[FRIAM] Climate Change

[uǝlƃ ☣]

Well, I mean "models" writ large.  Even when gathering and reducing observational data, there's a workflow for doing that. That workflow relies on a model of a sort.  And integrating different data sets so that they're commensurate also requires models.  E.g. correlating tree ring based with other climate data.

But you're ultimately right.  It's not so much about the models as it is the whole inferential apparatus one *might* use to drive policy decisions, including huge populations of expert climatologists.  There's probably a correlation to be drawn between people who distrust government and those who distrust the "scientific establishment" and/or the "deep state".  People tend to obey/trust whoever they regard as authority figures (e.g. greater shocks to another if a person in a lab coat tells you to do it).  Those of us who inherently distrust authority figures have a particular psychological bent and our impulse can go the other way.  It could be because we know how groups can succumb to bias, or how errors get propagated (e.g. peer review), or whatever.

*That* is why I think focusing on the workflows (modeling) is important.  Those of us who distrust the experts bear the burden of proof.  Hence, we have to really dig in and find the flaw in the experts' thinking.  To do otherwise is irrational.

Those of us who can delegate and tend to trust experts only need to dig in when/if a skeptic produces a defensible counter-argument.  If all a skeptic has to offer are blanket generalizations about human error or whatnot, then it seems rational to ignore that doubt and go with the conclusions of the experts.

If Pieter knows of a specific flaw in the way the experts do their work, then it would be a valuable contribution.

On 12/29/2017 12:41 PM, Jochen Fromm wrote:
> IMO it is not about models. Models are complicated and controversial. Climate change in the artic is a fact, melting arctic ice is a fact, melting glaciers is a fact. In the arctic regions we can oberve the rising temperatures most clearly.


-- 
☣ uǝlƃ