On the one hand the constellations are doing considerable damage to surface based astronomy. On the other hand they are funding an enormously lower cost of access to orbit and beyond for astronomy up in the sky.
If the satellites don't disrupt it, light pollution will anyway, or best case, it will be disrupted by atmospheric pollution.
There's no lack of clear skies in space.
There are so many things that you can only conceivably do from the ground. You can build vastly larger telescopes on the ground, you can install far heavier instruments (like cameras and spectrographs) on them, you can upgrade and repair components much more easily. You can build massive arrays of radio telescopes. You get the point.
There are some things you can only do from space, or that are better in space, but saying that everything will be done in space is like saying we only need laptops and don't need datacenters. You need both.
I'm just not seeing the argument for there being any telescope function that wouldn't be better done in orbit, at Starship payload prices.
There’s a much longer list of large radio based telescopes which have minimal advantages in space to the point where we haven’t launched any, and they are also negatively impacted by constellations. Also, several types of observations don’t really benefit from being in space. If you want to track killer asteroids space doesn’t provide much advantage even if you could get there for 100$/kg.
IMO the issue here isn’t satellites, it’s that they can harm multi billion dollar investments at zero cost to themselves. If you’re very clearly causing 10’s of millions in damages you really should be compensating the people affected.
What if they could be put behind the Moon, completely shielding them from Earth RF emissions? It's been a cost-prohibitive idea now, but might not be for much longer.
edit: italicized above. The Webb ended up costing $10 billion, so the savings could be even larger
You could build Webb for vastly, vastly less money if it launched on Starship, not just per unit of mass; volume is at much less of a premium in the new regime.
You can’t quite fit ELT pre-constructed into a single Starship, it’s true. But it’s built out of segmented mirrors. It doesn’t have to arrive on station in one piece.
B: yeah we're hurting you but if we didn't, A & C would
C: yeah we're hurting you but if we didn't, A & B would
I guess it’s nobody’s fault then
But well, any one is free to go fight the tide.
That mirror size pails in comparison to what can be accomplished via ground based observatories for much, much cheaper.
And the available modes of operation are completely different between space and ground. For example, space observatories are using decades old instrumentation techniques, hardware, and software. And that's just for starters.
We cannot do astronomy via space exclusively.
For me personally, I never wanted to see something like Starlink with my bare eyes, and I have several times. It's beyond annoying to me that a company has been able to unilaterally pollute the Earth's sky.
Because it had to get launched folded up, because the launch vehicles were so space- and mass-constrained. The unfolding mechanism was enormously complicated and added to much of the cost.
Now we can send up bigger, heavier objects for much less money. The 6.5m mirror fits comfortably inside Starship's cargo bay, unfolded. JWST is a one-off clockwork masterpiece; future space telescopes will come off an assembly line. There will be thousands in orbit, pointed at every part of the heavens simultaneously.
Yes, JSWT was launched into a L2 transfer orbit and not LEO, but from the above margins there is probably enough spare capacity for this.
The 39-meter primary mirror has 798 1.4-meter segments, each individually adjustable. The scaffolding required to hold all the segments is "significant" and must be assembled in orbit. We would stick this out next to the James Webb at L2, which means we would need the ability to travel to L2 to construct this mythical orbital extremely large telescope.
With our current technology, we can build ELT on Earth for a fraction of the cost of putting it in orbit.
It is much simpler to build the telescope equivalent of IKEA flatpack furniture and assemble it on-site than it is to build it and then move it.
Why? Just take existing ion engine designs but use a lot of them.
But it's beside the point. Such a singular mega-telescope would no longer be the only way to do astronomy in space. A world of radically lower launch costs gives more possibilities. We could have a fleet of thousands of independent small/medium sized space telescopes. That way, we wouldn't have to carefully ration imaging time between competing astronomy projects anymore. High quality data would become cheap and abundant, procured on-demand.
Electronics, sensors, and consumables fail or run out, which means you need to be able to get there to fix things. Then, as research projects change, you need to change instruments.
Yes, one could have thousands of 1-5 m telescopes. A terrestrial scope is about an order of magnitude cheaper than an orbiting telescope. So, the big question is who will pay to replace the thousands of scopes in that size range that are already operating on Earth.
Rationing will still needed (Research projects always exceed the number of available scopes). High-quality data, will not be as cheap or abundant as you think it might be. With the number of scopes we are talking about and the sensors that astronomers use, we can expect hundreds or thousands of petabytes a second of data. How will we get that data down from orbit?
A practical example this volume of data comes from the event horizon telescope. I could describe the firehose of firehoses but this paper from supermicro does a much better job of it https://www.supermicro.com/white_paper/white_paper_Black_Hol...
This is just one project. Could you imagine a thousand of these astronomical fire hoses running at the same time in orbit?
The next issue is where these telescopes should be placed. The more we learn about running telescopes in orbit, the more we realize that they need to be placed very far from Earth to increase the observable sky and the length of time one can study a chunk of it. As I said above, consumables need to be replaced, sensors need to be changed for the observing program, boils down to how you are going to get a repair crew out to wherever the telescopes are.
So launch another. Launch capacity is getting cheap... let's use it.
Consider that 1% of SpaceX's annual launch capacity is enough to put one Hubble in orbit every year. Instead of sending astronauts to fix the fucked-up mirror, you just launch one with a not-broken mirror.
We as a planet don't even build that many ground-based telescopes with a 2.5m+ mirror each year. Think about how astronomy would change if you could just take every telescope we build today and put it in space.
This scope is not built for space travel. This model for space telescopes from NASA put a price tag on all the major components for building and operating a single space telescope is counted in billions. https://ntrs.nasa.gov/api/citations/20110015780/downloads/20...
A couple of things not covered in other comments are the costs of mission control and end-of-life deorbiting. It is far cheaper to rent a car, drive to an observatory, mount your evolving experiment on scope, and debug it on-site than to put the same experiment in orbit. Terrestrial telescope mission control is ad hoc and usually in a heated/air-conditioned shack on the mountaintop.
That's going to change soon. The reason the ELT needs individually adjustable panels is to provide corrections for refraction in the atmosphere as it swirls around. A space-ELT doesn't need that (well, maybe some cheap slow actuators to set the focus on a large non-rigid mirror).
As the cost of launches drop, we will hit a point where it is cheaper to put the telescope in space where you don't need those expensive atmosphere correctors.
It's only barely technically feasible to construct these 30-m telescopes on the ground. The idea that they can be constructed in space is just fanciful.
Like I said: citations.
> Now we can send up bigger, heavier objects for much less money.
This hasn't even been demonstrated.
It is because you did not start your own re-usable space vehicle company. Bezos did but thus far he seems to treat it as a billionaire's toy, throwing a few fellow class members just over the Kármán line. There's a few others making attempts to actually get to space - Rocket Labs etc. - but mostly it comes down to Musk's creation which opened the door to radically lower costs, i.e. without them the usual suspects would have had no reason to keep on pushing 70's technology at whatever price they manage to extract from the governments and the market.
https://ourworldindata.org/grapher/cost-space-launches-low-e...
That's pretty good for theory.
Are you sure about that?
If you can launch a hundred tons to orbit for $5M, you can just make a huge dumb cheap telescope and throw a dozen of them up there. Quantity covers a multitude of sins.
Well, technically, I suppose you could argue that the direct launch costs are small. However, most of the cost of the telescope itself are driven by the high launch costs.
That's the classic problem for satellites: Launch is expensive, so you can't afford to fail. So you have to use mil-spec everything. Every screw has to come with a piece of paper attesting to the provenance of the chromium used to make the stainless steel. Every system has to be built and tested and have triple redundancy. You have to use teflon-insulated wires, because god forbid anything outgases while on orbit. You have to over-build everything, and test all the vibration modes on those giant vibration tables NASA owns. So now "the satellite is expensive". Oh, and since the satellite is expensive, the launcher can't afford to fail and costs just spiral out of control.
But look what happens when you have cheap launches: You don't care if the satellite has a 1% chance of failing. If it does, learn from your mistake, make that one thing better and launch again. Now that you only need two nines instead of five, all of your costs go down. You can use regular-grade chips, and hardware store screws. Instantly the price of everything drops by several orders of magnitude.