How limited is your confidence in climate models?

Question

Southern hemisphere sea ice trend (satellite data): http://nsidc.org/data/seaice_index/images/s_plot_hires.png

Climate model projections (bottom two panels are for southern hemisphere): http://www.ipcc.ch/publications_and_data/ar4/wg1/en/figure-10-13.html

I don't see a match (empirical observations vs. model output) but let's ask some experts: "... the Antarctic ice spread has the effect of "limiting confidence in [the models'] predictions"." http://www.theregister.co.uk/2012/11/12/antarctic_ice_growth_investigated/

The study mentioned in the above article also states in its abstract: "However, climate models have failed to reproduce the overall increase in sea ice."

How does this affect YOUR confidence in climate models?

Answer 1

Deniers seem to be the only ones who think models are exact, scientists certainly don't I would have thought that the fact the graphs you present which display 5 different models with varying error bars made that pretty obvious.



Have to agree with several others here jim z's comment is clearly from 'right-wing world' given that the LIA & MWP are quite clearly in the climate data, jim seems to be confusing models with proxy data, they are not the same thing as any real geologist would know.



Then we have your question deniers continue to claim the temperature data has fallen below the models, when it is fact near the top edge of one of the mid level models, all of which seems to prove denier can't read simple graphs.

Sadly that seems to be the case here as well, do you understand what error bars are or that Antarctic ice is still well within three of the error bars of models you show in fig D as it is up just 0.9% per decade while the Arctic is down 4% per decade.



What about Arctic ice it is in fact near the very bottom of the error bars for the models and at it current rate of decline will soon fall below the models (as they were in 1999)



jim, jim jim, "I figured the worst president in our lifetime" why bring GW Bush into this, or are you one of those "scientists?" who pretend Obama caused the GFC before even coming to office. That would tend to suggest your knowledge of politics and economics match's your knowledge of geology and climate, i.e it's none existent.

Answer 2

What is the relevance of you asking us how limited is our confidence in climate models? I really do not understand your point in asking this question.



Here is how the science works. Models are based on information available. When the models are ran they will give projected results. Those that have studied the science then assign their degree of confidence in as to how likely these predicted results will come to be.



Examples:

Should the climate models predict that over the next 50 years there will be a 300 foot sea level rise, then those that have studied the science would most assuredly assign a low percentage point (<10%) that this will be the case. This would illustrate a very low degree of confidence, by the scientist, that this would actually happen.



Should the climate models predict that over the next 50 years there will be a 3mm sea level rise, then those that have studied the science would most assuredly assign a high percentage point (>90%) that this will be the case. This would illustrate a very high degree of confidence, by the scientist, that this would actually happen.



Do either of these model runs show that in 50 years there will be exactly a 300 foot sea level rise or exactly a 3mm sea level rise? NO! They only show probabilities and then the scientist have to assign a degree of confidence in these probabilities.



Models do not predict absolutes. They never have and they never will. Models show probabilities based on the information given. These probabilities must then be evaluated, by someone that actually knows what they are doing, and assigned a degree of confidence in these probabilities. Either you do not how models were intended to work, you do not how scientist assign their degree of confidence in model predictions/projections or you intentionally try to create confusion here. Which is it? I am leaning towards, "All of the above"! Are you able to narrow this down a bit from here?

Answer 3

The only way to model something with 100% accuracy is to replicate it and the conditions it experiences completely. If you are looking for a perfect model of Earth's weather then you have to use the Earth and the Solar system as the model. A model is naturally an approximation with less scale, information and with shortcuts in it's processing so that the simulations can occur at a faster rate. In the early days of computers, it was said that it took a computer 24 hours to run a simulation 24 hours into the future.



I think that the models we have are extremely accurate considering the reductions in data and algorithms required. Of course a denialist always focuses on uncertainties and there is always uncertainties with models.



When there is a consistent difference between the model and observations, it means that some factor that should be modeled isn't. It doesn't actually impact the reliability of the model because you know there will be that difference. If a thermometer always read two degrees higher, you can still use it to measure the temperature.



My confidence in the model is not diminished by the observation that more young ice formed as it only means we haven't modeled everything yet, and I know that we haven't. Had the model predicted everything exactly then my confidence would be diminished because I know that a model will always be lacking in some way and a perfect simulation shows me no clues as to how or where the model is deficient.

Answer 4

So you are back on models again. I have no interest in models but am concerned about actual temperature readings that show a slight although continual rise in global average temperatures.



Although there is an increase in surface ice over both the Arctic and Antarctica in winter, the volume is still decreasing in the summer months. This is because the ice is melting in western Antarctica at a faster than normal rate from underneath due to the increase in water temperature. The phrase "...overall increase in sea ice" .is a misrepresentation of fact regarding the actual ice volume.

Answer 5

http://www.theregister.co.uk/2012/11/12/antarctic_ice_growth_investigated/



Hmmm, the similarity in topic is only coincidental, I imagine.



Of course, you don't even bother reading the paper that sparked the newspaper article, which basically says "Antarctic climate is complicated, and it's not surprising the global models get part of it wrong." More importantly, you ignore all the things climate models get right, because to acknowledge that broadly they are projected what is happening is kinda scary to you. Best deny it, sort of like the Republican pundits did with the polls last week. But it sucks to be wishing the real world would just go away.



Anyway, as the authors of the paper linked below state, the model projections of antarctic ice trends shouldn't be trusted yet. But to think this implies they (the models overall) are totally in error, given how they get global mean temperature and sea level rise correct, is ridiculous. Now nuance isn't the strong suit of you skeptics, but I'll try one more time. The thing is, when you are talking about regional scale predictions for effects covering relatively small areas of the planet, the agreement between model output and observations over short time periods is going to be poor in some cases. But the agreement between model predicted and empirically measured global variables will be better because the regional scale inhomogeneities get smoothed out. This is why ice cover in Antarctica is poorly modeled (the models get the details of the flow wrong) yet global temperature and sea level rise are predicted rather well.

Answer 6

There is no basis for confidence in the data the climate models use in the first place. There is little if any multiple independent verification. (Indeed there was a case where "evidence" of temperature change was presented that didn't adjust for the urban "heat island" effect and urbanization. You'd expect such an adjustment caveat at least from bright high school science fair entries.) Yet global warming enthusiasts expect others to accept predictive powers equal to fields whose standards they do not meet.

Answer 7

That website is a joke, right? Its a new version of "the onion"? I mean, did you look at the headlines under the "spotlight" section down the RHS?



Anyway, my faith in climate models is limited mainly by my lack of understanding of them. Since I don't understand them very well, I don't feel like I can criticise. What I can criticise is drawing conclusions about a trend of increasing ice mass which features such massive error margins (+/-66%!!) and then trying to compare these to images which DO NOT COMPARE THE SAME TIME FRAMES. The IPCC projections are seasonal projections, not annual projections, consisting of only 6 months out of each year; and you are comparing this to a decadal trend. Please Mike, enlighten us. Apart from the fact that they are not discussing the same thing, what is it about these two datasets that you think is so alarmingly different?



This is why people who don't understand science shouldn't be allowed to use the internet. If error bars cross over, that means there is no significant difference. If you aligned the projections with the dataset, (assuming you could given that you are not comparing corresponding data) where do you think the upper and lower thresholds of the data might be? Obviously you assume there would be a significant difference. Do you have any evidence that it is (apart from the joke article from the onion clone)?

Answer 8

I have little confidence that things will not be a lot worse than is projected by climate models. Don' t forget that climate models underestimated the amount of sea ice loss that would occur in 2007.

http://www.skepticalscience.com/Arctic-sea-ice-melt-natural-or-man-made-intermediate.htm



Contrary to what denialists say, computer models are not the basis of global warming theory. Scientists knew that models would predict that carbon dioxide would cause warming before they ran the computer models.



Jim Z







Read my response above. I have little confidencein climate models. I accept AGW because of thermodynamics, not models. Maybe I should call myself "Climate Thermodynamics Groupie."



Jim Z



That staement about calling myself "Climate Thermodynamics Groupie" was a joke.



edit



This question is a straw man argument. No one who has answered accepts AGW based on models.

Answer 9

My confidence in the ability of climate models to make precise specific predictions about the future is as limited as my confidence in models that predict where the stock market will be a year from now, but I don't spend my life trying to insult investment bankers or make snide insinuations that the entire economics profession is based on some great leftist conspiracy.



Edit: I do NOT "accept certain climate model predictions as gospel" but I do know, from hundreds of past examples, that JimZ is more addicted to idiotic lies he could never even pretend to prove than almost anyone else on this website.

Answer 10

The climate models have no demonstrable validity.



They are so complex, they have to run on supercomputers. They are so chaotic, the researchers have to make multiple runs, and then average the results in order to get anything stable. There are 50 or so of them; if the climate system were understood to a high level of confidence, there would need to be only one.



The government is funding millions of dollars to the modelers...to try and get them to simulate clouds and their effects better, which, admittedly, they do not do now.



The outputs do not agree with reality; all of them predict a hotspot in the upper troposphere over the tropics, but no one can find it. They predict rising sea temperatures, but, since the Argo system began sending data in 2003, there has been no trend in sea temperatures. The missing heat totals some 6.0 x 10^22 joules over the 9 years, the equivalent of the detonation of more than 200,000 Hiroshima class nuclear weapons--per day--*every* day--for nine years. Nobody can find the missing heat.



The actual temperatures are running below the lowest (modeled) estimates published by Hansen in 1988, and the IPCC.



The models fail miserably when tested with real-world data, both in forecast and in hindcast:



“We conducted a validation test of the IPCC forecasts based on the assumption that there would be no interventions. This test found that the errors for IPCC model long-term forecasts (91 to 100 years in the future) were 12.6 times larger than those from an evidence-based “no change” model...”

—J. Scott Armstrong, PhD, forecasting expert



The models that predict CAGW all include assumptions of large positive feedbacks (involving water vapor) which have never been validated.



The models predict accelerating sea level rise, but inspection of tide guage data indicate that the rate of sea level rise is not significantly different than it was in the pre-industrial era.



“Over the interval 1979 to 2009, model-projected temperature trends are two to four times larger than observed trends in both the lower and mid-troposphere and the differences are statistically significant at the 99% level.”

McKitrick, R., McIntyre, S. & Herman, C. (2010). Panel and multivariate for tests of trend equivalence in climate data series. Atmospheric Science Letters.



Models are not evidence. Models can be made to show anything, to fit anything, depending on the math, and the coefficients.



"With four parameters I can fit an elephant, and with five I can make him wiggle his trunk."

--John von Neumann, as quoted by Enrico Fermi



The nice thing about climate models is that the outputs are so far into the future that they are immune to invalidation. Nothing that cannot be falsified is scientific (see Popper's Principle of Falsifiability).



“No amount of experimentation can ever prove me right; a single experiment can prove me wrong.”

—Albert Einstein



“There is now no observational evidence to support the notion that global warming is caused by carbon emissions. None....The only current reasons for blaming carbon emissions are the predictions of climate models. The hypothesis that carbon emissions are to blame is currently falsified by the observational data. If the scientific method was applied, carbon emissions would not be blamed for causing global warming. The current situation is not the way science should be done. It isn’t science, it’s politics.”

Dr. David Evans, ex-carbon modeller for the Australian Greenhouse Office – Carbon Accounting Section



“It does not matter who you are, or how smart you are, or what title you have, or how many of you there are, and certainly not how many papers your side has published, if your prediction is wrong then your hypothesis is wrong. Period.”

—Richard Feynman

Answer 11

How limited is your confidence in climate models?

My confidence levels are pretty low. Models have their place but you shouldn't believe that a model is the same thing as reality. Clearly models have been used to remove the LIA and MWP by cherry picking proxies and portions of proxies. In the example you gave, it appears that reality failed to follow the alarmist's models. Reality is typically the alarmist's biggest enemy yet if you listen to them, you would think they are all about reality. Climate Realist is a good example. He believes he is a realist. A more appropriate name would be "Climate Model Groopie." I think climate models are limited and sometimes corrupt.



Dook admits that the models are no better than predicting the stock market a year from now but he still accepts certain climate model predictions as gospel. Those who are properly skeptical are deniers.



GCP, I too underestimated the leftist turnout. I figured the worst president in our lifetime wouldn't provide any reason even for his fellow leftists to get out and vote. I was wrong. The takers, the ignorant, and envious haters came out in droves and Romney didn't provide a big enough difference. Thanks again for proving who you are and why you believe. Maybe it wasn't bad polling. Maybe it was this:

http://townhall.com/columnists/rachelalexander/2012/11/11/obama_likely_won_reelection_through_election_fraud/page/2



CR, your contention that you believe only because of thermodynamics is just proving my point. I took Thermodynamics in school. I still don't believe models are reality. Are you really a realist or a groupie as you seem to admit.

Answer 12

Quote by Chris Folland of UK Meteorological Office: “The data don't matter. We're not basing our recommendations [for reductions in carbon dioxide emissions] upon the data. We're basing them upon the climate models.”



Obviously the greenies don't even believe in them for anything substantial. It has proven to be an expensive joke on the public.

Answer 13

VERY, because too many of the variables are filled in with "guesstimates" since we have NO verified FACTS for those variables.