U.S. has its wettest 12 months on record again • Hacker News Board

Re: U.S. has its wettest 12 months on record again

#21

Post by btilly » Tue, Jul 09, 2019, 10:22 PM UTC

billions wrote: post #4

mirimir wrote: post #2

Global climate change => higher air temperatures => more water in the air => more water available to fall out of the air.

The most powerful supercomputers in the world are focused on weather analysis and they can only project a few days out. Yet, some people think they've got it figured out.

The widespread willful ignorance about the differences between modeling weather and climate was amusing to me 20 years ago. But the joke is long past its sell-by date.

No, we can't predict whether it will rain on this day in August in Augusta Georgia. Detailed forecasts will probably never get a lot better a month out. But we can predict that December 2019 will be colder than August 2019 across most of the USA, and that 2020 will be warmer than 1980 was. And those forecasts are getting better over time.

That's reality. Deal with it.

Re: U.S. has its wettest 12 months on record again

#22

Post by fromthestart » Tue, Jul 09, 2019, 10:22 PM UTC

6d6b73 wrote: post #6

billions wrote: post #4

Earlier quoted context omitted.

The most powerful supercomputers in the world are focused on weather analysis and they can only project a few days out. Yet, some people think they've got it figured out.

The most powerful supercomputers in the world are not even close in processing power of a 4 year old. Isn't it possible that some people can see patterns that the supercomputers can't?

That's simply not relevant, nor true.*

Modeling climate change is much more complex than the simple heuristic pattern matching of the human mind. You're talking about a huge, complex, and fundamentally chaotic system; the latter property precludes the usefulness of heuristic based predictions by a 4 year old (or a human of any age).

There's a reason we run massive numerical simulations on multi billion dollar HPCs.

*Depending of course on how you define processing power. Sure, in terms of raw flops it is estimated that human brains are indeed orders of magnitude more performant than supercomputers, but for the type of operations required to simulate something like climate, humans don't come anywhere close to the accuracy or precision of results you can obtain from a gridded 3D model fed into a purpose built simulation - without which you really can't answer any but the most basic of climate change questions.

Re: U.S. has its wettest 12 months on record again

#23

Post by xmichael999 » Tue, Jul 09, 2019, 10:23 PM UTC

gxx wrote: post #15

I wonder if this will help to refill some of the depleted aquifers.

Yes, its about the only thing floods are good for.

Re: U.S. has its wettest 12 months on record again

#24

Post by ceejayoz » Tue, Jul 09, 2019, 10:24 PM UTC

justwalt wrote: post #5

billions wrote: post #4

Earlier quoted context omitted.

The most powerful supercomputers in the world are focused on weather analysis and they can only project a few days out. Yet, some people think they've got it figured out.

That does put things into perspective, a bit.

Nah, it's a common talking point, but it's highly flawed.

It's like claiming that because we can't predict if the stock market will go up or down tomorrow, we can't determine whether we're in a recession or a boom time.

Short-term weather is fairly chaotic (and we've gotten a lot better at predicting it over the last 50 years). Long-term trends are different.

Re: U.S. has its wettest 12 months on record again

#25

Post by rolltiide » Tue, Jul 09, 2019, 10:24 PM UTC

mirimir wrote: post #2

Global climate change => higher air temperatures => more water in the air => more water available to fall out of the air.

Its not that I disagree, I just want to point out this never explains what was going on the last time the records were hit and by proxy a snapshot of the current record

Re: U.S. has its wettest 12 months on record again

#26

Post by fwip » Tue, Jul 09, 2019, 10:25 PM UTC

trentnix wrote: post #13

billions wrote: post #4

Earlier quoted context omitted.

The most powerful supercomputers in the world are focused on weather analysis and they can only project a few days out. Yet, some people think they've got it figured out.

Not to mention we have an extremely limited data set to work from when identifying what “normal” is.

We're not talking about vague shifting definitions of "normal." We're talking about "the wettest 12 months ever recorded," in the 125 years we have data for.

Re: U.S. has its wettest 12 months on record again

#27

Post by ArtDev » Tue, Jul 09, 2019, 10:25 PM UTC

gxx wrote: post #15

I wonder if this will help to refill some of the depleted aquifers.

Expect flooding. So.. maybe?

Re: U.S. has its wettest 12 months on record again

#28

Post by conception » Tue, Jul 09, 2019, 10:25 PM UTC

6d6b73 wrote: post #14

chrisco255 wrote: post #8

Earlier quoted context omitted.

Sure, but those supercomputers aren't autonomous, they're programmed and designed by thousands of human adult experts.

And these people can make mistakes or be biased. If we can't predict what the weather in our hometowns will be how can we predict how will the climate change?

You can't predict where a bubble will form when boiling water, but you can predict that the pot will get to over 100C.

Re: U.S. has its wettest 12 months on record again

#29

Post by fwip » Tue, Jul 09, 2019, 10:26 PM UTC

6d6b73 wrote: post #14

chrisco255 wrote: post #8

Earlier quoted context omitted.

Sure, but those supercomputers aren't autonomous, they're programmed and designed by thousands of human adult experts.

And these people can make mistakes or be biased. If we can't predict what the weather in our hometowns will be how can we predict how will the climate change?

If you drop a glass full of water, I can't predict the exact splash pattern. But I can sure tell you the floor's gonna get wet.

Re: U.S. has its wettest 12 months on record again

#30

Post by trentnix » Tue, Jul 09, 2019, 10:26 PM UTC

btilly wrote: post #21

billions wrote: post #4

Earlier quoted context omitted.

The most powerful supercomputers in the world are focused on weather analysis and they can only project a few days out. Yet, some people think they've got it figured out.

The widespread willful ignorance about the differences between modeling weather and climate was amusing to me 20 years ago. But the joke is long past its sell-by date. No, we can't predict whether it will rain on this day in August in Augusta Georgia. Detailed forecasts will probably never get a lot better a month out. But we can predict that December 2019 will be colder than August 2019 across most of the USA, and t…

Not really. The delta in how data is collected in August and December is much smaller than 1980 to 2020. So such determinations aren’t as straightforward as you imply.

Not to mention none of the catastrophic climate consequences predicted in 1980 came to pass, but somehow we are supposed to be so much more certain about the prognosticators of today.