[nateweiss]

FWIW, my next-door neighbors, who were dairy farmers for many years, felt quite certain that it would be a very harsh winter this year. Their prediction was based on a few natural indicators: the fact that old apple trees along our road were bearing fruit for the first time in a while; and the coloring of a certain type of furry caterpillar (sadly, like a typical city transplant, I don't recall the details at the moment). There was also something about where birds were placing their nests.

My lovely neighbors are full of observations and predictions like this, sometimes preceded by references to the Farmer's Almanac... what one could call "country wisdom" with perhaps a smile on one's face... but it really seems like they are usually right about such matters. They were so convinced of the harsh 2013-2014 winter that they decided to leave for two months of it, which they haven't done in like 50 years or something--which turned out to be a great call. Of course it could be coincidence or random chance, and I probably don't hear about or remember the times when the predictions are wrong. But I do love that there are other constructs for measuring and predicting weather other than what we typically hear about from the weatherperson on the news.

[dllthomas]

Your neighbors should publish a paper, so we can fold these observations into our predictions, if they correlate with future weather while being uncorrelated with the other sources of information.

Or if it's really that accurate, sell the info to people trading orange juice and become stinkin' rich.

More likely, as you say, it's just confirmation bias. It'd be surprising if the relevant patterns were simple enough for caterpillar and tree biology to pick up on but too complex for our climatologists and meteorologists to spot in decades of PhD students looking for theses. It's not impossible, though.

[d23]

> More likely, as you say, it's just confirmation bias. It'd be surprising if the relevant patterns were simple enough for caterpillar and tree biology to pick up on but too complex for our climatologists and meteorologists to spot in decades of PhD students looking for theses. It's not impossible, though.

Right. Most caterpillars and trees I know have only had their PhDs for 5 years or so.

[dllthomas]

Well, I think it's unlikely, but not as unlikely as your quip (while amusing!) would imply. The computations would be performed on fuzzy caterpillar and apple tree hardware, but the correlations would have been discovered and tuned by natural selection over many iterations.

[lwhalen]

if there's anything this crowd can get behind, it's fuzzy logic and apple hardware.

[dllthomas]

Very nice! I had considered going the fuzzy logic route, but missed the other.

[manicdee]

Well, something that anemometers, thermometers, hygrometers and pyrometers can't tell you is what the air smells like.

If the animals that reacted differently, for example, to air with a different blend of gasses, and that different behaviour caused that animal to survive the surprise snow storm, you'd expect that this behaviour would become more widespread in later populations of that animal.

The genetic "memory" of the cold event would be carried by the surviving population, since only animals with behaviours that contribute to surviving the cold event end up procreating.

Now what if caterpillars change colour in response to atmospheric gasses, temperature, or diet? Would a caterpillar that changes colour with temperature be a suitable indicator of a coming cold snap?

http://mamajoules.blogspot.com.au/2008/10/caterpillars-chang...

So I recommend everyone start a macrophotography habit, keep photographing the insect life in your neighbourhood with geotagging turned on, upload to Flickr, and then the climate scientists of the world will have something to work on.

[dllthomas]

That's roughly how it would happen if it happens. I think it's unlikely that they predicted that this winter would be especially cold because I think the causes of this winter being especially cold are unlikely to be local. A caterpillar in Georgia could respond to the temperature, pressure, atmospheric mix, humidity, or whatever in Georgia; it cannot respond to, say, pressure differentials between Georgia and New York.

[ackydoodles]

Your lovely neighbours cannot be correct except by chance; if there was any validity to their predictions, it would imply not just that they knew the future of the weather, but that the future of the weather was knowable, which it certainly is not. Your tendency to see their predictions as accurate is likely a conflation of your confirmation bias and your choice-supportive bias.

Humans naturally search for meaning in chaos. That is all that their observations of caterpillars and trees and birds amounts to. They are trying to attach significance to a chaotic universe in order to reassure themselves that their place in the world makes sense. It does not. This is mathematically verifiable.

If the birds and the caterpillars and the apple trees could know the future of chaotic systems, it would imply that we live in a deterministic universe where birds and caterpillars and apple trees could not exist. QED.

[dllthomas]

"If the birds and the caterpillars and the apple trees could know the future of chaotic systems, it would imply that we live in a deterministic universe where birds and caterpillars and apple trees could not exist. QED."

Wut?

[ackydoodles]

Complex dynamics 101: life is a complex dynamical system. If the math of chaos is wrong, the processes underlying life would not operate as they do. Life would not exist as we observe it.

Also, your iPad would not work. As I said, QED. This is not controversial.

[Phlarp]

For something so "not controversial" that it doesn't require citations, your comment is pretty faded.

[nitrogen]

The Mandelbrot set is both chaotic and deterministic. QED

[Oletros]

> Also, your iPad would not work.

What?

[prewett]

Or, perhaps that weather has cycles and is not an uncorrelated random walk, and that 50 years of living with your life strongly influenced by the weather, along with several lifetimes of your parents' wisdom gives some insight into deep patterns of nature that are not readily apparent. Perhaps these persistent stories about old people predicting things based on caterpillars and trick knees are hints that the world might be more complicated than you think. And if you aren't so quick to assume that you understand the world because you think you understand a theory, you might investigate these hints. Maybe you will find out that there is something actually going on, and you'll have an exciting new world to explore.

[rayiner]

"Chaotic system" does not mean what you think it means.

[evantish]

"it would imply that we live in a deterministic universe..."

You say this as though it's impossible. You may not believe we live in a deterministic universe, but, that's just, like, your opinion, man. The rational stance (I think) is to: (1) admit you don't know, (2) acknowledge the possibility of a deterministic universe, (3) for whatever chance the universe is deterministic, delude yourself into believing that it's not, because the illusion of free will is probably good for your mental health.

I think you're doing #3 really well, but skipped over #1 and #2.

[DerKommissar]

Just because a system is chaotic doesn't mean nothing about it can be predicted.