Rubin Observatory's first LSST Camera release: 500k galaxies in the COSMOS field

Posted by MarcoDewey 1 day ago

Comments

Comment by toufka 1 day ago

The image size of this camera is unbelievably huge in its coverage.

Most telescopes take pictures of objects - this one takes pictures of the entire sky. And will do so in a time-lapse fashion for 10 years.

It's simultaneously more in line with a human's experience than an obscure wavelength of tiny object not visible to the naked eye - it takes a high-resolution picture that's similar to what you'd see if you look at the sky. But it's also profoundly grand beyond a human's experience because of just how much stuff is collected.

Comment by dylan604 1 day ago

I always laugh when I tell people I have a wide angle telescope, but Rubin takes that to an entirely different level.

Comment by irickt 1 day ago

Viewer from Aladin Sky Atlas

https://skyviewer.app/explorer

Comment by binarystargazer 1 day ago

I’m the person in the chain that takes the scientific data and maps it into these color images used in the viewer (among other responsibilities). I spent a long time researching and developing the software that as close as possible maps the data into human perception. It means so much to see people out there appreciating our work, and following what Rubin will achieve. So that you for taking your time to check it out!

Comment by graerg 1 day ago

Hey! Former astronomer here; I used to work a lot with SDSS data products. Do you happen to have a link to any sort of photometry DB/catalog for Rubin? Basically I’m looking to grab a ton of data but I’ve been out of the field for a while and I’m not sure where they publish their data sets and which ones you’d recommend as being relatively accessible.

Comment by binarystargazer 1 day ago

You want to check out the early science release info at https://rubinobservatory.org/for-scientists/resources/early-... . This is all early science from preliminary data. It is mostly to help people ramp up developing their software against the type of things they will be seeing in the future. There some known issues and gotchas outlined in the release. Hopefully this will be enough to whet your appetite. The data seen in this sky viewer is part of the Data Preview 2.

Comment by thegrim33 1 day ago

I hope the data releases are a bit better than what I've seen from recent telescopes.

The common pattern seems to be that the majority of data is locked away behind multi-year embargo periods, or never released to the public at all with no explanation why, with only the smallest little crumbs of raw data actually released to the public, and usually years out of date. Also you have to go through some kafkaesque chain of technologies/servers/bureaucracy to get access to that little bit of data.

It's a little depressing to me when the public funds these telescopes and then get snubbed when it comes to getting access to the data they produce.

Comment by binarystargazer 1 day ago

We are getting pretty far outside my area of expertise / knowledge on this topic here, but I know I and others agree with you here. There is a Rubin Observatory data access policy out there. Don't quote me on the exact specifics but I believe that most colleges and research institutions already have a mechanism to enable data access setup, or can get one. This is mostly an id auth type thing to make sure those accessing have data rights. There is a policy where people not associated with those can request access too, but I am not sure the procedure and would not be qualified to weigh in on anything more specific.

Comment by ricksunny 16 hours ago

Curious what were your thoughts about this story when it broke: https://www.theatlantic.com/science/archive/2024/12/vera-rub...

And whether it impacts your work? For example, if the unnamed agencies' efforts leave a telltale mark in areas they change, that would still be problematic, so presumably they have to place noise corresponding to adjacent pixels. Are you concerned about how data integrity affects your work output? Penny to understand the internal scuttlebut / any internal grousings about this.

Comment by zh3 1 day ago

Jaw-dropping stuff, looking forward to pointing 9-year old daughter (in "Future Astrophysicist" t-shirt) at it!

Possibly a naive question, after surfing the image for a while I didn't find any einstein rings (or arcs) - are they just rare/wrong scale or something else?

and thank you!

Comment by binarystargazer 1 day ago

They are both fairly rare, and or like you said wrong scale (for this data release at least). There is actually one in this field, but it is below the noise floor of this image, it is very small and very faint. There will be others in the future that will be easier to see. By the end of our survey we will have about 40,000 sq deg (a bit over half the night sky) looking like what you see here.

Be sure you and your daughter click the search icon, there are a few other fields of stuff you can check out. The links in the search will take you there without needing to scroll around the big black sphere.

I hope your daughter enjoys it, and we have a few activities at https://rubinobservatory.org/education/first-look-lasting-im... including a coloring book.

Comment by pulsartwin 1 day ago

Thank you for the amazing work! Do you have any publications or articles describing your findings and the process you developed for mapping the scientific data into perceptually meaningful colour images?

Rendering our observatory's images in true HDR is something I've been wanting to implement for some time (especially as HDR support on the web continues to improve), as there is so much more in the data than we typically get to see.

Comment by binarystargazer 16 hours ago

I have a publication that will be done soon™. I have been preparing it for a while, but I keep finding edge cases to fix, and ways to make things even better that I keep changing things fast enough that I need to re-work the paper. The code to produce it will all be open source, and I am even working on making it an independent package you can install separately from the rest of our software stack, specifically so other places can easily install and use it in all sorts of environments.

It's funny you mention HDR, I am _just_ about finished with the HDR extension to the image processing code. What you see in the sky viewer is all standard SDR formats. As you suspect it makes a HUGE difference. Once this is all done we will be working on rolling out many different ways to experience it. The HDR code path will also be part of the package and is not much different to use than the standard stuff.

Reading your comment has gotten me a little more energized to make more progress on this this morning.

Comment by skybrian 1 day ago

This is tricky to navigate because most of the sky survey hasn't happened yet. You need to zoom way out and look for patches that are mostly, but not completely black. And that gives you an idea of how huge the map will be once they've filled more of it in.

Presumably it will get easier after more of the survey is done, but having a list of bookmarks instead of just one might be nice?

Comment by binarystargazer 16 hours ago

If you click on the search icon, there are some quick links to fields across the sky so you don't need to scroll as much. There are some interface tweaks and ui stuff that will be coming down stream that will help with some of this as well.

Comment by relaxer 1 day ago

Just beautiful. Idk why, some part of me feels there's repeating fractals that differ in large chunks of time.

At the same time somehow every chunk is genuinely different. So much space my mind wants there to be a trick of some kind that reduces it

Comment by Aboutplants 1 day ago

Stunning

Comment by doodlebugging 1 day ago

This is great! I love to meander thru data like this trying to understand what I am looking at.

I found a couple of things that are probably processing artifacts from the spectral filtering and selection of colors.

One of them is a bright blue dipping segment that looks out of place. I can see blue filter artifacts on many of the starbursts. This one is very coherent but only for a short interstellar distance. I have labeled it "2" on the linked screenshots. [0]

The other odd duck is the one labeled "1" on the screenshots. It is a green artifact that is brighter and dipping opposite most of the other green artifacts in the starbursts.

I understand there are likely to be multiple passes of coherence filters, median filters, etc so these are probably related to signal/noise ratio enhancements in the raw data.

If you would like to see them for yourself I have included a couple of screenshots that will help locate them.

So much of this appears to be made up of long segments of stars that may twist through space. The individual stars look to maintain an internally consistent spacing in some cases. I think in many of these the individual stars or points of light are widely separated but from our vantage point they look related.

It is in some ways like using a flashlight to look at the dark sky from the inside of a curving windshield that has a lot of bug guts on it.

There's a lot going on out there.

[0]https://postimg.cc/gallery/WbFww2h

Comment by binarystargazer 16 hours ago

Most of the artifacts you see are actually asteroids. We don't take all the colors at once when making an image. The ccds just count photons that fall on them, so we have giant glass filters that allow specific wavelengths through. We then image the fields multiple times with multiple filters and then map the wavelengths of light into color images (see other post about trying to make them as perceptually correct as possible). However, because taking multiple images over time means that things which are changing (i.e. moving asteroids) not not appear at the same places when we combine them. Most of the time these time changing effects get filtered out in the combination step, but sometimes if they are bright enough, or persistent enough the make it through.

As an FYI, you can right click in the sky viewer and it will show the coordinates of the object, and there is a button to copy a shareable link that will take anyone right to where you are looking in the sky.

Comment by doodlebugging 10 hours ago

Thank you for this useful reply. I appreciate the description of image building using the glass filters. That makes sense to me. The asteroids seem like an imaging issue that would be difficult to deal with since they are likely near-field objects that are in motion relative to far-field objects. That would also explain some of the trails that one sees if the objects along that trail are the same object imaged over a multi-step process.

I figured there had to be a way to get real coordinates so that people could do research but I was not bright enough to push the buttons to find that link and had to use the tool I hate - Snip & Sketch. I have a license for SnagIt and Greenshot on another workstation, either of which is better than S&S but that installation is inaccessible due to hardware issues.