4 ms·

> It is interesting that this science has so many people trying to argue it away. Can they predict a moon eclipse accurately themselves? They still do not doubt

> It is interesting that this science has so many people trying to argue it away. Can they predict a moon eclipse accurately themselves? They still do not doubt the prediction. Can they predict the existence of the Higgs Boson, the effects of general relativity on the signal of GPS satellites, the exact path a solar system probe will take? There are so many things science predicts, but not a lot of people try to argue. Oh, there is SO much that could say about this stuff, but I won't because I've already spent too much time around here. I will leave you with this little nugget, though, because I found it quite hilarious at the time. Some years back I attended a computer conference where one of the folks there was receiving an award for using "actual science" in the production of computer animation. What he'd done was try and take orbital models and other data from NASA and use that to produce animation which correctly showed the path of a spacecraft traveling from Earth to Mars (I think it was), instead of just winging it like most animators would do. Except when he tried to actually use this stuff to generate the animation it didn't work, and in his simulations he ended up missing Mars by a wide margin. After some back and forth with NASA and taking some additional stabs at it, the situation was a bit better but in the end he still ended up having to draw a lot of that orbital path by hand - maybe 1/3 of it. He told some of us that he felt like an imposter because he was receiving an award for something that he felt like he didn't actually do. And I thought this was kind of funny because it seemed to accurately sum up the state of "science" these days. Here you have an authoritative source sending out models and data and such which doesn't appear to actually work, in large part. But, you know, if you just go ahead and fudge it enough to try and cover up the flaws with it then you can still win accolades for that!

kaybe 8y ago
Hm, yes. At least in this case we have proof NASA really knows what they're doing since the probes usually get to Mars. For climate science we will only know in a hundred years or so.
- southern_cross 8y ago
  Yes, but they didn't get those probes to Mars without doing a number of midcourse corrections (heck, you can't even get to the Moon without doing midcourse corrections), so it's not like you can just enter a model and run with it. If you try to do that then you're going to end up way off target. Which brings up an important point about all those climate models. Anybody who actually uses models to solve real-world problems will tell you that you can generally only trust those models so far. For anything long-term (where "long-term" can be remarkably short), you usually have to make the equivalent of midcourse corrections based on actual events. And for the longest time, at least, the climate change folks seemed far too reluctant to do that. And this may still be a problem, because I'm seeing reports now that "our climate models have been remarkably accurate", while others point out that they haven't necessarily been particularly accurate at all. To add insult to injury, there appear to have been several instances now where, when the data and the models diverged, instead of correcting the models they adjusted the data instead, justifying this by declaring that it had a "cool bias" or whatever.

kaybe 8y ago

Hm, yes. At least in this case we have proof NASA really knows what they're doing since the probes usually get to Mars. For climate science we will only know in a hundred years or so.

southern_cross 8y ago
Yes, but they didn't get those probes to Mars without doing a number of midcourse corrections (heck, you can't even get to the Moon without doing midcourse corrections), so it's not like you can just enter a model and run with it. If you try to do that then you're going to end up way off target. Which brings up an important point about all those climate models. Anybody who actually uses models to solve real-world problems will tell you that you can generally only trust those models so far. For anything long-term (where "long-term" can be remarkably short), you usually have to make the equivalent of midcourse corrections based on actual events. And for the longest time, at least, the climate change folks seemed far too reluctant to do that. And this may still be a problem, because I'm seeing reports now that "our climate models have been remarkably accurate", while others point out that they haven't necessarily been particularly accurate at all. To add insult to injury, there appear to have been several instances now where, when the data and the models diverged, instead of correcting the models they adjusted the data instead, justifying this by declaring that it had a "cool bias" or whatever.