Comments on: Computer Model Fail

By: Loco

Loco — Sat, 03 Jul 2010 12:21:25 +0000

I’m following the Sunspot Count here: http://www.landscheidt.info/?q=node/50
It’s remarkable how closely we’re following SC5 (the start of the Dalton Minimum). Oh yeah… there’s colder days ahead!

By: gerald.d.conklin

gerald.d.conklin — Sat, 19 Jun 2010 07:24:10 +0000

As it turns out the way to model the sun is not that complicated, the math was has been used for centuries.
this site has the details, and some interesting predictions about sunspot activity.
I’ll wait patiently to hear from Warren and the other more qualified visitors on this.

http://www.landscheidt.info/

By: Hey Skipper

Hey Skipper — Fri, 18 Jun 2010 22:38:48 +0000

Hunter (the Tourette’s one)

Two words: Plate Tectonics.

By: hunter

hunter — Thu, 17 Jun 2010 21:36:08 +0000

Ah, the blinding logic of the sceptic. A model of a physical phenomenon that is not the climate didn’t predict something with 100% accuracy, therefore climate models are all wrong!

“…relying on the models exclusively for proof of CAGW…”

What is “CAGW”? Who relies on models exclusively for proof of it?

“…NOT for predicting how that system will behave in the future”

What a remarkable misunderstanding of models, and indeed of science generally. All of science is about the construction of models. Some might be computer coded models. Others might be equations written on bits of paper. But they are all models. And they are all trying to predict how the systems being modelled will behave in the future.

By: Wally

Wally — Thu, 17 Jun 2010 19:43:06 +0000

Exactly ScepticalGuy,

That’s been a key point driven home by every course I’ve ever had in mathematical modeling. Its a circle. Come up with a theory, make a model, test your model (preferably with a controlled experiment), when your model fails analyze the data to attempt to figure out why, come up with a new theory, repeat.

By: ScepticalGuy

ScepticalGuy — Thu, 17 Jun 2010 05:00:37 +0000

To me, this demonstrates exactly what models are good for. They’re good for trying to explain how a complex system works, and to aid in understanding the interactions and internal processes – NOT for predicting how that system will behave in the future.
IMHO, models are actually most useful when they fail (as in this case) as that shows us what we DON’T know and forces us to investigate more.

By: Todd

Todd — Wed, 16 Jun 2010 23:53:08 +0000

[What Wally said]

By: Wally

Wally — Wed, 16 Jun 2010 21:06:52 +0000

This is exactly my gripe with the models, and relying on the models exclusively for proof of CAGW. If we can’t even accurately model the sun, what good are the models? You can’t get the sun right, you certainly can’t get cloud formation right, things like ocean cycles are in a similar state as the sun (we know of certain cycles, but those cycles don’t explain everything), what do you have right? The Earth’s tilt?