Posted on 03/09/2020 6:36:23 AM PDT by ConservativeDude
An animation of a GISS (Goddard Institute for Space Studies) climate model simulation made for the United Nations' Intergovernmental Panel on Climate Change Fourth Assessment Report, showing five-year averaged surface air temperature anomalies in degrees Celsius from 1880 to 2100....
(Excerpt) Read more at climate.nasa.gov ...
I’ll take “Things That Don’t Matter” for 200, Alex.
Goddard space center, Goddard space center, why does that name ring a bell? Oh yeah, my son used to work there when James Hanson was the BMOC there. I attended one of his big rah rah climate change rallies on campus. Great place, but the indoctrination is total.
Oh yeah, I forgot to ask, did they have asphalt paved parking lots in 1880?
Yet another massively-fraudulent study.
If these “climate models” used to predict “0.7 degrees C warming in 100 years” (or whatever the fraud du jour is) were any good, they could be initialized to the conditions of 1920, and a 100-year forecast made, and it would match the actual historical climate record of the last century. Including the heat-wave of the 1930s, the cool phase in the 1970s, etc.
If they could do this, they would have done it, and they would be shouting it from the rooftops.
But they haven’t done this, because they can’t do this, because this whole “man-caused global warming” is a total fraud intended to put most of us in mass graves and the rest in chains.
Did they have calibrated certified thermometers in 1880 accurate to 0.2 degrees?
Now I understand why they raised measured temperatures...
The data didn’t match the models.
Using what data? Why only from 1880? Shouldn’t the model be able to demonstrate the period from 1000BC to 2000AD, if it is a good model? How do we know the model is real, and doesn’t have the answer to historical data built into it, much like VW diesels did when they recognized they were being tested?
Oh, and GISS, is that the group which regularly and systematically lowers the historical temperatures through adjustments in order to make recent temperatures look warmer?
Hardly.
Science has to be repeatable and falsifiable. The warmists have already declared that there is nothing that will falsify their faith in warming.
The methodology embedded in the models must be published, reviewed, and demonstrated to be repeatable, even in the hands of skeptics. As the situation is now, all of the models are closely held secrets (to my knowledge), and therefore could easily be leavened with corrections on inputs or outputs to match known historical data. Computers literally do exactly what you tell them to. Nothing more, and nothing less.
Perhaps someone in the climatology field can let us know if the current generation of models includes water vapor, which has been noted as conspicuously absent in prior generations. This basic piece of the climate is (or was) ignored by climate models, which means that any predictions from the models came without understanding a huge factor in the actual workings of the climate. Adding water vapor in makes the calculations required by the model enormously complex, of course.
I specifically note the one large down-spike in the historical record. This is mimicked in the modeled data quite closely. Inflection points like this require specific attention to prevent fraud in the models, as prior knowledge of the inflection will influence the modelers to program in corrections specific to the time period, rather than being specific to the combination of events that caused the inflection.
Model based science is rife with opportunity for scientific fraud. It depends on complete and utter honesty on part of the scientists involved. We KNOW from Michael Mann that this basic honesty is lacking in the climate modeling community. Any study that relies on self-reporting from the model creators is structurally flawed.
NASA, itself, has vacillated on the honest/dishonest line, and therefore I do not consider them a viable arbitrator of truth on this subject.
Something modeling stuff in the future is modeled by another program that finds the first one correct?
Looks like GI = GO
Any study that relies on self-reporting from the model creators is structurally flawed.”
Thank you. And that goes to your pint that the methodology embedded in the models must be published. “Model” is not a talisman which they can simply pull out and magically wave.
Helpful response. Thanks again.
One is to take a look at the graph they provide. I suspect it is merely the most favorable sub-range of the whole data set. There is one interesting trend there that the article ignores. The "ensemble spread" of the models - the upper and lower limits (95% confidence interval) - is widening. In other words, as these models estimate further into the future their predictions increasingly vary. Not an indication of good accuracy.
But what about the average? Averaging is pablum for a weak mind. It is basically saying they really don't understand the climate yet. Some of the models apparently get certain aspects correct, while other models get other aspects correct. Run them together and average them over time and they come up with something that looks reasonable. But they don't know why one model spins off too high, another too low - with increasing disagreement between them.
Then there's this gem (emphasis mine):
To successfully match new observational data, climate model projections have to encapsulate the physics of the climate and also make accurate predictions about future carbon dioxide emission levels and other factors that affect climate, such as solar variability, volcanoes, other human-produced and natural emissions of greenhouse gases and aerosols. This study’s accounting for differences between the projected and actual emissions and other factors allowed a more focused evaluation of the models’ representation of Earth’s climate system.
The key there is that "accounting for" I'll bet you a good steak dinner that accounting for involved re-tuning the output of the model for each time step with the actual conditions (carbon dioxide). Of course a model is going to look pretty good if you keep re-calibrating it every few time steps with actual data.
The authors say that while the relative simplicity of the models analyzed makes their climate projections functionally obsolete...
Oops, were they supposed to admit that in public? They've been telling us for years these are "the" models. Now they're saying no, wait, those are obsolete, the models we have now are "the" models. Right...and in a few years???
“As climate model projections have matured, more signals have emerged from the noise of natural variability that allow for retrospective evaluation of other aspects of climate models — for instance, in Arctic sea ice and ocean heat content,” Schmidt said. “But it’s the temperature trends that people still tend to focus on.”
Another buried gem - they are admitting there are more "signals" - ie. variables to the system than they previously realized. As the models get better and are studied they'll no-doubt find more variables that are important - that help explain and reduce that variability between the models. Maybe at some point in the future they might even reduce or eliminate the need to keep "accounting for" or re-calibrating the models mid run to get a match to observational data. You see, that's the rub. They can make the models match observed data if they take into account all the things they didn't see coming. Well sure, I can predict the future too as long as you let me adjust my predictions to account for the things I didn't see coming.
The climate models are worthless as predictive tools. Worthless, perhaps even worse than worthless. This is not to say the computer models are worthless. They are only worthless in terms of prediction. They are very useful tools for developing an understanding of how the climate works. As the article points out, they "discovered" more "signals" - arctic sea ice, solar cycles, etc. This is great, this is exactly what computer modeling is good for - increasing our understanding. "Hey, the model doesn't match reality, let's figure out why...." some time later "ok, now the model is better, but this part is off..." some time later "Now we need to fix..." By building up the model they are learning what factors are important. The bigger and more comprehensive the model, and the longer you want to extrapolate with it - the more factors there are and the more even the subtle ones matter. So they're great learning tools, but they are not predictive, not even close.
A warmer Earth is better. The planet was warmer in the past. Increased CO2, which Follows temperature changes, leads to increased plant growth, and increased animal life, and better and easier farming: more food for people.
The Earth was warmer during the golden age of the Roman Empire. It was warmer when the Vikings explored and settled Greenland. If the Earth was warmer life would be easier in Scandinavia, Russia, and Canada. A warmer Earth means fewer crop failures in China.
excellent, thank you.
this really is more about handling/interpreting/manipulating data than it is hard science (ie, how much CO2 and so on).
your thoughtful response is helpful and superb. thank you for taking the time. :)
So they make a model of temperature predictions using historic data from the 1800s to present.
Then they run the model and find that it’s accurate on the historic data (plus or minus 2 degrees) and for a few years of new data.
So now they announce their model is accurate and they can predict the temperature for the next 100 years and it’s gonna get warmer! (Plus or minus 2 degrees which is well within normal atmospheric fluctuations and far beyond the disaster screaming of half a degree will kill us all!)
So...the best way to think about a model, even as it increases in sophistication, is “explanatory” (it backward looking) not “predictive” (forward looking), would that be correct?
Could I offer a challenge here?
Economic models that deal with commodity prices are often functionally predictive, right? Companies look closely at the data and make changes to their behaviors based on what they see in the data. And there is a whole trading industry which “helps” them.
Perhaps the response to that is, as every trader knows, yes, we try to make the data predictive.....and sometimes we are right, and sometimes we are wrong.
Is that the right way to think about the relationship between modeling and predicting?
And also that when we are talking about commodity prices (also a very complex phenomenon...not as complex as the global climate, but still, can be complex), we are talking at most a few years in advance....not projecting out to 2100.
Is that the right way to think about that?
Thank you.
The overwhelming majority of the EARTH’S HISTORY, since it has cooled and had a solid crust, and oceans, has been far warmer than ti is today.
The majority of that time, the earth has not had polar ice caps, and it it was not HUMAN ACTIVITY that caused that warming or cooling cycles.
The earth has a carbon cycle, it has a orbital cycle, the sun has cycles...the solar system itself travels through various parts of the galaxy over long cycles... all of which have, or potentially have, affects on climate... The earth, even with all the human activity is in no position where its in danger of dying from our activity.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.