Hacker News Reader

Comment by firesteelrain | Hacker News Reader

Not sure how that will work since JWST orbits the second Lagrange point.

It would be like asking for the GPS coordinates of the moon. NASA did test receiving weak GPS signals on the moon with LRO in 2023. It wouldn’t be useful for navigation though (not yet unless someone has like a way to do reverse GPS on the moon but not sure how that would work)

Reason this works for the ISS is because of the subsatellite point. It can receive GPS signals regardless of altitude above the Earth’s surface.

Also TLEs apply to the ISS because it’s earth orbiting.

TLEs are designed for satellites in Earth orbit, where they define position and velocity using orbital elements interpreted by models like SGP4.

View on HN

> It would be like asking for the GPS coordinates of the moon

No problem at all, just give the location where the moon is at the Zenith and use the distance as the altitude.

> Reason this works for the ISS is because of the subsatellite point. It can receive GPS signals regardless of altitude above the Earth’s surface.

No, wether the object can actually receive GPS signals is completely irrelevant to wether its location can be described in the GPS coordinate system.

You could describe the location of the Sun in GPS coordinates too, the altitude value would just be very large.

You can use GPS to describe a point on Earth. To use the moon or sun is kind of weird because of their size to use GPS coordinates for this

I was referring to finding your position on the moon using Earth referenced GPS signals.

> You can use GPS to describe a point on Earth.

No, you can describe any point in the universe using GPS coordinates. You just lose some resolution the further away from earth you are because it's basically spherical coordinates (like polar coordinates but for 3D). And the system isn't inertial but earth-fixed, of course, so you would have to give the coordinates together with a time.

And if you're describing the location of the moon and the sun, you would probably pick their center of gravity.

I believe this isn’t true otherwise NASA would be doing this

- Earth isn’t a universal reference

- GPS uses WGS84

- GPS is bound to the Earth’s surface and center

- It’s Geodetic

- There's no universal “equator” or “prime meridian” beyond Earth

- Space uses inertial frames or celestial coordinate systems (right ascension and declination, or galactic coordinates)

That's exactly what I said. It isn't very practical for space ops, but you can absolutely give a current GPS position for every object you want.

That’s conceptually misleading.

They are meaningless for things not near Earth because they’re tied to Earth's shape, rotation, and gravity field

I wouldn't call it meaningless if it can be converted back and forth with a (non-linear) transformation.

You can do a lot of things…

There are a handful of Earth centered, geocentric standard reference frames. The most used today is the Geocentric Celestial Reference System (GCRS). It should be obvious that if you want to compute where to point a telescope, a transformation of coordinates will involve a step through such a coordinate system. GPS is it's own system, but there are transformations to and from the GCRS and GPS frames. Which one makes sense depends a lot on your application.

I agree that GCRS and other Earth-centered inertial systems are fundamental for astronomy, tracking, and spaceflight.

And yes, you can transform between GPS coordinates and GCRS (or ECI/ECEF), depending on what you’re trying to do. The key distinction I’m making is about contextual meaning and application. GPS (WGS-84) coordinates are geodetic.

Once you're transforming positions millions of kilometers away (eg. L2, the Moon, or Mars) into lat/lon/alt, you're applying a system optimized for geodesy to a domain where the altitude is arbitrary, and "longitude" rotates with Earth.

The person is simply saying you can. That’s it! As they said, it isn’t useful for things not near Earth. Perhaps not meaningless, but the “meaning” of the coordinate would decrease the further away from earth the object is.

If you’re saying it physically isn’t possible, then you’re basically saying the universe has some topological structure that cannot be projected onto the surface of a sphere. Which now that I’m thinking about it, is probably the case because of black holes? Lol maybe if the object is “in” a black hole then the coordinate actually is completely meaningless, but any other object would have some amount, even if low, of “meaning”

You're right that any 3D point can be mapped to a spherical coordinate system like GPS. I don’t disagree.

GPS coordinates are geodetic, tied to the WGS-84 ellipsoid and Earth’s rotating reference frame. They were designed for terrestrial navigation and near-Earth orbit, not for interplanetary space. Once you're dealing with the Moon, the Sun, or Lagrange points, you're so far outside the system's intended domain that projecting lat/lon/alt onto those locations introduces more confusion than clarity.

You're also right to point out that it's not physically impossible to describe a faraway object in that coordinate system.

There is just no practicable utility.

NASA and others use inertial frames, barycentric coordinates, and RA/Dec for a reason.

Also, love the black hole tangent - that’s where coordinate systems get seriously weird :-)

I could build a house with my pinkie if I excuse the fact I’ll use a team of laborers to do the work and accept that they are so inaccurate that I would be lucky to end up with a shed… if I only cared about technicality.

> NASA did test receiving weak GPS signals on the moon with LRO in 2023.

I doubt very much that the position of the ISS in the article is being sent from the ISS at real time. It's more likely calculated using NORAD / Celestrak orbital elements plus orbital calculations.

I remember having a Windows desktop app to show the satellites locations, I'd have to download those text files to keep the information accurate. For the information beyond the snapshot, the app has to calculate distance and trajectory to estimate "If NORAD said it was here at this point in time, and heading that way with that speed, then right now it should be around here.". A bit like "If a train left Chicago 5 hours ago going 60 mph, where is it now?".

Nowadays it's all online of course: https://in-the-sky.org/satmap_worldmap.php .

> doubt very much that the position of the ISS in the article is being sent from the ISS at real time. It's more likely calculated using NORAD / Celestrak orbital elements plus orbital calculations.

Yes, this is how the referenced site knows the approximate position of the ISS via TLEs. TLEs are updated regularly for space objects

That doesn't matter for the problem at hand though. You can calculate the current GPS coordinates from any TLE, even if they aren't derived from GPS measurements but from Satellite Laser Ranging or some other method.

You can derive Lat and Lon and Altitude on Earth. Thats the one point of the TLEs. But they aren’t GPS derived coordinates.

Yes, but you don't need GPS derived coordinates for the DNS LOC entry.

Correct because the site referenced uses N2YO which is using NASA provided TLEs which some backend that provides an API. GPS and TLEs are not the same.

Yes, I realize not-having initially understood what LOC DNS actually is. As mentioned, this could of course be applied to Hubble.

Any MEO or LEO satellite

Hubble operates in LEO so it’s eligible

Reply on news.ycombinator.com