Comments on: Computer Models

By: Paul Hanlon

Paul Hanlon — Sat, 11 Oct 2008 01:09:24 +0000

Four words – Long Term Capital Management. This was a company that boasted two Nobel Laureates on the payroll with supercomputers to analyse investment opportunities, that almost brought down the financial system in 1998. I actually believe in the ability of computers to help us figure really complex things out, but they are only as good as the data they are fed, and these climate models are not factoring in anything like enough variables to get useful data out. I’m sure I read something on RealClimate that said that 12 variables are used in calculating their models. Even if they break these down into smaller subsets of related data, this isn’t anything like enough to get an accurate assessment of something as complex as climate.
paulhan

By: Frank

Frank — Mon, 06 Oct 2008 19:48:14 +0000

Looks like they do a good job of hindcasting only with CO2. http://www.skepticalscience.com/climate-models.htm

By: Scottar

Scottar — Sun, 05 Oct 2008 05:37:45 +0000

http://news.bbc.co.uk/1/hi/sci/tech/6320515.stm

Models 'key to climate forecasts'

By Dr Vicky Pope
UK Met Office's Hadley Centre

"The only way to predict the day-to-day weather and changes to the climate over longer timescales is to use computer models."

"Now, current state-of-the-art climate models include fully interactive clouds, oceans, land surfaces and aerosols, etc. Some models are starting to include detailed chemistry and the carbon cycle."

"Much longer predictions are run, typically reproducing the last 150 years and predicting the next 100 to 1,000 years"

This is what the Hadley Center claims but their rants don't include the recent cooling or they gloss over it. It may take several cold dark winters to have them back to square one scratching their posteriors. But by then many people will be suffering from lack of energy supplies and costs and Congress will be in the deep shit dumpster.

McCain needs a wheel alinement as he still supports the Cap and Trade proposal and wants to include AlGore in his cabinet. What a nut job. I think I'm going into hibernation for the next 4 years.

By: Mesa Econoguy

Mesa Econoguy — Thu, 25 Sep 2008 17:55:55 +0000

Computer "models" are wrong all the time, which has to do with the limits of statistics, and the underestimation of the probability of extreme events.

Unlike climate models, asset-pricing and transaction models can actually be tested for accuracy and effectiveness.

No climate model would pass Wall Street muster.

By: RPJ

RPJ — Thu, 25 Sep 2008 15:54:29 +0000

Al Gore has argued that computer models can be trusted to make long-term forecasts, because Wall Street has been using such models for years - where has he argued this?