I often wonder why the global warming doom-gloom-soothsayers have so much traction in the world. Like right wing conservatives, they like to claim that they are victimized by a hostile establishment press, but the NYTimes, a pillar of the establishment, is certainly with them. Check out the 230 comments on Andy Revkins DotEarth blog regarding the recent email disclosures from the CRU. The Editors’ Selections, with the purpose of
…highlighting the most interesting and thoughtful comments representing a range of views.
includes 4 posts, all firmly in the camp of “How dare they publish this! This is just normal science. Face it, global warming is a fact!!” So much for a range of views…but no matter.
But why do intelligent and scientifically literate people, including some who are quite reasonable, e.g., Andy Revkin, feel so confident that the AGW hypothesis has been established beyond doubt? Frequently – check out those Selections – references are made to mountains, avalanches, piles…etc. of data that prove the point. I think something is missing here: I think it is the global circulation models (GCM) run on super computers that clinch it. But there is very little peeking into those models – they are essentially a black box for most people: numbers go in, Apocalypse comes out!
Without the models, there would be no terrifying scenarios, disturbing graphs showing steeply rising temperatures over decades to come, no tipping point doomsday model runs. There would be some hard data (CO2 rising), a mountain of ice core, satellite, and surface data from which some would infer a clear trend, correlation, and causal mechanism; there would be an interesting hypothesis about positive feedback amplifying the otherwise manageable temperature rise that might be caused by CO2 increases and that might or might not happen; there would be the same endless scientific haggling and argument over the way the numbers are handled by statistical routines and whether this or that presentation of the data is appropriate and meaningful; there would be no consensus. The advocates of AGW would be a determined and inventive bunch, but they would be hard pressed to demonstrate that the rest of the world should abandon the null-hypothesis, i.e., climate and CO2 have always fluctuated- what’s so different now? - and adopt their hypothesis. Computer models change all that.
The GCMs give the AGW crowd the cover to say that they can predict (not with certainty, of course…) the future trend of the climate. It gives them the supposed justification for stating that they have uncovered the “forcing function” that precisely quantifies the impact of CO2 concentrations on the climate. It provides them with a rationale for assserting that their understanding of feedback mechanisms is corrrect and that their predictions are reliable. This role of computer models is not often examined, rarely questioned, certainly not in the popular press.
It’s worth taking a look at the writing of Daniel Botkin, a scientist who was present at the creation of computer modeling in ecology, and who has a lot to say on the role of models in scientific investigations. His basic point is that models are valuable tools for understanding a natural system, for trying out ideas of how changes in one thing may affect another, but they are not very good for making predictions. His essay, Science and Soothsaying, is a good starting point.
Another critical view of computer modeling is the Pilkey’s book Useless Arithmetic. Orin Pilkey (not to be confused with the climate scientists father and son, Pielke Sr. and Pielke Jr., also with a jaundiced view of modelers’ work) is most known for his controversy with the US Army Corps over its penchant for pouring millions of dollars into pouring sand on eroding beaches. These wasteful projects are often supported by very impressive computer modeling.
In thinking about this topic, I keep returning to a book published almost twenty years ago, Ice Time. In its chapter, The Machine’s Eye, the author makes the point that the study of climate had become, in large part, the study of climate models. He traces the rise of supercomputing in the investigation of climate, and notes that it has become “big business.” The author is relatively uncritical of the use of the models, but he focuses more on their use to understand the mechanics of the climate system rather than to predict the future. The chapter is the only extended discussion in layman terms that I have ever seen of just what computer models of the climate do, and how they are put together. For that, it remains a very useful discussion.

Posted by lichanos 
Posted by lichanos 


Posted by lichanos 
























