5,580
edits
Changes
Jump to navigation
Jump to search
Line 197:
Line 197: + + + +
Talk:Science of global warming (view source)
Revision as of 01:17, November 5, 2007
, 01:17, November 5, 2007→Section Needed on the Saturation Problem
::::I can't say I know definitively that they're using logarithms, because I've never seen them publish their models. I can say the evidence suggests otherwise. Firstly, I'd think it odd if every model-builder is using a logistical curve, but they uniformly say they're using a logarithm. Surely someone would actually use the right word? Secondly, logarithms are one of the four basic curves that are normally used to build mathematical models; logistical curves aren't on that list. Finally, while both curves would tend to fit the data pretty well for lower levels of greenhouse effect, they would diverge at higher levels. In fact, the models have over-predicted recent warming, and had to be tweaked downwards--that is, they've behaved exactly as you would predict if they'd mistakenly chosen a log function. [[User:QBeam]] 2:25pm Nov. 2, 2007 (EDT)
::::I can't say I know definitively that they're using logarithms, because I've never seen them publish their models. I can say the evidence suggests otherwise. Firstly, I'd think it odd if every model-builder is using a logistical curve, but they uniformly say they're using a logarithm. Surely someone would actually use the right word? Secondly, logarithms are one of the four basic curves that are normally used to build mathematical models; logistical curves aren't on that list. Finally, while both curves would tend to fit the data pretty well for lower levels of greenhouse effect, they would diverge at higher levels. In fact, the models have over-predicted recent warming, and had to be tweaked downwards--that is, they've behaved exactly as you would predict if they'd mistakenly chosen a log function. [[User:QBeam]] 2:25pm Nov. 2, 2007 (EDT)
:::::First of all, how can you say "the models aren't built on first principles" if you've "never seen [any published] models?" (I presume you are talking about predictive models; descriptive models are ''not'' based on fitting the data). Like I said before, I don't know that they ''aren't'' using a logistical function, I just know (with my measily 3+ years experience in a major which has a very small focus on modeling techniques) that no scientist worth their salt would use an entirely wrong function, just because. If you can point me to any evidence of this, I'd love to read it.
:::::At any rate, how do you know they are using the wrong function at all? If scientists who study these sorts of things for a living all use a logarithm, how do you know they are wrong? [[User:HelpJazz|Help]][[User talk:HelpJazz|Jazz]] 20:17, 4 November 2007 (EST)
:*The problem is, you are assuming good faith. Unfortunately on the Internet that is at the top of the charts for never happening. QBeam was stringing together buzz words and silly science in the hope they would lend creditability to his/her post, a fact you just proved, HelpJazz. --<font color="#1E90FF" face="Comic Sans MS">[[User:TK|şŷŝôρ-₮K]]</font><sup><font color="DC143C">[[User_Talk:TK|/Ṣρёаќǃ]]</font></sup> 01:01, 20 October 2007 (EDT)
:*The problem is, you are assuming good faith. Unfortunately on the Internet that is at the top of the charts for never happening. QBeam was stringing together buzz words and silly science in the hope they would lend creditability to his/her post, a fact you just proved, HelpJazz. --<font color="#1E90FF" face="Comic Sans MS">[[User:TK|şŷŝôρ-₮K]]</font><sup><font color="DC143C">[[User_Talk:TK|/Ṣρёаќǃ]]</font></sup> 01:01, 20 October 2007 (EDT)