Search This Blog

Tuesday, January 7, 2014

The willfully scientifically illiterate...

Over at Ars Technica there is yet another silly post titled "On Moderation in Climate Discussion."

Ars writes a lot of stuff from Ubuntu Unix to particle physics - most of which is reasonable stuff.

However, they also have a "contributor" named John Timmer who writes on "climate science."

Today I see the post linked above which basically says "this is a politically charged topic about which people rant" so we are tightening our "moderation."

Tightening our moderation of reader comments, hmmm...

I suppose little wonder when you see statements about climate change like this (from the link above): "...scientists all recognize the reality of climate change and take it seriously."

See, its all dogma and clearly if you don't agree there is something wrong with you, dear reader.

So shut up, stop posting, and let us get on with our weekly dogma report (or drum banging for "global warming").

I have written about this contributor before and his commentary and content.

Here is a recent post by him: "Solar variability has a small effect on climate change."

And what do we find here?

Well, climate science apparently has a number of "scientific" models that represent the earth's climate, many in fact (43 as Ars claims in this related article which we will get to in a minute).  Each produces disparate results as they relate to predicting climate.

(NOTE: Unlike physics which has, for example, "the standard model," which, while certainly not perfect represents, at least to me, a much more reasonable model of what a scientific model is all about.)

In any case the "Variability" article goes on to discuss how some people are making, at least to the author, unsubstantiated (or unsubstantiatable) claims "[the sun's output] .. hasn't seemed to have changed enough [over time] to have left a mark on the recent climate record."

As an "answer" to these unsubstantiated claims we are presented with the following example of "climate science" (from the "Variability" article speaking about some climate study):

"..they compared a climate model's output with reconstructions of the Northern Hemisphere's temperatures for the last 1,200 years (Northern Hemisphere data is much more complete than Southern)."

So we have at least 43 different climate models, and we picked one - apparently at random though no mention of the reasoning behind the selection is provided.

We compared this random model to some incomplete data that we happened to have and then adjusted the level of "influence" of solar activity (here it is described in what passes for detail):

"Within the climate model, they set both large and small values for the influence of solar activity on the climate.

And the large values simply don't work very well. With a high value for solar influence, nearly three hundred of the 1,000 years of the comparison failed to line up—the model output failed to match the historical record. In contrast, with a low value of solar influence, the number of mismatched years was cut by more than half.

So, the summary here seems to be that with one model solar "setting" the error rate when comparing to "reconstructed" data is 30% and with another (presumably the one the author thinks is important) its 15%.

Of course, its totally unclear what this means given that only 1/43rd of the climate models where tested against "reconstructions" of other, admittedly incomplete, data.

Timmer's conclusion (underline mine): "...the study is another point against the idea that the Sun's variability has had a significant influence on the historic climate."

So what's interesting here is that the above reasoning, at least in the article's authors mind, this study has some sort of scientific value - a data "point" in what sense I wonder...

Let's compare this to, for example, the detection of a new particle in physics (see this on the Higgs Boson in Wikipedia): "To conclude that a new particle has been found, particle physicists require that the statistical analysis of two independent particle detectors each indicate that there is lesser than a one-in-a-million chance that the observed decay signatures are due to just background random Standard Model events—i.e., that the observed number of events is more than 5 standard deviations (sigma) different from that expected if there was no new particle."

Certainly in many other endeavors, both scientific and engineering, a supposed "choice" between a 15% and 30% error rate where not all the facts/data are known is a meaningless comparison.  Meaningless because you admit up front that you are simply "picking" data from your problem space.

The other article I linked "Nailing down climate uncertainty hints at greater future warming" is equally short on science.  Here the discussion, much as I described above for solar influence, revolves around the effect of "clouds" on climate models.

The conclusion here: "Still, in this case, the models worked exactly how they were supposed to, in that they identified a key process that could then be studied empirically. And, with a major source of uncertainty identified, it increases the chances that we can find the remaining sources and study those in more detail as well."

So, basically, the model is full of uncertainty.

But I thought that "climate science" was decided???

So if I were to post a summary of my conclusion here, that "climate science" is a giant fraud the moderator of the forum at Ars Technica would give me a much needed warning: "...throwing out phrases like "the whole thing is a giant fraud" is a quick way to get a moderator's warning."

So I guess the point of Timmer's articles are to promote dogma - and only the pro "climate science" dogma - all others be damned.

Makes you wonder about "the willfully scientific illiterate..." - doesn't it...?