Are computer models data?
Modeling is valid science, for example, we'd never have developed nuclear weapons without them. The real question is are these models valid and I'm not competent to judge that (are you?), so I trust science to learn, test, and refine them.
Modelling is NOT data. Or do you not understand that simple fact? Data is OBSERVED, models are
created, thus they contain whatever biases are programmed into them. Thus they are NOT valid. And yes, with a PhD in geology I am very competent to judge them, and have even used them in my own work. However, when we use them they are very specific, and extremely focused.
Take a look at every single climatology study and you will see this term used in the Abstract "We use a simple model".....So they are admitting to using simplistic computer models to try and recreate the most complex engine on the planet, namely the climate.
Put another way, climatologists have yet to be able to come up with a model that can recreate what happened yesterday. A model that can't do a hindcast is worthless. That too is a fact.
So we can agree that modelling is valid science? As for the accuracy of models that is another issue. Geologists have modeled the earth's interior. Have they ever been there? No but their models generate data that fits what little information does exist from earthquakes. Scientist will use the best models available even though none are ever going to be 100%.
No. Models are not valid science. They can be a
tool, used by scientists, but anyone who bases the body of their work on models is a fool. We actually have a lot of data about the crust, mantle and interior of the Earth. No, we have not been there, but we can use sound waves to map out what the interior looks like. We have ophiolite sequences on the surface that we can dissect to generate an understanding of the rock that makes up the mantle, thus we have physical examples that we can then use to test the models to see if they are valid.
That is a mountain of difference between the models that we use, and the models that climatologists use. They have NO means of testing them for validity, they just assume they are, but when the models have been tested against real world observations they are shown to be 100% (no, that is not an exaggeration) wrong.
The most complex models on the planet are those used by Formula One race teams. They are called Computational Fluid Dynamic models and they are concerned with very few variables. They cost tens of millions of dollars to create, and they cost tens of millions of dollars to operate, and more than 99% of what they produce is trash. The modelers will take a part, and they will develop thousands of design changes to that part which they will then run through the model. Out of over a thousand runs they will chose maybe three that show promise which they will then make and test in a wind tunnel. Most of those then fail. For every 10,000 parts that are modeled, they will produce one that actually gives a race benefit.
And that is from the most complex, expensive, and well run computer model systems in the world.