Author Topic: More Information for Hubble Zoo (beta)  (Read 33559 times)

Michael Parrish

  • Newbie
  • *
  • Posts: 21
    • View Profile
More Information for Hubble Zoo (beta)
« on: June 04, 2010, 03:59:29 pm »
Hi all,
For those of you that may have missed the introductions on the blog (http://blogs.zooniverse.org/blog/2010/05/11/filling-in-the-background/), I've been working on providing some more information for Hubble Zoo.

Very soon (probably Monday), we'll be including a new data view that will link back to this thread asking for your help to improve it.
Until it is on the live site, I'll just start a discussion around a screenshot and some explanation  :)



  • ID:  The Zooniverse ID.  When referring to a galaxy, this is the reference to use as we can guarantee it won't change.  More about this in the future.
  • Survey:  The survey that produced the image.
  • Survey reference:  Any ID given by the survey team.
  • RA/Dec:  Viewable as either decimal or hours/degrees - minutes - seconds.
  • Magnitude:  For each available band as well as the error.
  • Kron/Petrosian radius:  For each available band as well as the error (if available).
  • Redshift:  Also in terms of light travel time and universe age at the given redshift.

I've also linked galaxies to additional resources where I've been able to:
  • AEGIS:  There isn't a perfect match between our dataset and the EGS catalog, though I'm linking into their data viewer where there appears to be a match.
  • GOODS-N:  All of these are linked to two cutout tools (stsci.edu and irsa.ipac.caltech.edu) where fits files are also available.  Unfortunately, their servers are somewhat flakey so one or both may not be online at times.
  • SDSS and Stripe 82:  Linked to SkyServer
  • GOODS-S and GEMS:  I'd like to use the GEMS Browser at http://archive.stsci.edu/prepds/gems/browser.html, though there isn't a way to link a galaxy directly.
  • Additionally, all galaxies are linked to a position-based search on NED.

We are by no means done adding information and this page is very likely to change fairly frequently.
We really want feedback (positive or negative) and ideas from anybody who's even moderately interested.

When it is live, you'll be able to access this information from http://www.galaxyzoo.org/my_galaxies.  The direct url will be http://www.galaxyzoo.org/examine/[ID], where [ID] is the Zooniverse ID.  In order to be backwards-compatible with any old IDs floating around, survey IDs (i.e. SDSS ID) will also work in that url.

ElisabethB

  • Hero Member
  • *****
  • Posts: 13200
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #1 on: June 04, 2010, 05:11:45 pm »
Hi Michael,
this looks great already !  ;D

paulrogers

  • Guest
Re: More Information for Hubble Zoo (beta)
« Reply #2 on: June 04, 2010, 05:28:09 pm »
Wow, that looks great.  Just a couple explanations for us nonprofessionals though?

What's the "Red (606nm)" & "Infrared (814nm)" things mean?  And the significance of the "Kron/Petrosian radius"?

Addition?  If I were to go to NED, et al, 9 times out of 10 it would be because I want to see a spectrum.  Any chances of getting that "All spectra" thing we had on SDSS surfaced to this page?  And as long as I'm wishing, the next most likely thing I would have done is change view resolution, i.e. zoom, or move around, i.e. translate point of focus.  A couple buttons for that would be nice. ;D

I've been waiting for some upgrades to the My Galaxies anxiously, so I can promote Recent's to Favorites, and demote them too.

JeanTate

  • OotD posters
  • Hero Member
  • *****
  • Posts: 2976
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #3 on: June 04, 2010, 05:59:04 pm »
Hi all,
For those of you that may have missed the introductions on the blog (http://blogs.zooniverse.org/blog/2010/05/11/filling-in-the-background/), I've been working on providing some more information for Hubble Zoo.

Very soon (probably Monday), we'll be including a new data view that will link back to this thread asking for your help to improve it.
Until it is on the live site, I'll just start a discussion around a screenshot and some explanation  :)



  • ID:  The Zooniverse ID.  When referring to a galaxy, this is the reference to use as we can guarantee it won't change.  More about this in the future.
  • Survey:  The survey that produced the image.
  • Survey reference:  Any ID given by the survey team.
  • RA/Dec:  Viewable as either decimal or hours/degrees - minutes - seconds.
  • Magnitude:  For each available band as well as the error.
  • Kron/Petrosian radius:  For each available band as well as the error (if available).
  • Redshift:  Also in terms of light travel time and universe age at the given redshift.

I've also linked galaxies to additional resources where I've been able to:
  • AEGIS:  There isn't a perfect match between our dataset and the EGS catalog, though I'm linking into their data viewer where there appears to be a match.
  • GOODS-N:  All of these are linked to two cutout tools (stsci.edu and irsa.ipac.caltech.edu) where fits files are also available.  Unfortunately, their servers are somewhat flakey so one or both may not be online at times.
  • SDSS and Stripe 82:  Linked to SkyServer
  • GOODS-S and GEMS:  I'd like to use the GEMS Browser at http://archive.stsci.edu/prepds/gems/browser.html, though there isn't a way to link a galaxy directly.
  • Additionally, all galaxies are linked to a position-based search on NED.

We are by no means done adding information and this page is very likely to change fairly frequently.
We really want feedback (positive or negative) and ideas from anybody who's even moderately interested.

When it is live, you'll be able to access this information from http://www.galaxyzoo.org/my_galaxies.  The direct url will be http://www.galaxyzoo.org/examine/[ID], where [ID] is the Zooniverse ID.  In order to be backwards-compatible with any old IDs floating around, survey IDs (i.e. SDSS ID) will also work in that url.

Great news!

"I'll just start a discussion around a screenshot and some explanation" - did you omit the screenshot?

"Survey": presumably the universal set is SDSS, Stripe 82, GEMS, GOODS-N, GOODS-S, and AEGIS (or EGS)?

"Redshift": will that be whatever value(s) the relevant survey provides (together with its confidence, and whether it's photometric or spectroscopic)? Or is there some plan to also produce a Zooniverse estimated redshift?

Is there some absolute minimal set of data that will be associated with each (Zooniverse) ID? I'm interested because I suspect artifacts (e.g. diffraction spikes) won't - or shouldn't? - have unique IDs. Also, what sort of "N/A" or "null" text will be used where there are no data from the survey?

Michael Parrish

  • Newbie
  • *
  • Posts: 21
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #4 on: June 04, 2010, 07:30:52 pm »
What's the "Red (606nm)" & "Infrared (814nm)" things mean?  And the significance of the "Kron/Petrosian radius"?
It allows you to view the measurements (magnitude and radius) for each wavelength.  The radius basically describes the distance from the center of the image that contains the brightest part.  For instance, if the brightness is highly concentrated into a point, it will have a small radius.

I've been waiting for some upgrades to the My Galaxies anxiously, so I can promote Recent's to Favorites, and demote them too.
Check again, it's there  :)

"I'll just start a discussion around a screenshot and some explanation" - did you omit the screenshot?

"Survey": presumably the universal set is SDSS, Stripe 82, GEMS, GOODS-N, GOODS-S, and AEGIS (or EGS)?

"Redshift": will that be whatever value(s) the relevant survey provides (together with its confidence, and whether it's photometric or spectroscopic)? Or is there some plan to also produce a Zooniverse estimated redshift?

Is there some absolute minimal set of data that will be associated with each (Zooniverse) ID? I'm interested because I suspect artifacts (e.g. diffraction spikes) won't - or shouldn't? - have unique IDs. Also, what sort of "N/A" or "null" text will be used where there are no data from the survey?
The screenshot is being hosted on an old server of mine (http://kachunk.sl.siue.edu/examine.png), sorry if that's making it tough to view.

COSMOS should be on that list at some point.

As far as redshift goes, I used the values provided from our catalog which also has its own selection criteria between spec-z and photo-z.  For the most part, I also have photo-z, spec-z, photo-z error, and some quality flags for spec-z available.  So, it is definitely something I could add.

The rule we're using is "if it's in the zoo, it gets an ID."  If for some reason we ever take it out of the zoo (don't know if we'd ever do that though), we wouldn't recycle an ID.

Where values are unavailable, we'll just list it as "Unavailable"

** edit ** I've also been prototyping a spectrum viewer.  Digging through loads of FITS files takes some time  :)
« Last Edit: June 04, 2010, 07:34:09 pm by Michael Parrish »

Half65

  • Hero Member
  • *****
  • Posts: 10288
  • "Eagle Nebula"
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #5 on: June 04, 2010, 08:47:25 pm »
That's really great. Whit this data I can extract data from overlapping thread
Thank you.

swengineer

  • Sr. Member
  • ****
  • Posts: 312
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #6 on: June 05, 2010, 04:11:03 am »
What's the "Red (606nm)" & "Infrared (814nm)" things mean?  And the significance of the "Kron/Petrosian radius"?
It allows you to view the measurements (magnitude and radius) for each wavelength.  The radius basically describes the distance from the center of the image that contains the brightest part.  For instance, if the brightness is highly concentrated into a point, it will have a small radius.
Is it possible to also show which wavelengths (channels?) map to which colors in the composite color images?  Regardless, I really look forward to trying out the new data view interface.  Thank you!

JeanTate

  • OotD posters
  • Hero Member
  • *****
  • Posts: 2976
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #7 on: June 05, 2010, 03:34:58 pm »
What's the "Red (606nm)" & "Infrared (814nm)" things mean?  And the significance of the "Kron/Petrosian radius"?
It allows you to view the measurements (magnitude and radius) for each wavelength.  The radius basically describes the distance from the center of the image that contains the brightest part.  For instance, if the brightness is highly concentrated into a point, it will have a small radius.
Is it possible to also show which wavelengths (channels?) map to which colors in the composite color images?  Regardless, I really look forward to trying out the new data view interface.  Thank you!
I think the new capabilities (tools) will be accompanied by a number of informative articles, giving some background, some details of the various processes involved in data analysis, and so on; a bit like what you can find on the SDSS site, but written perhaps more directly for zooites.

In general, the color transformation (mapping) in astronomical images maintains relative order.

For example, the SDSS filters are u, g, r, i, and z, in order of increasing wavelength. Images made from data taken through three filters - g, r, and i say - will nearly always be displayed with the shortest wavelength mapped to B (g in this case), the middle to G (r), and the longest to R (i; I don't know if mappings are made to color schemes other than RGB, but I doubt it).

In the case of images created from just two filters, the simplest thing to do is to map the shorter wavelength one (say "Red (606nm)") to B, and the longer (say "Infrared (814nm)") to R. However, I think it's common to create a G image, by some kind of combo of the B and R ones, and generate a (pseudo) three color image.

What's been done for the three (more?) Hubble-sourced images we'll have to wait for one of the zookeepers (or astronomers) to tell us, but my guess is that the '90' series images were created using a different mapping than the '1' and '50' series ones (blue/cyan is common, but the '90' series seems to use orange, whereas the other two use yellow).

paulrogers

  • Guest
Re: More Information for Hubble Zoo (beta)
« Reply #8 on: June 06, 2010, 06:38:21 am »
What's the "Red (606nm)" & "Infrared (814nm)" things mean?  And the significance of the "Kron/Petrosian radius"?
It allows you to view the measurements (magnitude and radius) for each wavelength.  The radius basically describes the distance from the center of the image that contains the brightest part.  For instance, if the brightness is highly concentrated into a point, it will have a small radius.

But we'll need to know what to make of this information when we're deciding whether to bring a classification to the forum. :-\

Quote
I've been waiting for some upgrades to the My Galaxies anxiously, so I can promote Recent's to Favorites, and demote them too.
Check again, it's there  :)
Excellent! :D  I so didn't want to lose my little "mushroom",

or my pretty blue spiral,

and this one was the second one I got, and it's screaming AGN.

All saved as favorites now! :D

JeanTate

  • OotD posters
  • Hero Member
  • *****
  • Posts: 2976
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #9 on: June 06, 2010, 02:53:13 pm »
What's the "Red (606nm)" & "Infrared (814nm)" things mean?  And the significance of the "Kron/Petrosian radius"?
It allows you to view the measurements (magnitude and radius) for each wavelength.  The radius basically describes the distance from the center of the image that contains the brightest part.  For instance, if the brightness is highly concentrated into a point, it will have a small radius.

But we'll need to know what to make of this information when we're deciding whether to bring a classification to the forum. :-\
I'm afraid I don't follow you here paulrogers.

Do you mean you think you need that info (and understand it) in order to make a classification?

Or, having made a classification, decide in which Forum thread to post the interesting object (that you've just/already classified)?

Or something else??  ???

paulrogers

  • Guest
Re: More Information for Hubble Zoo (beta)
« Reply #10 on: June 06, 2010, 05:22:26 pm »
But we'll need to know what to make of this information when we're deciding whether to bring a classification to the forum. :-\
I'm afraid I don't follow you here paulrogers. Do you mean you think you need that info (and understand it) in order to make a classification? Or, having made a classification, decide in which Forum thread to post the interesting object (that you've just/already classified)?  Or something else??  ???

Yes.  But primarily the second.  For the first, we're primarily doing shape classification.  Straightforward enough one might suppose, but I'm wondering how looking at a particular slice of the spectrum affects the shape we see.  Might not looking at a narrow wavelength, say OII, make a spiral look more clumpy and ring-shaped?  We are by nature multispectral creatures, in the 4000-7000A range.  I once walked into a friend's darkroom and my red parka turned unexpectedly "white".  I'm supposing it would really matter when trying to do our initial amateur interpretations of what's "interesting" for posting in the forum, to know what these wavelengths enhance or don't see.  There's some debate in the "severely misaligned" thread about what's going on because the orange/red and blue images don't seem to line-up by simple translation.  They seem to be showing us different things.  Perhaps an explanation why these particular wavelengths were chosen might be "illuminating". ::)

c.f. http://www.galaxyzooforum.org/index.php?topic=277799.msg468529#msg468529
« Last Edit: June 06, 2010, 05:32:49 pm by paulrogers »

lpspieler

  • OotD posters
  • Sr. Member
  • ****
  • Posts: 377
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #11 on: June 06, 2010, 06:38:04 pm »
Hi,

the info page looks very helpful. I especially like the inclusion of the NED search because I often had the case that the object I was interested in didn't seem to have an SDSS id. NED made sure that there definitely was no catalogued object at that particular position. As for the Hubble surveys: I take it that the links to the survey pages will allow for a search for the survey id. Are there also possibilities to search for ra/dec?

Another little thing: The "infrared (8..nm)" info at the top is coloured. I guess that means it's linked somewhere. Where does the link go? Or does the colour mean something else?

cheers

Michael Parrish

  • Newbie
  • *
  • Posts: 21
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #12 on: June 08, 2010, 04:15:48 pm »
The update's live now.

I'm also still looking into some options to provide some multi-wavelength views on the galaxies.

elizabeth

  • Hero Member
  • *****
  • Posts: 33783
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #13 on: June 08, 2010, 05:04:15 pm »
 ;D Awesome!! Thanks

chicgeek

  • Newbie
  • *
  • Posts: 15
  • A long time ago, in a galaxy far, far away. . .
    • View Profile
Re: More Information for Hubble Zoo (beta)
« Reply #14 on: June 08, 2010, 07:23:03 pm »
In general, the color transformation (mapping) in astronomical images maintains relative order.

For example, the SDSS filters are u, g, r, i, and z, in order of increasing wavelength. Images made from data taken through three filters - g, r, and i say - will nearly always be displayed with the shortest wavelength mapped to B (g in this case), the middle to G (r), and the longest to R (i; I don't know if mappings are made to color schemes other than RGB, but I doubt it).

In the case of images created from just two filters, the simplest thing to do is to map the shorter wavelength one (say "Red (606nm)") to B, and the longer (say "Infrared (814nm)") to R. However, I think it's common to create a G image, by some kind of combo of the B and R ones, and generate a (pseudo) three color image.

What's been done for the three (more?) Hubble-sourced images we'll have to wait for one of the zookeepers (or astronomers) to tell us, but my guess is that the '90' series images were created using a different mapping than the '1' and '50' series ones (blue/cyan is common, but the '90' series seems to use orange, whereas the other two use yellow).

What?? I can't seem to get my head around how this redshift thing works! u, g, r, i, z? If those are shortest to longest wavelength, is u blue and z red? What do the numbers actually tell me? I clearly have no idea what I'm talking about. Can anyone point me to an idiot's guide to redshift?  ???
The universe never did make sense; I suspect it was built on government contract.
-Robert A. Heinlein