Radio Galaxy Zoo Talk

Press release: Volunteer black hole hunters as good as experts

  • 42jkb by 42jkb scientist, admin

    Our first Radio Galaxy Zoo paper was officially published today on MNRAS and we put out a press release in regards to this.

    CAASTRO: http://www.caastro.org/news/2015-rgz

    ICRAR: http://www.icrar.org/news/news_items/black-hole-hunters

    Posted

  • csunjoto by csunjoto

    Congratz everyone and team scientist.
    May i know if there's another paper from RGZ in progress? Or this is the final end?

    Posted

  • 42jkb by 42jkb scientist, admin in response to csunjoto's comment.

    Oh yes we have a few in the pipeline.

    The hybrids paper has been submitted and in the second round of revision.

    The WAT discovery paper is being written and hopefully ready by the end of the year.

    We are working on analyzing the green DRAGNs and a paper will follow

    There are two catalogue papers coming out. (1) on the RGZ catalogue and (2) RGZ cross matched to galaxy zoo

    Posted

  • csunjoto by csunjoto

    About the RGZ catalogue, will be a list of all the Host galaxies volunteer have classifed and confirmed by team scientist? No matter how the morphologies & size? Am i correct ?

    There's no giant RGZ paper ?

    Thanks Julie

    Posted

  • JeanTate by JeanTate in response to 42jkb's comment.

    But no SDRAGN paper 😦

    Posted

  • 42jkb by 42jkb scientist, admin

    Yes there is a giant paper in the works and the SDRAGN paper. I'm trying to keep up with all the projects and there are so many! Which is wonderful of course!!!

    Posted

  • 42jkb by 42jkb scientist, admin in response to csunjoto's comment.

    The catalogue paper will include the host galaxies to the radio morphology as identified through RGZ. We will attach a "likeihood" or "probability" to the identification as explained in the current RGZ paper as each one will not be confirmed by a team scientist as there are too many. So we will have the number of radio components and the host galaxy in the first data release of RGZ.

    Posted

  • JeanTate by JeanTate in response to 42jkb's comment.

    To what extent will zooites' RGZ Talk Comments (and material in Discussions) be used to create (or add to) the catalog?

    Posted

  • ivywong by ivywong scientist, admin

    @JeanTate: while most of the catalogue will be made using Kyle's KDE algorithm, the RGZTalk comments are very useful for individual projects such as the ones that you've all been working on such as the ones named by Julie as well as those being conducted by Larry's and Heinz's respective teams. From previous conversations, Larry's and Heinz's teams are filtering through most if not all your comments and hashtags in search of their objects of interest.

    Therefore while Kyle's catalogue will be based on the main questions being asked, many more projects and catalogues of individual classes of radio galaxies will be using increasingly the value-added information that RadioTalk is yielding. Does this answer your question?

    cheers,
    Ivy

    Posted

  • JeanTate by JeanTate in response to ivywong's comment.

    Thanks Ivy.

    I'll write more later, when I have time, so just some quick comments for now ...

    • false positives: Kyles' "KDE algorithm" will certainly produce some false positives, radio/IR associations that are not real/physical; this will happen in several ways
    • one will be where the nearest IR object is a bright star; the real association will - almost always - be with a faint IR source invisible in the WISE image
    • another will be with artifacts, of several kinds
    • for example, the FIRST sources we are classifying certainly include those with FIRST side-lobe probabilities > 0.5 (per Radio Galaxy Zoo: host galaxies and radio morphologies derived from visual inspection there is no cut for this)
    • another is the radio sources in the images we classify which have contours include many which are just noise/artifacts (and many which are not!), especially the 'few pixel contour' sources; many will have IR associations that are not real
    • whatever method is used to handle radio sources which appear in more than one ARG field, there is no way to properly associate radio sources with IR ones in cases where at least one physically associated radio source is #overedge (this is true of many giants, but also many non-giants)
    • #doublelobe #ifrs: I do not know how you plan to address these, but I'm sure dealing with false positives for these will be tricky; the result may be a substantial underestimation of how many of these there are

    Posted

  • sisifolibre by sisifolibre in response to 42jkb's comment.

    I love that there are so many papers in preparation and review. I'm happy if RGZ volunteers have helped in these papers, I can't wait to read them!

    I was collaborating on this project about a year ago, but i was frustrated at a little thinking i didn't do it well and set aside to support other. But after the news of this publication in the Facebook page of zoouniverse and read the paper i encouraged me again to take part in this project, so i would say that publish press releases as these it is important to encourage more people to participate in citizen science projects such as this.

    Although I repeat and sounds corny... thank you for your work and for giving us the means to help to the extent that we can. Only I hope to continue learning and keep the height.

    Posted

  • ivywong by ivywong scientist, admin

    Thank you @sisifolibre for your kind and generous support. We really do appreciate your all your help and am glad that we are finally starting to showcase some of the cool results that the science team is able to produce thanks largely to our RGZooites. Keep up the great work! We look forward to more discoveries since we still have ~57% of the project left to finish 😃

    Posted

  • ivywong by ivywong scientist, admin in response to JeanTate's comment.

    I will leave Kyle to comment on his algorithm but these are issues that you highlighted are definitely in the forefront of our minds as we compile the final catalogue. 😉

    Posted

  • csunjoto by csunjoto in response to ivywong's comment.

    Whoa 57% ? O_O i'm feeling like almost all interesting pict already commented & collected. If i may know, how much pict not yet classified by anyone? And how much pict is 57%

    And i agree with @sisifolibre. Announcement or update really give additional excitement to classify more. For example until now i don't really have an exact judgement if a source is hybrid or not. So that's way i'm curious with RGZ Hybrid paper & i recently knew, team RGZ doing observation for this (via RGZ Twitter). I know it's out of topic but if you don't mind, can you share info about hybrid observation. Where are you going to do this? And what kind of data is collected? Are there difficulties to make sure it's hybrid or not?
    😃

    Posted

  • 42jkb by 42jkb scientist, admin in response to csunjoto's comment.

    There are 177281 galaxies in the current RGZ database and we have retired 43% of those so there are many more to look at and I'm sure we will find even more weird and wonderful objects.

    Anna is the lead on the hybrids. We currently have observations going at the VLA (should probably mention this somewhere) and a proposal was just submitted for time on Keck.

    Posted

  • KWillett by KWillett scientist, admin, translator in response to JeanTate's comment.

    Hi Jean,

    • Regarding false positives; these are possible, and pointing out the association with bright stars is a good point. We can mitigate that automatically by searching the WISE catalog for previously identified stars; if those are the cross-match from RGZ clicks, then we either move to the second-most common peak (if present) or mark it as "No IR source".
    • Noise artifacts are mostly ruled out by the fact that only peaks with > 3 sigma flux density are shown. For the tiny noise blips that still make it above those, we're working on an additional cutoff by angular size (which would eliminate the smaller peaks).
    • The automatic pipeline has no good way to identify the overedge galaxies (and I can't think of a way in which that'd be possible). This is where the RGZ volunteer tags and manual collection will be critical. We may add tags to the catalog if a subject has been marked as #overedge by several users (which will allow other astronomers to search for them), but I don't think there's a good automated way to combine the separate lobes.

    Posted

  • JeanTate by JeanTate in response to csunjoto's comment.

    There seem to be a lot more zooites on the site lately, so if it is indeed at least partly due to the PRs (etc), let's see if we can produce conditions from which more PRs (etc) could be issued! 😃

    Like what? I hear you ask; well, like more RGZ-based papers and preprints 😄

    Posted

  • JeanTate by JeanTate in response to KWillett's comment.

    Thanks Kyle.

    Regarding false positives; these are possible, and pointing out the association with bright stars is a good point. We can mitigate that automatically by searching the WISE catalog for previously identified stars; if those are the cross-match from RGZ clicks, then we either move to the second-most common peak (if present) or mark it as "No IR source".

    Bright stars sometimes produce WISE IR image difficulties far from the location of the star itself. E.g. diffraction spikes (diffspikes) are just as bad as the saturated 'halo' itself, and my impression is that diffspikes are longer and wider than from the same star in SDSS.

    One problem with marking "No IR source" is that it's 'sky dependent': the reason there's no apparent IR source in a WISE image is not that such a source may be faint, but that the sky, locally, is vastly brighter than in general. Also, there may be a WISE source ~near the host, but just not visible in the WISE image we zooites use to mark the location of such sources. This may be particularly interesting to investige, using other WISE bands ...

    Noise artifacts are mostly ruled out by the fact that only peaks with > 3 sigma flux density are shown.

    I think such an approach would generally work well in SDSS, or WISE; however, for FIRST, noise artifacts frequently have peak fluxes well above "3 sigma". Obvious ones are those around really bright radio sources (e.g. ARG00039fw, ARG0003a1h), but they also occur well away from the bright source, e.g. ARG00029z3, ARG0003i9u; here's the FIRST image:

    enter image description here

    For the tiny noise blips that still make it above those, we're working on an additional cutoff by angular size (which would eliminate the smaller peaks).

    Good. However, quite a few 'tiny blips' are, in fact, core radio emission from a host; sometimes these are #compact but often #triple (or #doublelobe without the core), e.g. ARG00006g5

    I agree that overedge sources will be almost impossible to search for, automatically; however, these are - as we have seen - sometimes very interesting!.

    New question re the upcoming catalog: how will you identify 'No IR source' radio sources?

    I ask this because I think a delicate balance to be struck between false negatives and false positives here. The false positive case is obvious, and particularly nasty (?) in the case of true #ifrs sources (esp those with wide doublelobes!), but the false negative case is also tough (or so it seems to me). Consider, for example, an asymmetric double lobe ... the host may not be even close to the midpoint between the lobes, and may also be WISE IR faint; a minority of zooites may (correctly) identify this host, but the majority may miss it. Ditto if there are several faint IR sources as possible hosts; the KDE algorithm may produce no clear 'winner'.

    Posted

  • JeanTate by JeanTate in response to 42jkb's comment.

    One other: is anyone working on #ifrs? If so, how imminent is a paper?

    I'm particularly interested in big (between 10" and 1' in extent, say) and bright #triple and #doublelobe #ifrs objects ...

    Posted

  • 42jkb by 42jkb scientist, admin in response to JeanTate's comment.

    As of right now, no one is working on the IFRS. I will send a note off to the science team to see if anyone wants to pick it up. If not, it will have to wait until we get through the GRGs, Hybrids, green DRAGNs, and the catalogue papers and the other ones I mentioned above.

    Posted

  • akapinska by akapinska scientist in response to 42jkb's comment.

    Julie, isn't Ray working (or planning to) on IRFS?

    Posted

  • akapinska by akapinska scientist in response to 42jkb's comment.

    Hi all,

    update on the follow-up observing of hybrids is here: http://radiotalk.galaxyzoo.org/#/boards/BRG0000008/discussions/DRG00007e6

    @csunjoto I will respond to you on the hybrids shortly!

    Posted

  • 42jkb by 42jkb scientist, admin in response to akapinska's comment.

    I know he is interested in IFRS but I'm not sure if he is going to take this one on as the lead.

    Posted

  • csunjoto by csunjoto in response to akapinska's comment.

    As newbie here and with limited knowledge about astronomy , your post about the follow up is very helpful. Before that i never know process from proposal to observation take more time. Good luck for the paper

    Posted

  • ivywong by ivywong scientist, admin in response to JeanTate's comment.

    Thanks @JeanTate. I concur that more papers are great but I would think that we (both the science team & RGZooites) have more impact if each paper we publish is of a very high quality. I think that if we were to bombard PR with lots of papers/PR which are less significant, we would not achieve the same impact as people will begin to tire of the "noise", so to speak.

    Posted

  • JeanTate by JeanTate in response to 42jkb's comment.

    Is anyone working on #restarted?

    Posted

  • akapinska by akapinska scientist in response to JeanTate's comment.

    ha ha no @JeanTate! 😉

    I would absolutely love to, but physically have no time - completely overloaded with all the other things.

    Restarted radio galaxies are really worth looking into. We know they are there, and wonder whether all radio galaxies restart or only some of them (and if the latter then why only some and not all), BUT we don't know THAT many of restarted sources. If I recall correctly there may be 20-30 known ones? So again, a rather rare type of a radio source...!

    Posted

  • JeanTate by JeanTate in response to akapinska's comment.

    Thanks akapinska. 😃

    BUT we don't know THAT many of restarted sources. If I recall correctly there may be 20-30 known ones? So again, a rather rare type of a radio source...!

    For what it's worth, putting "#restarted" into the RGZ Search tool produces 205 'hits':

    • 169 Objects
    • 35 Discussions
    • 1 Collection

    The one Collection is super-zooite antikodon's RSRG, which was started over two years' ago, and also last updated over two years' ago. It contains 38 ARG fields.

    Even allowing for false positives and recovering a high %age of previously known 'restarteds', it seems we zooites may have found several times the already known ~30! 😮

    Posted

  • akapinska by akapinska scientist

    @JeanTate - 169 candidates sounds really great indeed. But these need to be cross checked to retrieve only the truly restarted ones.

    To give you an idea. With hybrids I started with 94 candidates before writing the 1st paper on them, but ended up with only 12 true hybrids. The candidates went through quite a number of experts before settling on the final true selection.

    Posted

  • zutopian by zutopian

    Referring to above four posts concerning restarted jets:

    I would like to mention following paper, which is concerning restarted jets, isn't it?

    Double-double radio galaxies from the FIRST survey

    The radio structures and optical identifications of a sample of 242 sources classified as double-double radio sources by Proctor (2011) from a morphological study of sources in the FIRST (Faint Images of the Radio Sky at Twenty centimeters) survey (2003 April release, 811,117 entries) have been examined. We have been able to confirm only 23 of these as likely to be double-double radio galaxies (DDRGs), whose structures could be attributed to episodic nuclear activity in their host galaxies. A further 63 require either higher-resolution radio observations or optical identifications to determine whether these are DDRGs. The remaining sources are unlikely to be DDRGs. We have examined the luminosities, sizes and symmetry parameters of the DDRGs and the constraints these place on our understanding of these sources.

    Authors: S. Nandi, D. J. Saikia
    (Submitted on 9 Aug 2012)
    http://arxiv.org/abs/1208.1941

    Posted

  • akapinska by akapinska scientist in response to zutopian's comment.

    @zutopian Thanks! I wasn't aware of the paper. By the way Proctor (2011) is really a good resource for candidates only,not for confirmed type of a source.

    Posted

  • zutopian by zutopian in response to akapinska's comment.

    You are welcome! Actually, I had (wongly) guessed, that you mean the paper by Nandi & Saikia, when I had read in a previous post by you following comment.:
    " If I recall correctly there may be 20-30 known ones? "
    So I wrongly guessed, that you probably mean the 23 ones, which had been confirmed Nandi & Saikia .

    The "Radio Zoo Talk"- sample of restarted jets should be cross-matched with the double-double sample, which had been discovered by Proctor and checked by Nandi & Saikia!
    So we would know the number of the new candidates, which were found by RGZ volunteers!
    BTW, the paper by Proctor (2011) is one of the References in the paper "Radio Galaxy Zoo: host galaxies and radio morphologies derived from visual inspection", as you know!

    Morphological Annotations for Groups in the FIRST Database
    Author: D. D. Proctor

    The morphology of selected groups of sources in the FIRST (Faint Images of the Radio Sky at Twenty Centimeters) survey and catalog is examined. Sources in the FIRST catalog (April 2003 release, 811117 entries) were sorted into singles, doubles, triples and groups of higher-count membership based on a proximity criteria. The 7106 groups with four or more components were examined individually for bent types including, but not limited to, wide-angle tail (WAT) and narrow-angle tail (NAT) types. In the process of this examination, ring, double-double (DD), X-shaped, hybrid morphology (HYMOR), giant radio sources (GRS), and the herein described W-shaped and tri-axial morphology systems were also identified. For the convenience of the reader separate tables for distinctive types were generated. A few curiosities were found. For the 16,950 three-component groups and 74,788 two-component groups, catalogs with probability estimates for bent classification, as determined by pattern recognition techniques, are presented.

    (Submitted on 19 Apr 2011)
    http://arxiv.org/abs/1104.3896

    Posted

  • akapinska by akapinska scientist in response to zutopian's comment.

    It was a while ago when I was looking into it, and I think one of the authors was indeed Saikia. Some earlier references I was referring to are also:
    http://adsabs.harvard.edu/abs/2006MNRAS.366.1391S

    http://adsabs.harvard.edu/abs/2009ASPC..407..137J

    But what you found is much more updated 😃

    BTW, the paper by Proctor (2011) is one of the References in the paper "Radio Galaxy Zoo: host galaxies and radio morphologies derived from visual inspection", as you know!

    Yes it is, and..? I'm really sorry I'm not sure I understand what you point here.

    Posted

  • zutopian by zutopian in response to akapinska's comment.

    Yes it is, and..? I'm really sorry I'm not sure I understand what you point here.

    Well, the paper by Proctor (2011) is relevant in the context of restarted jets and besides it has some relevance related to the Radio GZ paper! Actually, the RGZ paper is the subject of this Talk discussion, isn't it? The discussion about restarted jets is "out of topic". Mentioning the RGZ paper was meant as a "reminder"/allusion, that the subject of this discussion is actually the RGZ paper. Interestingly, in both papers (RGZ paper and Nandi & Saikia) the paper by Proctor (2011) is given as a reference, though Proctor's paper has been cited curiously just five times so far according to ADS.

    PS: I hope, that I could make clear, what I mean.

    Posted

  • zutopian by zutopian in response to 42jkb's comment.

    I noticed today, that there is following discussion (in the "Journal Club"), which had been started, when the paper had been accepted.:

    Radio Galaxy Zoo: host galaxies and radio morphologies derived from visual inspection
    http://radiotalk.galaxyzoo.org/#/boards/BRG0000009/discussions/DRG0000c95

    Besides there are following two blog posts.:

    First Radio Galaxy Zoo paper has been submitted! dated 2 March 2015
    http://blog.galaxyzoo.org/2015/03/02/first-radio-galaxy-zoo-paper-has-been-submitted/

    First Radio Galaxy Zoo paper has been accepted! dated 28 July 2015
    http://blog.galaxyzoo.org/2015/07/28/first-radio-galaxy-zoo-paper-has-been-accepted/

    Posted