I have been working in astronomy for a few years now and I recently have gone back to school for a PhD in orbital dynamics.
Rubin is going to make a big difference in our knowledge of the asteroid belt, it will likely more than triple the size of our known catalog of asteroids. Its actually somewhat difficult to know exactly how much it will increase our knowledge. The bigger the telescope we build, the fainter the asteroids we see. The difficulty is that while we can make a pretty educated guess as to how many smaller ones there are, this is such a jump that the error bars on that guess is quite large.
I am quite excited to see how the catalog of asteroids changes, I expect we will be finding a LOT of smaller rocks near the Earth.
Asteroids have a broad range of albedo, basically the brightness of the surface can vary from the blackest coal (like 3-5% of the light reflecting) to concrete (up to about 50%). All visible range telescopes will be susceptible to a bias in their observations, since a big black rock will be as bright as a smaller paler rock. We know that the asteroid belt favors the dark material.
In a couple of years, the Near Earth Object Surveyor (NEOS) space telescope will launch. NEOS is an IR telescope and will not have the same albedo bias. The trade off is that it will measure the black body radiation, meaning asteroids have to be nearer the sun. Broadly these are very complementary surveys, Rubin will be fantastic for filling out the main belt, and NEO Surveyor will do a great job on our neighborhood.
Source:
I worked at Caltech on NEOS, I wrote the code they use to predict known asteroid orbits:
I love your selfishness. Learned a lot, while sipping my second coffee and eating my breakfast before diving into work. Thanks kind stranger.
I loved astronomy as a kid and the town I grew up in has a "solar system way", basically a long street where a kind fellow (who in his freetime taught astronomy to nerds like myself and had built his own oberservatory in his backyard) had - with the blessing of the city - built a scale model of the solar system on a length of about 1.5 kilometers (a bit less than a mile).
I always found it fascinating when walking "through our solar system" with about 13 times the speed of light (normal walking translated into the distance at that scale) how veeeeeeeeeeery far apart things get in the solar system.
Edit: And yes - Pluto is still in there. It was built before the demotion - and they kept him in when doing the renovation last year (I was not in my home town since then - I so need to see it, when I visit the next time).
No need for all that extra stuff, just be rational enough to realize humans do not exist in a vacuum, and trust those feelings that tell the healthy mind to be kind, and indeed, that other have those feelings too.
If you're just here looking at the HN comments: check out the article. It's really well-written and has a bunch of nice visualizations if you like astronomy things.
The starlink streak issue is real, but isn’t this type of survey uniquely well suited to compensate for it, because it takes repeated exposures in relatively quick succession, meaning the odds of a given pixel being obscured by a streak in successive images get very low? Still not ideal but seems manageable compared to other telescopes running non-automated surveys.
Why is the substrate glass? Lack of reactivity? Ability to remove imperfections? As a layman with almost zero knowledge of telescope construction, I feel like a heavy amorphous solid would not be my first choice for the base layer underneath the reflective/mirror coating.
Its not your normal soda-lime glass, its more of a glass-ceramic material that have very low coefficient of thermal expansion, something like zerodur: https://en.m.wikipedia.org/wiki/Zerodur , which means it can keep its shape and focus even under varying temperature
Q for astronomy people: This is tracking the sky movement as it takes the pictures right? Also, with the atmosphere moving, is there a limit of how large the telescope can be and take photos from earth, before it can't get more quality?
At some point you'll be diffraction limited even at the scale of Earth. The larger your effective aperture, the better you can resolve. Adaptive optics helps get big telescopes closer to diffraction limited performance. That's the best you can do with a given optical system, barring some funky microscope setups. Trying to beat the diffraction limit has occupied a lot of very smart minds.
Practically to go really big you need to use interferometry. There are radio experiments that can do this at Earth scale - the Event Horizon Telescope can image very small objects (black holes) by making simultaneous observations from all over the world. The telescopes point at the same place and use very very good timestamping. At the South Pole we have a hydrogen maser for that. Then all the data gets sent somewhere for correlation and a lot of processing. The analogy I like most is imagining you have a big mirror (Earth) but you've blacked out almost the entire surface except a few points where the telescopes are.
Radio is particularly amenable to this because you can build big dishes more easily than for visible light, and the diffraction limit is lower because it's proportional to wavelength/aperture. So in addition to big dishes, you're observation wavelength is much much longer (mm vs nm).
There's a log-log plot on that wiki page which is quite difficult to read, but the important point is that radio is all the way at the top and the best we have is the VLBA.
The atmosphere is always an issue, we can correct for many of the effects using adaptive optics, but there are always limits. The advantage you get from going bigger on the ground is that you are making a bigger "bucket" to put photons in. More photons = fainter objects are visible, this is the motivation for projects like:
Typically for non-adaptive optics telescope the atmosphere will limit you to the scale of about an arcsecond. Meaning objects which take up less than an arcsecond of the sky will appear as points. Adaptive optics telescopes however have much better resolving power.
- about 20TB per day, around 100PB expected for the whole survey
- 0.5PB ceph cluster for local data
- workloads on 20 nodes kubernetes cluster/argocd
- physical infra managed with puppet/ansible
- 100Gbs(+40Gs backup) fiber connection to US-based datacenter for further processing
> The vast archive, growing by 20 terabytes each night, will after 1 year contain more optical astronomy data than that produced by all previous telescopes combined.
Storage densities these days are kinda amazing, it's not that much of a datacenter. Assuming you chunk it with triple redundancy, that's 220k TB raw. 10k 22 TB disks, you put them in one of those 4U 50 disk storage pods. 200 pods, 10 of those in a rack with some space left for a switch and power, so that's only 20 racks.
The Rubin Observatory will generate approximately 20TB of raw image data per night, with an annual data production of about 15PB for the 10-year survey.
It's been over 10 years since they started. I don't know what the funding details are, but overall this is not really working on a scale that 5 months would change.
It looks like the current administration may kill other projects mid-mission:
"Among the other programs set to lose funding are a craft already on its way to rendezvous with an asteroid that's expected to pass close to Earth in 2029, and multiple efforts to map and explore the acidic clouds of Venus. Researchers worry that abandoning missions would mean investments made by earlier generations might be lost or forgotten."
I have been working in astronomy for a few years now and I recently have gone back to school for a PhD in orbital dynamics.
Rubin is going to make a big difference in our knowledge of the asteroid belt, it will likely more than triple the size of our known catalog of asteroids. Its actually somewhat difficult to know exactly how much it will increase our knowledge. The bigger the telescope we build, the fainter the asteroids we see. The difficulty is that while we can make a pretty educated guess as to how many smaller ones there are, this is such a jump that the error bars on that guess is quite large. I am quite excited to see how the catalog of asteroids changes, I expect we will be finding a LOT of smaller rocks near the Earth.
Asteroids have a broad range of albedo, basically the brightness of the surface can vary from the blackest coal (like 3-5% of the light reflecting) to concrete (up to about 50%). All visible range telescopes will be susceptible to a bias in their observations, since a big black rock will be as bright as a smaller paler rock. We know that the asteroid belt favors the dark material.
In a couple of years, the Near Earth Object Surveyor (NEOS) space telescope will launch. NEOS is an IR telescope and will not have the same albedo bias. The trade off is that it will measure the black body radiation, meaning asteroids have to be nearer the sun. Broadly these are very complementary surveys, Rubin will be fantastic for filling out the main belt, and NEO Surveyor will do a great job on our neighborhood.
Source: I worked at Caltech on NEOS, I wrote the code they use to predict known asteroid orbits:
https://github.com/dahlend/kete
Edit: I failed to mention that Rubin is a big deal for a lot of time-domain astronomy, I'm just being selfish talking about asteroids only.
I love your selfishness. Learned a lot, while sipping my second coffee and eating my breakfast before diving into work. Thanks kind stranger.
I loved astronomy as a kid and the town I grew up in has a "solar system way", basically a long street where a kind fellow (who in his freetime taught astronomy to nerds like myself and had built his own oberservatory in his backyard) had - with the blessing of the city - built a scale model of the solar system on a length of about 1.5 kilometers (a bit less than a mile).
I always found it fascinating when walking "through our solar system" with about 13 times the speed of light (normal walking translated into the distance at that scale) how veeeeeeeeeeery far apart things get in the solar system.
Sadly only in German: https://www.muenchberg.de/erleben/tourismus/tourismus-und-fr...
Edit: And yes - Pluto is still in there. It was built before the demotion - and they kept him in when doing the renovation last year (I was not in my home town since then - I so need to see it, when I visit the next time).
Selfishness tends to work well for everyone when practiced by a rational individual ;)
I’m pedantic: s/rational/romantic|idealist|not-cycical-person/
No need for all that extra stuff, just be rational enough to realize humans do not exist in a vacuum, and trust those feelings that tell the healthy mind to be kind, and indeed, that other have those feelings too.
If you're just here looking at the HN comments: check out the article. It's really well-written and has a bunch of nice visualizations if you like astronomy things.
The starlink streak issue is real, but isn’t this type of survey uniquely well suited to compensate for it, because it takes repeated exposures in relatively quick succession, meaning the odds of a given pixel being obscured by a streak in successive images get very low? Still not ideal but seems manageable compared to other telescopes running non-automated surveys.
Discussion a year ago (75 points, 22 comments) https://news.ycombinator.com/item?id=39927682
The mirror coating timelapse video is pretty awesome https://www.youtube.com/watch?v=Gg9UPS7ndRA
Why is the substrate glass? Lack of reactivity? Ability to remove imperfections? As a layman with almost zero knowledge of telescope construction, I feel like a heavy amorphous solid would not be my first choice for the base layer underneath the reflective/mirror coating.
Its not your normal soda-lime glass, its more of a glass-ceramic material that have very low coefficient of thermal expansion, something like zerodur: https://en.m.wikipedia.org/wiki/Zerodur , which means it can keep its shape and focus even under varying temperature
Interesting demo by Huygens Optics: https://youtu.be/qi8jmEbWsxU?si=rj0I3k-l74Xhg7vC
Q for astronomy people: This is tracking the sky movement as it takes the pictures right? Also, with the atmosphere moving, is there a limit of how large the telescope can be and take photos from earth, before it can't get more quality?
At some point you'll be diffraction limited even at the scale of Earth. The larger your effective aperture, the better you can resolve. Adaptive optics helps get big telescopes closer to diffraction limited performance. That's the best you can do with a given optical system, barring some funky microscope setups. Trying to beat the diffraction limit has occupied a lot of very smart minds.
https://en.m.wikipedia.org/wiki/Diffraction-limited_system
Practically to go really big you need to use interferometry. There are radio experiments that can do this at Earth scale - the Event Horizon Telescope can image very small objects (black holes) by making simultaneous observations from all over the world. The telescopes point at the same place and use very very good timestamping. At the South Pole we have a hydrogen maser for that. Then all the data gets sent somewhere for correlation and a lot of processing. The analogy I like most is imagining you have a big mirror (Earth) but you've blacked out almost the entire surface except a few points where the telescopes are.
Radio is particularly amenable to this because you can build big dishes more easily than for visible light, and the diffraction limit is lower because it's proportional to wavelength/aperture. So in addition to big dishes, you're observation wavelength is much much longer (mm vs nm).
There's a log-log plot on that wiki page which is quite difficult to read, but the important point is that radio is all the way at the top and the best we have is the VLBA.
Visible interferometry is much harder...!
The atmosphere is always an issue, we can correct for many of the effects using adaptive optics, but there are always limits. The advantage you get from going bigger on the ground is that you are making a bigger "bucket" to put photons in. More photons = fainter objects are visible, this is the motivation for projects like:
https://en.wikipedia.org/wiki/Extremely_Large_Telescope
Typically for non-adaptive optics telescope the atmosphere will limit you to the scale of about an arcsecond. Meaning objects which take up less than an arcsecond of the sky will appear as points. Adaptive optics telescopes however have much better resolving power.
Is there a figure somewhere on how many TB of images this will produce per day when running in automated sky survey mode?
I found this pdf presentation with lots of great technical details about data management and a devops infra oriented view of this telescope: https://ci-compass.org/assets/602137/2025jan23_cicompass_rub...
Worth a read for the devops guys around here!
Insanity - love it
> The vast archive, growing by 20 terabytes each night, will after 1 year contain more optical astronomy data than that produced by all previous telescopes combined.
73 PB over the full runtime of the survey. That’s a nice new datacenter filled to the brim with images.
Storage densities these days are kinda amazing, it's not that much of a datacenter. Assuming you chunk it with triple redundancy, that's 220k TB raw. 10k 22 TB disks, you put them in one of those 4U 50 disk storage pods. 200 pods, 10 of those in a rack with some space left for a switch and power, so that's only 20 racks.
Probably about 10 racks if using dense HDDs.
The Rubin Observatory will generate approximately 20TB of raw image data per night, with an annual data production of about 15PB for the 10-year survey.
Has this not been affected by USA science cuts?
No, but its space-based counterpart was cancelled (the 2.4 meter wide-field survey telescope).
https://en.wikipedia.org/wiki/Nancy_Grace_Roman_Space_Telesc...
It's been over 10 years since they started. I don't know what the funding details are, but overall this is not really working on a scale that 5 months would change.
It looks like the current administration may kill other projects mid-mission:
"Among the other programs set to lose funding are a craft already on its way to rendezvous with an asteroid that's expected to pass close to Earth in 2029, and multiple efforts to map and explore the acidic clouds of Venus. Researchers worry that abandoning missions would mean investments made by earlier generations might be lost or forgotten."
https://phys.org/news/2025-06-trump-dozens-nasa-missions-thr...
So I'm not sure any US government-funded science project is safe.