5 ms·

or we've been lucky so far with the predictions, or been able to fit our predictions to match the data. This is just PREDICTIONS. it's not settled science - m

• btilly 13y ago
  The old "weather is chaotic so we don't understand climate" argument. Thereby letting you dismiss what scientists agree on without even looking at their evidence. If you know the science, then you'll understand that predicting exactly what the weather will be on one particular day in one month is fundamentally impossible. But predicting what it is likely to average out to in June in Los Angeles over the next decade is both possible, and routinely done (albeit with significant error bars). There is absolutely no contradiction between these two facts. Let me illustrate with a good example showing how scientists can understand both the predictability of climate and the unpredictability of weather at the same time. I will illustrate with an important example that is not widely understood by the lay public. If you know what the Hadley cells are, you can predict at what approximate latitude we'll have jet streams, and why. (Google that term if you don't know what a Hadley cell is - they aren't that complex.) You'll also be able to predict that the jet streams will cause weather to follow them, resulting in (for instance) a general west to east flow of weather over much of the USA and Canada. The fun part is that if you know more fluid mechanics, you can also show that the exact location of the jet stream is chaotic. There will be bends and kinks of with a predictable wavelength (about 2x the width of North America if you are interested) and a generally predictable amplitude, but entirely unpredictable location. Those bends will cause warm weather to be pulled north in one place, and south in another. (Therefore when the East Coast is unseasonably cold, the West Coast tends to be unseasonably warm, and vice versa.) But you can't predict WHERE those bends will actually be. So this tells us a lot about climate (for instance where it will be wet, where you have rain shadows, etc). It tells us a lot about causes of significant variation, including how big the variation tends to be. But it also says that we can know nothing about which way the variation will go in New York City next October. Your argument boils down to, "You can't even tell me what New York City will be like next October, you obviously aren't doing science here!" Which is simply ridiculous, since in fact the mechanics, average, and average variation are all understood along with the reasons why we cannot get more exact answers! Everything that I am describing is standard and has been known for decades. In fact I learned it from my fluid mechanics text over 20 years ago. Please stop using the, "We can't predict the weather so this can't be settled science" fallacy. It is a cheap cop-out to avoid listening to what actually is known by smart people who have devoted their lives to studying the subject. (Oh, and you learned something kind of fun. The next time you look on a weather map and see the jet stream, you'll know why it isn't running straight across. And just watch - if you pay attention over the next few months you'll notice unseasonable weather on one coast of North America that is the opposite direction of unseasonable weather on the other. And on the weather maps you'll see the jet stream, and know why it happened.)
  • flagnog 13y ago
    It's not a fallacy or a cheap cop out. You're of the opinion that what we know about the weather now is all that we need to know, but it's not. How does the orbit of the moon affect weather? What about changes in the magnetic currents in the Earth's core? Earthquakes? Solar winds? Can you tell me what solar activity will be like in 10 years? "Everything that I am describing is standard and has been known for decades." How many times have we heard this, just before big leaps in understanding have occurred? Several times, just in my lifetime. It's hubris to believe we know that much about the weather.
    • btilly 13y ago
      Comprehension time. I never said that we know everything that we need to. I've repeatedly pointed out that our models have big error bars on them, and there are long lists of things that we need to improve, of which your list doesn't even hit on the most important parts. But the fact that we have significant unknowns doesn't mean that we can't put order of magnitude estimates on our uncertainty, plug those into models, and come up with error bars. In fact that is how those models are developed, and how we develop error bars. "Everything that I am describing is standard and has been known for decades." How many times have we heard this, just before big leaps in understanding have occurred? It's hubris to believe we know that much about the weather. I said that after getting through a description of how the jet stream works, and was clearly referring to that description. I did not say - nor did I imply - that the jet stream is everything to do with the weather or climate. Important? Yes. Everything? No. It is certain that there will be major discoveries over the next 20 years about how climate works. Why is Greenland showing far more melting than expected? How does soot from China affect everything from arctic ice to rainfall in the USA, and how does that affect climate? How much methane will be released as permafrost melts, and what will the climate impact of that be? These are all areas of active research. But I'd be shocked if scientists 20 years from now disagree with scientists from 20 years ago about how the jet stream works.
      • flagnog 13y ago
        Wow, speaking of comprehension - the last sentence that you quoted emphasizes that I was talking about the weather, in general, not the jet stream. And order of magnitude estimates are about as accurate as a neophyte programmers time estimates.
• rayiner 13y ago
  "Chaotic" doesn't mean what you think it means. E.g. There are a lot of things in aerodynamics that are chaotic (especially turbulent flows), yet we can make large scale predictions of how, say, a supersonic wing will behave just fine.