Search This Blog

Friday, May 25, 2012

Wolfram and a New Kind of "Climate" Science

I found another interesting article on "climate science" at Ars Technica today written by one of the same authors that I wrote about in "AnthroBioBlame: A New Kind of Sience." a while back (I think the intellectual pun on Stephen Wolfram's "A New Kind of Science" was lost on the authors).

I think that a little background on Wolfram and his "New Kind of Science" is in order first off.

Wolfram is a physicist and mathematician.  About ten years ago he released a book on what he termed as literally "a new kind of science."  His main thesis in the book is "computation irreducibility."

Computational irreducibility means that there are systems in the nature, the world, and the universe for which the only mathematical description possible of that system is the system itself (see this chapter).  The book also covers cellular automata and other examples related to how he discovered this principle.

Now Wolfram's conclusions are by no means "accepted science" yet until there is some plausible argument against this thesis one must believe that it may not be possible to reduce certain complex systems to predictable mathematics.

Wolfram's book points to what I see as the largest and most fundamental flaw in "climate science" - basically the notion that simple human "climate models" have any predictive nature whatsoever, i.e., that the system of climate on our planet is computationally irreducible.

What does this mean?

It means that while we can run around taking as many measurements as we like of the climate process on earth the only way to know what the climate is going to do in the future is to watch it evolve into its future state.

There is no amount of predictive computational horsepower to do otherwise.

One must conclude this after reading Wolfram's book.

Secondly climate is an "open system."

This means that you cannot simply measure the temperature of things and make a prediction.

Why not?

Because things like asteroids and volcano's, which are not normally part of the climate, pop up periodically and introduce new and unforeseeable consequences.

An example of what Wolfram's book says to me is summed up in the image at the top of the article.  I have seen many like it over the years - some with smaller time scales, some with differing temperature and C02 scales - but similar layouts none-the-less.

The image implies that somehow someone somewhere was able to create an accurate and reliable "global temperature/CO2" plot over the history of the earth.  Here is another one commonly shown:

Now clearly during the last 400,000 years there hasn't been consistent temperature monitoring of the planet (nor for the 600 million as the other chart shows).  This data is extracted from indirect measurements - for example Antarctic ice cores - or from what are often government-funded "educated" guesses.

(Now, depending on your perspective we're either headed for CO2 armageddon or warming up from a prior disastrous global temperature plummet.)

So while these may be the levels shown indirectly we have no real way of knowing what the causes and effects were.

Were there outside influences?  Where there geological influenced?  No one can say.

Third "climate science" itself is not actually science in the sense that the "scientific method" is followed.  There no, at least to my knowledge, theory of climate which predicts future behavior based on some past data and a model.

I suspect that this is true for exactly the reasons Wolfram writes about - its simply not computationally possible.

So while climate scientists can say what they like about man's affect on climate no one can be sure if its true because we don't have a fundamental understanding of climate in the first place.

Now when I say "not a science" I am talking about those making bold predictions about future consequences based on "climate science."

Certainly there is much value to having satellites and weather prediction and  so forth going on - it certainly adds to our base of knowledge.

But there is no hard data on the past, only guesswork and inference.

Its the certainty of the conclusions that I find troubling.

I don't expect that "climate scientists" will agree with my thesis here.  Primarily because things like Wolfram's book do not fall under the purview of the kind of things they study.

But Wolfram's ideas are along the lines of others who have made other interesting discoveries about the tools of science, e.g., Kurt Gödel.

Now at a secondary level we can run around worry about the short term future, i.e., as it says in the Ars Technica article.  For example the discussion that satellites "imaged the Greenland ice sheet—again, not because of some sort of bias, but because the sheet is very big and very significant."

This is another flaw in the climate science "chicken little model."  The Greenland ice sheet is significant to climate science because, if it all melted, the ocean levels would presumable rise (by presumably I man assuming that all the water goes into the ocean and not, for example, into the air as water vapor).  Yet some how these conclusions of what to study is not bias.

Antarctica is significant too in that regard, but so are many other things: asteroids, volcanoes, and so forth.

The bottom line for humanity is that the future is unpredictable - most likely because of outside factors and computational irreducibility.

"Climate science" says that "we might be killing ourselves so we need to do X."

But where is the research to support X as a safer alternative? 

Maybe X will be worse than the present...

How does anyone actually know and how can they actually know it?

And finally, another measurement of the quality of science is can the hypothesis and theories be used to make accurate predictions, i.e., scale up.

If you are reading this on a computer screen then you are relying on hard science and engineering.  Climate science is not yet at that point so even if we did understand it what could we do with that knowledge?

Some have discussed, for example, deploying large solar reflectors to "shield" the earth from the sun in order to cool it.

Interestingly though, you never see these sorts of solutions accompanied by any sort of hard evidence or mathematics that will make the consequences predictable.

No comments:

Post a Comment