Additionally it’s not like these models are untested. Their predictions are continually compared to actual reality, including by taking know data, and asking the model to predict more known data. Pretty basic stuff for testing the capability of any model, and guess what, that data shows these models are pretty darn good.
Everything in this world is complex and chaotic, how cars behave in collisions is complex and chaotic. But no one doubts our ability to use computer models to design safer cars. Equally most people generally trust the one week weather forecast, despite the fact that it’s substantially more difficult to predict weather than it is to predict climate.
Just because weather is chaotic and unpredictable doesn't mean that climate is.
That's a good analogy of my skepticism for AGW. We are saying temps went up 1c over 100 years, taking daily measurements, and comparing them to ice cores which are our measurements of historic temps... Only those ice cores aren't daily- they're thousands of years apart. So the ice cores data is irreparably smoothed in a way the recent data is not.
It’s perfectly reasonable to take daily measurements, average them into yearly measurements. Then compare them to yearly averages taken from ice cores.
The only reason we can even use ice cores for temperature measurements is because they’re have high enough resolution that we’ve been able to observe the impact of global temperature on the most recent layers of ice. It’s not like someone drilled an ice core one day and discovered a bunch of temperatures sharpied down the side. They drilled many ice cores, then spent months correlating observed patterns to temperature records so they could build a model that allows them to understand the link between temperature and observed patterns in the ice.
Don’t mistake a record measuring thousands of years of history for a record that only records every thousand years.
What I am talking about is charts like this from the NOAA: http://www.climate.gov/media/5147
Thes resolution is yearly going back only about 2000 years. Everything beyond that, the resolution is much lower. Where is that data from and don't you see how much "smoother" it is?
The thrust of the underlying article is looking at the exact question you’re asking, can ancient ice records that have a more limited resolution of 20-500 years be compared to centennial or millennial scale data, and can you draw useful conclusions about time periods of that range.
Incidentally, the evidence for AGW, and indeed geologically sudden climate change without precedent, goes far beyond ice cores. Very little in science in predicated on lone observations that can't be cross checked. We have a wide variety of information sources that all paint a coherent picture:
https://en.wikipedia.org/wiki/Global_temperature_record
Measuring historical temperatures is not unlike measuring distance. We know, for instance, that the Andromeda galaxy is 2.537 million light years away. How can we possibly make such a precise assertion so confidently? We started by measuring small, earthly distances, and used those to triangulate larger ones. We worked our way up the distance scale by a series of inferences. Thus, we built up a model of how things look at various distances. It's more or less the same type of methodology - yet mysteriously, the anti-AGW pundits don't seem to have a problem with that.
https://en.wikipedia.org/wiki/Cosmic_distance_ladder
Climate scientists, believe it or not, know what they're doing. It is the height of arrogance to flippantly dismiss their efforts simply because you personally don't understand them, because you can't be bothered to research even basic things like the resolution of ice core data.
That said, we can predict, with some level of accuracy the stability of of orbits in a number of cases. Approximate solutions can be managed with various levels of theory, perturbations, and numerical simulation.
The difference is, Astronomical models can be simplified significantly, whereas climate models must be high fidelity to produce accurate results.
To me the critical question is at what fidelity are the models useful for our (humanities) purposes?
If our purpose is to predict climate and its impact we will need at least +/- 0.1C at a resolution of 100km at a timestep of 1 month. It must be possible to do mesh refinements to determine how accuracy changes with changing resolution, and then combine that with the prior resolution conditions to choose an appropriate mesh.
As far as I know most modern climate models are accurate to within a degree for at least a 30 year range. I haven't looked at the mesh sizes for a while but I'm sure someone has done something similar.
A climate model can’t you whether or not is going to rain in exactly 300 years time, but it can tell you how much rain will fall during an entire year, 300 years from now.
Climate modelling is macro scale modelling to weathers quantum scale modelling. You’re operating at entirely different scales, and further you zoom out, the more you can ignore the minutia and still produce useful predictions.
Additionally taking about absolute time periods is the wrong approach. The real question is now many time steps can your model take before it starts to significantly diverge from reality. A car collision simulation will be operating with time steps measured in milliseconds to nanoseconds, a climate model in time steps measured in days to weeks. Both of these models will probably be predicting a similar number of time steps into the future, and both will have a similar number of quantised simulation areas. For the car you’re simplifying cubes measured in millimetres, for climate, zones measured in km.
Again, climate models have been measured, tested and repeatedly proven their predictive power. If empirical results aren’t enough to convince you, then what is?
Perfect example of this is quantum effects. Zoom in far enough and quantum effects make the world entirely unpredictable, and yet I can still accurately predict where I throw is going to land with no knowledge of quantum effects. That doesn’t mean those effects don’t influence the ball, it just that noisy chaos of those influences is tightly bound enough that it doesn’t matter. Newtonian physics is enough to make a prediction, despite being an incomplete description of the physical system at play.
I.e. we can say thing X has a 90% probability of happening. But we can’t say thing X will happen.
And of course a strong probabilistic prediction is incredibly powerful and useful. I don’t want to suggest that it isn't, but it’s not an absolute prediction. And once again usefulness of a probabilistic prediction is almost always as useful as an absolute prediction (I would also argue that absolute prediction don’t really exist in the real world, they’re just extremely high confidence probabilistic predictions, but you haven’t bothered to model the actual noise).
That does not mean you cannot do it accurately. It means you have to continually refine and improve and throw out and restart and refine and....
Scientists as a group are by far the most honest group of people you can point at. No other profession comes even close. Alluding to some dishonesty in the work behind climate change is straight up denialism. Yes some of the science has not panned out. That's how science works. It's part of the bones.
You can't really model anything "accurately", but you can assign a subjective (probability is subjective) probability to certain outcomes based off an analysis of your model, and what in your best judgement are the correct parameters.
This is why the papers containing uncertainty analysis of varying quality, the press releases will have some handwaving, and the articles in the newspapers will contain nothing of the sort.
Its more than that. Large models have hundreds of adjustable parameters, which each may be tweaked within reasonable ranges to simultaneously backfit historic data and potentially get any kind of future result. It's very easy to get biased outputs which look totally reasonable, which is especially dangerous if the discipline is subject to a rigid orthodoxy.
Have you looked? Because the data is available.
https://e360.yale.edu/digest/extreme-weather-events-have-inc...
The issue is that it's hard to get these across correctly to the general audience. It requires analytical thinking and patience and attention to details, which none are compatible with how today's media and social networks are incentivized and structured (short, catchy, flashy, clickbaity, infuriating, etc.)
It makes no sense to talk about the 'chaotic' nature of a system without relating to the scale and scope of the model. Weather is a chaotic system but this doesn't mean we can't predict it at all what it'll be like an hour later or tomorrow. As we scale it out further, the accuracy decreases but we can predict with high certainty that it's not going to snow in Los Angeles six months out from today.
For example, we have fairly detailed climate related data sets going back many thousands of years, coming from many different kinds of sources. You input these as starting conditions to a model, run the model forward, and then see how accurate its predictions are against later historical data.
Despite your feelings about is or is not possible, there is very strong research that shows that climate modeling works.
We can look at the predictions (inference) from older models and check how accurate they were at predicting the future. No back testing required. Those models have been very accurate. Current models have only improved and reduced error.
We can test the accuracy, you have to be honest about it.
https://www.science.org/content/article/even-50-year-old-cli...
If climate is only model able by chance, we got lucky a lot of times. Someone should be buying a lot of lottery tickets because the science community is way luckier than the expected value. That dozens of independent research teams were about to exactly cancel out errors in precisely the right way (but different from one another) to have good predictive power. Certainty too lucky for a purely random process. So what's the bound? Clearly this process is repeatable.
The models are accurate and science can deal with probabilities. Or rather that scientists (experts) know what they're doing.
The odds on the first conclusion are too high. The second is easier to justify. I think it's easier to justify that you're not as informed on the subject and that there's nuance you're missing that the experts know (I'm confident they know more than me).
If your model includes most of determining factors for the system your modelling, then it’s going to produce pretty good results. There will be places where the model drifts, but you still expect the model to be pointing in pretty much the right direction. As you introduce more factors, your model starts to model details with increasing accuracy, but again you don’t expect it to substantially change the macro result.
I’m short we’re perfectly capable of knowing that the warming will happen without knowing about aerosols, the direction of travel is perfectly predictable, aerosols will only change the amount of travel, and not by an order of magnitude.