have some impressions of model building, tuning and testing that are not quite right. There seems to be widespread misunderstanding about these things. I’m not accusing you of anything or making any assertions, but as you well know, expertise in a field counts for something.
Meteorological forecasts are based on the same chaotic, complex systems. Are they useless?
Kristi writes
Can you describe the fundamental difference between a weather forecast and a climate projection in terms of what the models have to do?
Kristie,
Regarding your question on whether Meteorological forecasts are useless?
Beyond 10 days, basically yes.
And they have actual data entered into the models.
‘What those kinds of models dont do is take the loads and stresses at points in time and feedback those loads and stresses back into themselves to see how the bridge design evolves. Because it simply doesn’t evolve based on the history.”
Actually, this got me thinking about cracks. It seems like once you have a crack, the freeze and thaw cycles will work on it in a kind of feedback? Or a bit of metal exposed to corrosion that shouldn’t be, causing spread within the formerly protected area?
(Apart from that, your description doesn’t apply to climate models any more than it does to bridges.)
Kristi Silber
From Campus Reform ^ | Mar 23, 2018 at 9:45 AM EDT | by Toni Airaksinen
Kristi writes
You could certainly model failure like that. And the stress at the crack would be well known at every time step given the modeled reduction in component strength so any uncertainty would come about from the modeled specifics of the component’s condition at points in time.
But none of that alters the bridge design per se. Well it doesn’t “alter” the bridge until complete failure anyway. And at that point its again, a simple calculation of loads and stresses given that one component has failed.
So can you model a component failure by repeated freeze and thaw cycles on a crack? Sure…but that isn’t a feedback in the same way a CGM does it. Not by a long shot.
To say that it is impossible to model the climate in all possible worlds may not be correct,but,is cycles within cycles as the world chases its tail trying to find equilibrium. Outside of these worldly cycles are more cycles within cycles.
The sun and the moon and space weather, then we have seemingly random events like volcanoes and earthquakes that stir the oceans with tsunamis, these are but a few of the epicycles that stir the pot in a chaotic world system trying its best to get back to peace.
Those that are so egotistical that they think they can model the climate, then tell us all good citizens that any change is our fault need to take a good hard look at themselves, in the mirror.
Some models have utility. To date climate models have demonstrated none.
@RACookPE1978, Reading that makes me want to vomit – preferably on the fools that contributed to writing it.
I don’t give a damn about how many of what gender, race, nationality or anything else enter a given field. In particular, I WANT engineering to be forever a “meritocracy” – it’s more important that the bridge doesn’t collapse than it is how many of its designers were [fill-in-the-blank “protected group”]. FFS!
AGW is not Science –
I agree with you. What’s wrong with meritocracy? I have no idea why that was directed to me.
I worry that conservatives (generalization!!!) have been complaining about academia so much that young conservatives will not want to go to college, get advanced degrees, and become academics themselves. Education needs a diversity of views, and the liberal dominance is not good for the country. But it’s up to conservatives to do something – encourage their kids to get PhDs and become professors rather than complain about academia. The country needs highly educated conservatives. (I’m not saying one can’t be educated without a degree – not at all. But a PhD is no small accomplishment.)
Cheers all,
Kristi
The above link shows you how a models output is shown to be right when the observations show it to be wrong. Just make the confidence level wide enough.
Why are they bothering? I thought the science was settled.
Of course, silly me, no-one would give them a proper job to do and they have to keep reassuring themselves that they are right.
I attended the lecture. The auditorium was perhaps 75% full, perhaps less. I thought that both speakers did extremely well .. until the bridge analogy. Just how many “well modeled” bridges have failed?
Gavin’s discussion about models was excellent. He admitted that there were errors but deflected that by covering what the models got mostly right. Of course, his temperature graph did not contain the pause, but he convinced me that he truly believes that his group is doing the best science possible.
At the end, there were questions from the audience – written on cards, reviewed by someone, and given to the moderator. There were lots of cards turned in – but (IIRC) only one was read. I was really hoping to hear how they responded to the skeptics. After the cameras were off, they went through the cards. Unfortunately, I did not get to hear what was said.
If anyone was on the fence, this presentation was good enough to convince them that the presented view was the only correct view. Extremely well done.
And if the models get the thing they are designed to model -future temperatures – wrong, who cares what they supposedly “get right”? And if they get right some stuff but get wrong other stuff, then they are getting it right through chance. It always amuses me for example when people claim theur models are right if they underestimate what happens. No, your model is just as wrong as if you overestimate – and may in fact be more wrong.
Yet when we compare proxies to known, modern temperatures we know the proxies can show the opposite of the instrument record.
And since we can never actually know whether the reconstructions are accurate, we should always treat them with a great deal of scepticism.
The stupid thing about Alley is that he was the first to announce that rapid climate change was normal and that it had happened many times in the past. His analysis of GISP proved it.
Two fricking clowns on one stage.
Is this an early April fool, or a comedy show by Schmit and Alley?
“In this special program, NASA Goddard director and earth scientist Gavin Schmidt and Penn State climate scientist Richard Alley discuss what makes our paleoclimate models so robust—and how they can be used to confidently predict the future.”
Confidently predict the future. ??? Predict the future? What utter crap. Never has happened; never can happen and I confidently predict that it never will happen.
(need I put a tag on this)
Now this should have been in the Friday funny section
Nassim Nicholas Taleb lectures at length on the fragility of robustness. Things are robust only in the anticipated dimension, hence the hazard of the Black Swan hiding camouflaged in fractally complex reality.
Not at all surprised this would be from Smithsonian Magazine.
My in-laws for the past 5-6 years have been giving subscriptions to Nat Geo and Smithsonian. I’m too polite to tell them they they are NOT the same quality from about 40-50 years ago.
In the latest issue from Smithsonian:
“Do Trees Talk to Each Other?
A controversial German forester says yes, and his ideas are shaking up the scientific world”
https://www.smithsonianmag.com/science-nature/the-whispering-trees-180968084/
Some plants, when stressed by pests, emit pheromones that attract predatory insects. That’s a far cry from communicating with an actual language, though.
I’ve heard of this forester from his earlier book, The Hidden Life of Trees, mostly unsubstantiated anthropomorphising little understood natural relationships. He is best described as a naturalist, modern term being environmentalist. He might actually be onto something. I know many foresters in Vermont who have expressed similar viewpoints based on their experience.
Paleoclimatology is as “robust” as a whorehouse at low tide.
Ha ha
You misspelled “warehouse.”
There are lots of things that are robust in paleoclimate studies. Such as the LIA, the medieval warm period, the Roman warm period, co2 concentration increase following warming…….and so forth………and so forth….
The WARMEST PERIOD IN THE CURRENT EPOCH, the Holocene Climate OPTIMUM,…
“Robust” – ORIGIN – mid 16th century: from Latinrobustus – ‘firm and hard’
Like a cadaver?! LOL
Paleodata has its place. It helps pinpoint ice advances in the past. It helps find and date volcanic eruptions that clearly took place given abundant ice core evidence, and helps date evidence of ancient human taking of prey when obvious tools and human remains are absent.
Schmidt is a good example of how once-respected institutions become corrupted by stacking the decks with the like-minded – and they can go a long way on exploiting trust established in past generations.
In a nutshell, that’s also what’s wrong with modern universities – close minded communities of opinionated ideologues never challenged by reality.
Having experience with Transportation Gravity Models, used in long term planning studies, I can appreciate the thinking behind the climatological models. In my case, it was dealing with only a 20-30 year horizon, in their case an 80–100 year period. The major difference being the transportation model used well known relationships validated by on the ground historical measurements.
Even then (almost 50 years now!), I noticed that the cumulative errors caused a very gray scene when rendered 20 years in the future. The fog was caused by the uncertainty of the model, even though the current temporal relationships were well understood. It was because there were too may exogenous variables involved. Economics, politics, land use (even birth rates) being the major questions. Because of that, the results were not taken as gospel and only used as a general guide, with the understanding that things would be routinely reviewed every 5 years and completely reevaluated every 10 years.
The emphasis placed on the current climate models is way overblown. I have a strong feeling that the current crop of modelers, if honest, feels the same way. No one their right mind would ever consider a model’s output for a future scenario as “robust”.
I have no trouble believing that Alleys Gisp-2 paleo-climate reconstructions are very robust but since they show that current warming is much less than medieval warming, roman warming and the climatic optimum warming then they don’t actually support any argument for man-made warming. Neither btw do they show CO2 and past warming in lock-step. He clearly believes in man-made warming because for some reason he he wants to believe it despite the fact that his own data utterly refutes it!
The kicker for paleo data is that when it comes to modern warming the warmists say we must look at the arctic to see manmade warming but not at the antarctic because that is naturally cooling yet for past climate they say look at the antarctic but not at the arctic (for reasons explained above). This willful blindness to adverse data and constant contradiction is typical of climate ‘science’.
Exactly why they believe in thermageddon remains a mystery to me but it sure as heck isn’t scientific. Perhaps moral panic, middle class angst, funding, anti-capitalist dogma or an inexplicable desire to blame fossil fuels for anything bad; having already tried to blame them for global cooling and acid rain that didn’t happen either.
Reality is what is real. There is no such such thing as a reality which conflicts with another reality.
One of them is unreal.
In palaeoclimatology measurements of past climates, in the few places where the measurements are taken ,are used to generalise the temperatures and weather conditions in surrounding areas. Can this always be done? I don’t think it can.
More caution in the correlation of data would be a good idea
If we can’t be sure of the basic measurements we can’t be sure of the computations made from those measurement whether we use an abacus or an electronic device. We don’t have much knowledge of the past so we can’t use that small amount of knowledge to predict the future.
Uncertainty is a fact of life. That is my assertion. One can never be sure what happened in the past and we may have missed the important bits. That’s the way it is.
( Given certain news about Facebook in the last few days it may not be a good idea to use it until it is possible to regulate it’s behaviour with user’s contacts)
I’m sorry, I saw the names “Gavin Schmidt and Richard Alley” and misread “robust” as “a bust”.
Any honest examination of the Earth’s climate will show that the general trend is toward cooling. Any suggestion of anthropogenic must founder on the inescapable fact that mankind did not even exist during the most radical eras.
Please enter “influence” between anthropogenic and must. Sorry about that.