ESASky | Hacker News Reader

ESASky

sky.esa.int

355 pointsxioxox55 comments

Since no one has explained what esasky is: esasky allows professional astronomers (and anyone else because astronomy is all about open science) to view science observations from observatories across the world in one place.

If you look in the top-left corner you'll see a bunch of controls.

The first one (with three layers) chooses what is displayed. You can choose from loads of surveys (some which cover the whole sky, others that only cover a part) at loads of different wavelengths (and because the sky does change, search for LSST, there's an implicit time aspect as well). These are the original images (ignoring the underlying reduction process each survey does), these are for science not outreach. This is the button to play with (see how the same spot looks in visible vs. radio vs. x-ray).

Next button allows you to pick specific observations (as the all-sky part has implications about how you tessellate images), not that useful unless you understand what you're doing in more details than I have space for.

After that is catalogues of objects. This information will be compiled by survey teams, and is derived from various sources (including other catalogues). The magical astronomy keyword here is "TAP" for Table Access Protocol.

Spectra and timeseries is the next button, you're (generally) looking at spectra/timeseries from individual objects here, but things get more complex here.

The remaining buttons are really of no interest outside the profession sphere (though the multi-messager (i.e. not light, think gravitational waves/neutrinos etc.) button might be interesting when LIGO is running).

(I'm an astronomy RSE, but I don't work on esasky, nor for ESA).

Beautiful, but could do with using quaternions for rotation to avoid gimbal lock. At least I assume that’s the problem with getting stuck on the poles

For interactive navigation on a sphere (remember, this is just a sphere looked at from the inside) you actually want absolute euler angle based rotations, otherwise compound rotations around any closed loop on the sphere's surface will change the orientation of the camera and make further navigation extremely confusing. Source: many years spent designing 3d applications. Also, Google Earth and every other major app do it the same way.

Im fact, if you zoom out all the way, you'll see that we're actually looking at a sphere from the outside, as if the stars are situated on a celestial globe. It seems to be a relatively common convention for interactive night-sky maps.

Inverting the Z axis purely as a UI quirk/implementation detail is one thing, but more interesting would be to embrace it as a means of representing the time scale of the universe - as you zoom into the earth, you zoom into the deeply redshifted ancient data, which conveniently is from a smaller universe so projecting it onto a smaller shell of the sphere makes sense. Until you reach the core, which is a picture of the cosmic background radiation.

Not nearly as easy to navigate and discover as apps like Stellarium.

Now, if they had the database with hundreds of terabytes of objects that NASA has for their OpenUniverse simulation they’re running for the upcoming Roman space telescope (see https://www.jpl.nasa.gov/news/nasas-roman-mission-gets-cosmi...), then maybe I could understand why it is so confusing.

As it is, I don’t really understand why it has to be so confusing.

I'm confused why you think this is confusing? Stellarium is something you use to point a telescope with (and more focused on amateurs), whereas esasky is designed to browse through objects/images/spectra (with science data for scientists)? I don't know how much the overall data would be (given esasky is using a bunch of astronomy standards would be), but we're at least in 100s of PBs here.

I don’t see evidence of 100s of PB of data here. I do see multiple green stars, which I know is extremely rare in the real world. That’s just not a common color spectrum for stars.

And using a viewpoint of a large sphere from the outside just seems wrong. It should be viewed from the inside.

https://imgur.com/a/r2n3RMz

Are you really, really sure the hundreds of millions of images &c. it links you to aren't in the hundreds of petabytes in size?

I saw no numbers on any of the icons when I visited in my web browser.

It looks like they have a deficient web page design that doesn’t work well on mobile. Thus leading to a much more confusing interface.

It's not for you, obviously. It's for academics and nerds that are interested in these kinds of data sets and used to have much more trouble finding and consuming them.

I think if they had a better web page interface, then a lot more people would be able to make use of the site.

And I think that would probably also serve the professional astronomers or astronomical researchers better.

To put it nicely: you're not the target audience. This is the astronomy equivalent of the $50000/yr bespoke mechanical engineering simulator. This is not an outreach tool (or a citizen science tool). You don't take this to a dark field with a telescope at night (it's not a skymap), this will be used on a big screen to plan observations, or at an observatory (where you'll have at least 3 screens running). These kind of applications are developed with deep and continuous input from their users, and there's nothing superfluous there.

What do you even mean by "better"?

Who is it that can manage the niche software needed to digest the data sets but have such trouble figuring out how to navigate that web page? Of what use would they be to the academics and weird nerds?

It took me less than two minutes to figure out navigation and how the menu works. If you really want to make an impression in this area it would probably be better to spend time on improving VO-software and organising study groups where you live.

Agreed, especially with regards to Stellarium, although this seems more of a researcher's tool than a hobbyist's tool.

There's an excellent web version of Stellarium, if anyone's interested: https://stellarium-web.org/

Wow, so many Starlink satellites overhead!

Does anyone know what's the deal with the fairly consistent discoloration of the individual images? The edges are often orange and the middle blue. You'd think they'd colour correct this out when doing the stitching...

Also, are these [0] artefacts a result of adaptive optics since they shine out those lasers to keep track of distortions? And these [1] which seem to be the same but larger and less focused. I remember seeing similar ones on Google Sky years back but never really figured out what causes it.

[0] https://sky.esa.int/esasky/?target=94.25875681534997%2020.97...

[1] https://sky.esa.int/esasky/?target=218.7659069213465%20-59.6...

Maybe you could link to the images themselves rather than positions on the ESASky map? Those positions are associated with hundreds of thousands of images.

Does it not open properly on your end? The artefacts seem pretty glaringly obvious in those locations and I've tested the links in another browser to make sure it's not caching the zoom or whatever. What you see when you load is the image I'd take.

Anyhow, whoever can actually answer will surely know which ones I mean.

Aren't you referring to just the map? You find the images under a menu in the upper left corner.

Likely a "ghost" caused by a reflection from a bright star somewhere in the instrument: https://www.eso.org/%7Eohainaut/images/imageProc.html

Ah that's it! So it does actually show the primary mirror of the telescope, with the hole in the middle probably being the struts holding up the secondary.

When I scroll out, I would expect to reach a point close to Earth or our solar system, but somehow, I look at a sphere from the outside.

It's just a projection, you're looking at the sphere of the sky as seen from a (stabilized) Earth position. When you zoom out, you increase your field of view beyond that of the human eye to encompass the whole sphere. Because there's no way to compress a sphere into a plane, you have to choose a projection; this is a common projection (stereographic) which lets you zoom out as far as 180 degrees, by presenting the field of view as if it were the surface of a sphere as seen from an infinite distance.

A fun fact I learned while visiting the Globe Museum in Vienna, is that they used to make globes in pairs; one would be a globe of the Earth, and the other would be a Celestial Globe which is a projection of the sky onto a sphere. What you're seeing is a celestial globe

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

You are The One.

I found some kind of artifact @ 70.3169798 +19.0238259 FOV: 1.3°x2.4°. What is that? Some kind of antenna?

I think it is the shadow or a reflection of the camera itself within the telescope. It's visible in multiple places.

https://sites.astro.caltech.edu/palomar/about/telescopes/ima...

exactly. I see it as well in multiple other spots as well. eg: 1. 05 49 04.008 +01 09 47.27 FoV: 1.3° X 2.2° 2. 05 44 35.074 +01 52 52.36 FoV: 3.4° X 5.9° 3. 06 37 17.795 -47 18 20.97 FoV: 1.4° X 2.4°

many more that i didnt list.

I wish there was something for 3D level of exploration, preferably open source, something like Stellarium but in 3D, something similar to the navigation interface[1] in the Expanse TV series.

[1] https://www.joycenho.com/wp-content/uploads/2015/04/EXP-Stil...

I'm not quite sure what your exact specifications are, but I think that Celestia[0] fits?

[0] https://github.com/CelestiaProject/Celestia

https://spaceengine.org/ , though it’s partially fictional if I recall correctly. Edit: oh and not OSS. Misremembered.

How fixed is all this? It would be nice if we could scroll through time.

Pretty much everything beyond our solar system is essentially fixed on a human timescale. Over 2000 years, a typical star will move about half a degree. That's the width of the moon in the sky. There are of course notable exceptions like Barnard's Star, whose movement is pretty obvious on photographs taken over several decades.

If you want to explore how space changes over time, I recommend you look into something like Celestia[0]. It allows you to simulate star movement over thousands of years, and show you how the night sky looked to the Ancient Egyptians.

[0]: https://celestiaproject.space/

Some what related, two awesome links:

Nearest star to us, over time: https://commons.wikimedia.org/wiki/File:NearSunStarsSimple.j...

Star closest to the north pole, over time: https://en.wikipedia.org/wiki/Pole_star#cite_ref-Meeus_6-2

All these stars, and every one of them is unique. It boggles the mind.

Only the sharp dots are stars - all those fuzzy little orbs in the background are galaxies

> every one of them is unique

How sure of this are we? Feels like we've can only confirm that's true for a small selection of the vast total.

Think of them like snowflakes

So not all unique.

To quote [0]:

> Although snowflakes are all the same on an atomic level (they are all made of the same hydrogen and oxygen atoms), it is almost impossible for two snowflakes to form complicated designs in exactly the same way. While snowflakes can be sorted into about forty categories, scientists estimate that there are up to 10^158 snowflake possibilities. (That’s 10^70 times more designs than there are atoms in the universe!)

[0] https://ssec.si.edu/stemvisions-blog/are-all-snowflakes-real...

I don't think this is what people are talking about when they say "no two snowflakes are the same". Treated this way, no two of anything could ever be the same. Why even bother having the word "same"?

Sure, but you could say the same for white t shirts.

We can’t really confirm it for anything. It’s all up to the axioms you choose. Equally reasonable ones would have them all distinct, uniform, and everything in between.

"ESASky is an application that allows you to visualise and download public astronomical data."

Reply on news.ycombinator.com