Hacker News Reader

Comment by dylan604 | Hacker News Reader

I'm never unimpressed when we can make these kinds of advanced notices. I know it's just math, but there's a lot of objects out there to run all of the simulations against to see what will be where when while in our line of view.

View on HN

All stars are fixed point-like, basically static coordinates, so it's more of when one object encroaches another within a few micro-degrees.

3-body problem implies crazy perturbations and chaos theory and invoking of pendulum analogies, and all that's true, but due to the magnitudes of difference between human timescales and cosmological, we know the eclipses, occults, and all other matter of quirky alignment and other miscellaneous minute for millennia to come.

I can’t quite put my finger on why, but it blows my mind that its that “simple” in the sense that all the factors can be accounted for despite being incomprehensibly far away. My instinct is that there would almost certainly be some tiny unaccounted for perturbation that throws everything massively off over huge distances yet that’s not the case.

I can guarantee "simple" comes at the cost of (probably) multiple tens of thousands of pages of thrown away scratch paper (or equivalent) over many years and lives.

It is, it is just so far into the future that it seems trivial.

The Voyager spacecraft, when being gravity assisted by Jupiter, stole enough energy from it, that in 5 billion years, Jupiter will be 2 millimeters behind in its orbit around the sun.

All the stars we can see with our naked eyes are close enough to that, yes.

But you did make me think of this: https://www.eso.org/public/videos/eso1825e/

it's interesting seeing all of those stars in such random orbits instead a unified orbit where everything is going in the same direction. That one star really looks to be moving, and is pretty much the only way to know where the "center" is. Everything else looks like blobs floating in a petri dish.

I’ve been looking for a resource for a long time but haven’t found much. What I am curious about is how people 500+ years ago predicted solar / lunar eclipses (for example). Columbus was able to persuade some islanders that he was some sort of god by making the moon darken/turn red at a predicted time due to an almanac he happened to have.

Would I be able to make the same prediction using only math and whatever instruments could measure the positions of the sun and moon that accurately? What instruments and equations do I use?

If you're starting from scratch you would need some way to determine the apparent position of the Sun and the Moon. Probably some kind of long tube that you can rotate precisely. Ancient people would use buildings for that purpose. Then you need to make observations probably over several, possibly many, years to determine the Earth's perihelion and apohelion (at least their timing), the length of the year, and the precession. For the Moon you determine its orbital period and plane, and if you want to know whether the eclipse will be partial or total from a given location you also need the perigee and apogee (again, at least their timing).

Once you have all the data you find all the times when the Moon crosses the ecliptic plane (about once every 14 days), which are the only times when eclipses can happen. You discard the times when the Moon will not be aligned with the Earth and the Sun. You will be left with a handful of times per year when an eclipse is possible. If the eclipse is solar you compute the orientation of the Earth to see where it will be visible from. It's all trigonometry and linear algebra.

My understanding is that at Columbus time they did not measured sun / moon positions, but instead based their predictions on the eclipse cycles - if you have long enough record of past eclipses, you can notice that they are not random but the same timing sequence repeats after some time:

https://en.wikipedia.org/wiki/Eclipse_cycle

What are the odds we have a repeat of a substantial celestial object hitting Earth? Like devastatingly so, can that be quantified in some sense?

We have the (incomplete) track record of hits from the last billion years or so to extrapolate from.

Don't know that there has been much since that Dino Doom Rock 65m years ago?

Ever see notices of an asteroid making it's closest approach? Sometimes, those are closer than the moon is to earth. Sometimes, they're further out. There does exist a team tracking objects for the sole purpose of deciding if there is a potential of a collision path.

https://eyes.nasa.gov/apps/asteroids/#/home

Has some fun graphics to show what we know about. The scary part are the ones we don't know about. Some say that this groups budget is not enough.

Reply on news.ycombinator.com