Inter-Galactic Memo
To: All Personnel
Fr: W. Leavitt
Re: Acidic Oceans
8-24-2010

Well, they’re at it again. Some fool-hardy scientists, more full of themselves than a bloated mosquito, have released findings from yet another research project involving Computer Models. This time we are being warned that the oceans are becoming more acidic and IT’S OUR FAULT!

Here’s the relevant quote:
According to this research, a decrease in pH means an increase in acidity. In 1750, the global mean ocean surface pH was at 8.2, and now it is at 8.1. If carbon dioxide emissions are not cut, the researchers' simulations predict that the pH could decrease to as low as 7.7 by 2100. On the other hand, if carbon dioxide emissions are controlled, the simulations predict that the pH won't fall below 8.0 by 2100. Research indicates that there will be an emissions peak in 2016, then it will decrease by five percent each year after.

Fascinating. No, really. Translated, it means “we know next to nothing about ocean acidity and it’s phases, but this model makes it sound as if we do.” I wonder how many factors they were able to program into their model. 10? 50? 100? And I wonder how they chose these factors? By committee? The head of the project? An RPG die? These are important questions because we know on the face of it that the vast majority of factors involved in a process as complex and lengthy as this one, remain undetected, unthought-of, and unknown.

Let’s be nice and give them 100 factors. This is a lot for a computer model, but they’re getting better at it all the time. Now let’s estimate (a technical term meaning “speculate”) that there are actually 10,000 +or -. I’m guessing it is more likely to be greater by a factor of ten, but that’s just me being cynical. How accurate a picture are we likely to get with a ratio like that?

Here’s a handy analogy. One hundred reasons (evidences) to commit murder are probably sufficient for a conviction, and are all the police would bother discovering. But for a forensic psychologist, or sociologist, enough digging, research, experimenting and hypotheses would likely reveal a lifetime of complex interconnections and decision-paths leading up to the murder, which would offer a completely new and different story. Much more thorough, and useful, from a predictor standpoint, as well as medical, in terms of treatment and prognosis. But she probably still did it.

Once again, I have no complaint with modeling complex systems on cool Macs, with those sleek design features. They are a useful tool, a powerful weapon in the arsenal of science. But they are not reality. Not the real thing. And they do not inform to the extent that they should be substituted for reality, especially by really smart people who should know better.

The modeling is fine. It should be combined with lots of other things in order to make educated guesses on the way to that elusive goal of “actually knowing.” And it’s okay to come and say, “hey, this is what we’re studying, and this is why, and this is what we think might be happening, but it could be this as well.” Instead, the preferred method these days is to release the findings prematurely, and in isolation, to some faction of the press, in this case usually an online geek-parade like Daily Tech. (I love Daily Tech.) One wonders what the actual motive is for such behavior. It is hardly professional. Has little to do with the scientific method. Is often politically-motivated. Absolutely inappropriate. And smacks of a new kind of über-geek narcissism.

And as long as they keep having the bad taste of doing it this way, I’m going to keep calling them on it. (Until I’m proven completely wrong by a precocious ten-year old.)