Galaxy Zoo Starburst Talk

Spectroscopically quasi-confirmed companions of quench sample objects

  • mlpeck by mlpeck

    JeanTate suggested something like this as a potential side-project, and I've had time for some followup. Here, as quickly as possible is what I've done so far.

    I used the skyserver cross-id tool to search for all spectra within 2 arc-minutes of each quench sample object -- I used the entire untrimmed sample of 3000 for this. That produced a list of 7157 objects which includes each QS object, repeat observations of same, all neighbors on the sky regardless of redshift difference and repeat observations of those. Next I calculated the distance on the sky from each QS object to every object in the list of 7157 and excluded "neighbors" with a distance within about a fiber radius and more than 2 arc-minutes away. This should have eliminated repeat observations of QS objects, but not of neighbors and still no redshift cuts are applied. This produced a table with 3654 rows.

    Finally, I went to CasJobs and did a join of this list of candidates with the SpecObj view. SpecObj is supposed to contain only "primary" objects so in principal this should eliminate repeat observations. I also at this point restricted redshift differences to |Δz| < 0.01 and only selected objects classified as 'GALAXY' or 'QSO'. This produced a final table of neighbor candidates with 560 rows.

    Out of those there are 548 unique specObjId's of neighbors (no I don't know yet why there are repeats) and 447 unique specObjId's from the original quench sample. Of the 548 seven are classified as 'QSO', which basically just means they are broad-line AGN.

    I intentionally made the selection criteria for neighbors fairly loose anticipating that the final number of candidates would be small enough to make final cuts by hand. It turns out that the relative velocity differences ranged from ~ ± 2700 km/sec. with about 80% between ~ ± 550 km/sec. Projected transverse comoving distances range from < 1 kpc to ~660 kpc with half < 125 kpc.

    I've just briefly looked at the sample using the SDSS image list tool. What I see include:

    1. different spectra of the same object (these will be different plates with slightly different positions or in a few cases nearby galaxies).

    2. mergers in various stages. Some of these are quite spectacular. One of the 'QSO's for example is evidently having its AGN fed by a QS object.

    3. Common group/cluster members that may or may not show signs of interaction.

    4. Accidental associations (quite possibly half or more of the sample). And of course there's always the possibility of

    5. Errors in the spectro pipeline redshift measurements.

    I've published the table on CasJobs with table name "qneighbors". My user name there is mlpeck54. There's also a CSV version shared on my dropbox account at https://www.dropbox.com/s/621xmisibtv7h3o/quench_neighbors.csv. The table contains (DR8+) specobjid's, positions and redshifts of QS objects and neighbors and a bunch of data taken mostly from the MPA pipeline on each neighbor.

    Comments and questions welcome.

    Posted

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

    Wow! Impressive! 😄

    Some very quick comments:

    1. different spectra of the same object (these will be different plates with slightly different positions or in a few cases nearby galaxies).

    Where these are of QS objects ("QS spectra"), I wonder if they'd also be selected (as "QS objects"), if the same Chen PCA were repeated?

    1. Accidental associations (quite possibly half or more of the sample).

    Worth doing a quick check of the overlapping pairs catalog (Keel et al. (2013))? Maybe there are some as-yet-unrecognized overlaps?

    1. Errors in the spectro pipeline redshift measurements

    Yeah. It's usually very reliable, but sometimes it fails, rather spectacularly.

    search for all spectra within 2 arc-minutes of each quench sample object

    From work I did earlier, I know this will find almost every case of multiple sources/same galaxy ... but not all. There are some very nearby spirals - both Eos and not - which have two (or more) spectroscopic objects ("objects with spectra") that are further apart than 2'. Are any among the 3002 QS objects? I intend to take a look ...

    Posted

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

    There are some very nearby spirals - both Eos and not - which have two (or more) spectroscopic objects ("objects with spectra") that are further apart than 2'. Are any among the 3002 QS objects? I intend to take a look ...

    The answer seems to be "no".

    There are a handful of QS Eos which measure > 2' end-to-end, and one that's got spectra of two regions, a long (angular) way apart (~1.4'), AGS00001b0:

    enter image description here

    And there's another that is much larger, and has spectra of two regions ... but those regions are much closer, AGS00000to:

    enter image description here

    Posted

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

    There are some very nearby spirals - both Eos and not - which have two
    (or more) spectroscopic objects ("objects with spectra") that are
    further apart than 2'. Are any among the 3002 QS objects?

    NGC 4330 has 2 spectra roughly equidistant from the presumed (but hidden) nucleus that are more than 2' ~1.4' apart (per Jean). You won't find the second one in DR9 or DR10 navigate however (edit) nor is it in this list because the second spectrum isn't a science primary and therefore isn't in specObj.

    An interesting case that one would be tempted to call an overlap accidental juxtaposition but probably isn't is NGC 1268/NGC 1267, which is among the neighbors sample. The have a rather high relative velocity of ~2000 km/sec. but both are considered secure members of the Perseus cluster. I don't think this is in Keel's catalog, but his expertise could be useful. Note in the DR10 finder chart image below that NGC 1268 (a QS object) is somewhat extended to the south and overlaps the low surface brightness outer parts of NGC 1267.

    enter image description here

    Posted

  • mlpeck by mlpeck

    This had been my first candidate to remove from the QS objects with neighbors sample. It has the highest redshift in the preliminary list (z = 0.298) and the largest projected distance from its prospective neighbor (~ 600 kpc*), but:

    The quench sample object is above and just to the left of center while its identified neighbor is SW of the center in this finder chart image. It turns out there's a BOSS target that's in the DR10 release that's closer on the sky and has about the same redshift. There's at least one other object in the general area in the same redshift range and many fuzzies with photometric redshifts in the same general range. This particular object is in Goto's 2007 catalog of K+A galaxies and there's a cluster candidate near this position in Lopes et al. (2004).

    enter image description here

    K+A galaxies were originally discovered (~1980 iirc) in intermediate redshift clusters, and that I think is what we're looking at here. So I'm moving this one from likely accidental juxtaposition to probable common cluster members. Below are zoomed in DR10 images of the QS object and its spectrum:

    enter image description here

    enter image description here

    (*) What's the correct distance measure to use? Angular diameter distance? Comoving transverse distance? I don't know! The quoted distance is the comoving transverse distance multiplied by the angular separation in radians.

    Posted

  • mlpeck by mlpeck

    I was checking to see why some specobjid's show up more than once in the neighbors list, and not too surprisingly most of them turn out to be associated with the pair of spectra that belong to the same galaxy. It really doesn't hurt to take a look at some of these and zoom out a bit: this particular galaxy is a group/cluster member with no obvious dominant central galaxy. In the DR10 finder chart image the program object(s) is the dusty disk a little ways SW of center. Many of the extended objects with spectra in the field are fellow group members.

    enter image description here

    Posted

  • mlpeck by mlpeck

    Update: I repeated the steps outlined in the first post using DR10 data throughout and got a somewhat expanded list of potential neighbors. There are now 597 unique DR10 specObjid's of potential neighbors of 472 unique specObjid's in the original quench sample. I've updated the data tables on both my Casjobs published tables and my dropbox account; the Casjobs version has 610 rows, which includes 13 repeated neighbor candidates.

    Here is a histogram of relative velocities of neighbor candidates with QS objects, given by the formula

    vrel = c(z-z0)/(1+z0)

    where z0 is a reference redshift -- in this case the redshifts of QS objects. This histogram looks like a mixture of distributions, so I used the R package "mclust" to fit a Gaussian mixture model to the data. Sure enough Mclust likes a mixture of 2 gaussians, both with means right around 0 and with standard deviations of 182 and 1243 km/sec. The mixture of densities is the blue curve in the histogram. Mclust makes a classification of group membership for each observation and also calculates the uncertainty of each classification. In this case the uncertainty is a maximum at about ±416 km/sec. and that is taken as the dividing line between the inner distribution and the more diffuse one.

    So, statistically at least and considering only one dimension the inner core of the distribution should contain the most promising candidates for physically associated galaxies. That's about 3/4 of the sample.

    enter image description here

    Here's another plot that's a standard tool in determining cluster membership, used for a slightly different purpose here. I've plotted relative velocities as defined above against projected distances. I'm still not exactly sure what's the appropriate cosmological distance measure, but the horizontal axis here is transverse comoving distance multiplied by angular distance (in radians). Red points here would be probable accidental juxtapositions by the relative velocity based classification. As I noted the other day one of the pairs separated by ~600 kpc. are actually probable common cluster members, so we'll see eventually (probably when I get back to where I can keep multiple windows open on a pair of monitors). The DR10 based sample update has added a second pair with projected separation > 600 kpc. I have not yet checked that one.

    enter image description here

    Posted

  • mlpeck by mlpeck

    I'm not going to consistently post GZ uid's, but this is AGS000000d, which is at least interacting with the merging pair to the east. Not previously noted as merging:

    enter image description here

    Posted

  • mlpeck by mlpeck

    Trouille and Zutopian caught this as a merger, but the GZ classifiers did not: AGS000000x. The other conspicuously white galaxy to the NW of center doesn't have an SDSS spectrum, but according to NED it is UGC 06361 at z=0.024, which makes it a physical member of this triplet.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS000000w appears relatively isolated in this finder chart image, but it and the faint disk galaxy to the NE belong to the same group as the previous example along with a few others nearby. Projected distance from the merging triplet is ~ 125 kpc.

    enter image description here

    These examples were from the first page of DR10 image list thumbnails of neighbor candidates.

    Posted

  • mlpeck by mlpeck

    One more for now: AGS00001tp. Not previously noted as a merger, but the disk appears warped in this DR10 image and the face on disk nearby seems to have suffered some tidal disruption. There are 5 spectroscopic objects in this finder chart field all with z ≈ 0.03.

    enter image description here

    Posted

  • mlpeck by mlpeck

    DR10 finder chart image of AGS000005m - spectroscopically confirmed companions at z = 0.06172, 0.06127 (QS galaxy is on the right).

    enter image description here

    Posted

  • mlpeck by mlpeck

    DR10 finder chart for AGS000005q, another pair of spectroscopically confirmed companions. The linear feature that appears to connect them is a background galaxy

    enter image description here

    Posted

  • mlpeck by mlpeck

    Spectroscopically confirmed companions. GZ classifiers saw "Tidal debris" in the program object AGS000005v

    enter image description here

    Posted

  • mlpeck by mlpeck

    Yet another confirmed pair (z=0.0604, 0.0599), not previously noted as a possible merger. The SDSS spectro pipeline calls the barred spiral a QSO, which basically means it has a broad-line AGN. AGS0000063

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS000007y. This 4 way merger has spectra for each clump. There are also repeated spectra of some of them including the one in the center of this finder chart image, which is the QS program object.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000dh: not classified as merger, but it looks tidally distorted to me in this DR10 image. Companion is spectroscopically confirmed (z = 0.0435, 0.0406)

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000li: spectroscopically confirmed companion (north of the centered QS galaxy; z= 0.264, 0.261). These are in a small group.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000lz: not previously noted as a potential merger. The object to the left of the QS object is a starburst galaxy and spectroscopically confirmed neighbor (z= 0.129, 0.132).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000o6. Spectroscopically confirmed merger (z = 0.168, 0.167).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000ok: spectroscopically confirmed companions (z= 0.068, 0.068). GZ classifiers called this a merger. The smudge to the upper left obscured by the label is a foreground galaxy.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000s2: the clump to the west of the QS object nucleus has a spectrum. Not previously noted as a merger candidate.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000un. The faint smudge to the NW of the program object has a spectrum, and its a confirmed companion (z = 0.0269, 0.0267). Not previously noted as a possible merger.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000vk: confirmed companion (z = 0.1406, 0.1401). Classified as "tidal debris".

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000wb: neighbor is spectroscopically confirmed companion (z = 0.0855, 0.0852). Not previously identified as a merger candidate. The pointlike white object left of center is the program galaxy and the spiral is the companion. The red ETG is in the background.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00000z6: GZ classifiers correctly called this a merger. The program object is the small galaxy. The larger galaxy is a BOSS target and therefore new to DR9 or DR10.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS000011x: the redder galaxy NW of the program galaxy is actually in the background -- its redshift is z = 0.089 while the QS galaxy is at z = 0.069.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS0000150: not previously noted as disturbed, but the spiral galaxy to the left and ETG to the right are confirmed companions (z = 0.0772, 0.0784, 0.0775).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS000016t. Not previously noted as a merger candidate, the galaxy to the south is a confirmed companion of the QS galaxy (z = 0.0881, 0.0885). I would call this disturbed based on the DR10 finder chart image.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS000017p: these appear to me to be interacting and are confirmed companions (z = 0.0626, 0.0628). Not previously noted as merger candidate. QS object is SW of center.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS0000182: this was called a merger, even by Zutopian but it's not! At least if the current redshift estimates are correct. The QS spectrum is in a region below the nucleus, which is a BOSS target. The red "cloud" is actually a background galaxy at z = 0.095.

    enter image description here

    Posted

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

    AGS0000182: this was called a merger, even by Zutopian but it's not! At least if the current redshift estimates are correct. The QS spectrum is in a region below the nucleus, which is a BOSS target. The red "cloud" is actually a background galaxy at z = 0.095.

    Yes, you are right.: It is indeed an overlap. It is listed in the Galaxy Zoo overlap catalog.

    PS: I deleted it from my collection "QS mergers" just now.

    Posted

  • mlpeck by mlpeck

    I don't think this has been noted before, and this is the first example I've seen. AGS0000190 and AGS0000191 are both in the quench sample and are spectroscopically confirmed neighbors (z = 0.0347, 0.0342).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00001bl: maybe a coincidence but our program galaxy is NGC 4465, a possible member of the Virgo cluster, and on the sky it's in the outskirts of the giant elliptical NGC 4472. Redshift of the program object is z =0.024, while NGC 4472 is at z ≈ 0.003.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00001cn. This is the second Perseus cluster galaxy in the quench sample. Looks disturbed to me in this DR10 image

    enter image description here

    Posted

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

    AGS00001cn. This is the second Perseus cluster galaxy in the quench sample. Looks disturbed to me in this DR10 image

    Curiously, the redshifts in SDSS and NED/SIMBAD differ.: SDSS:z=0.017 versus NED:z=0.0205 and SIMBAD: z=0.0207

    1237670960021111003 (z=0.021*) might be a companion, if the QS galaxy's redshift in SDSS is wrong.

    *SDSS redshift (no redshift in NED/SIMBAD)

    Posted

  • mlpeck by mlpeck

    Curiously, the redshifts in SDSS and NED/SIMBAD differ.: SDSS:z=0.017
    versus NED:z=0.0205 and SIMBAD: z=0.0207

    1237670960021111003 (z=0.021*) might be a companion, if the QS
    galaxy's redshift in SDSS is wrong.

    Its redshift in SDSS can't be wrong by that much. Hα and [N II] are fairly prominent in emission and Na D and other prominent absorption lines are present and all consistent with z = 0.0175. But in any case both the program galaxy and its nearest neighbor with an SDSS redshift are almost certainly members of the Perseus cluster. The cluster velocity distribution is ≈ 1300 km/sec. (Struble and Rood 1999) -- these two galaxies have a relative velocity of ≈ 1000 km/sec. and both have peculiar velocities < 1000 km/sec relative to the adopted redshift of the cluster, which is 0.0179 according to NED.

    Posted

  • mlpeck by mlpeck

    AGS00001ns: not previously noted as a merger candidate, but there appears to be a bridge of material connecting it to the ETG to the south. Spectroscopically confirmed companions (z = 0.0774, 0.0777).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00001w6: not previously noted as a merger candidate, the object to the east is at the same redshift (z = 0.0307, 0.0306). Also, the QS galaxy has two spectra -- one is a BOSS spectrum centered on the nucleus.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00001wf. The nucleus and a bright clump to the NW both have spectra. Not sure whether to call this one object with two spectra or a merger (GZ classifiers called it a merger, probably because of the galaxy to SW).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00001y5: not previously noted as a merger candidate, but looks obviously disturbed to me in this DR10 image. QS object is N of center. The compact object near center and spiral to the south are companions (z = 0.0344, 0.0343, 0.0342).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00001yl: not previously noted as a merger candidate, but note the apparent long tail of material from the QS object on the opposite side of the larger ETG (z = 0.0684, 0.0682).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS0000217. Confirmed companions (z = 0.1631, 0.1635). GZ classifiers saw tidal debris.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS0000272 has two spectra. QS spectrum is the northern one.

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS000027g: confirmed companions not previously noted as merger candidate (z = 0.0497, 0.0497).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS000028y. Confirmed companions not previously noted as merger candidate (z = 0.1677, 0.1685).

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS000029a. SDSS pipeline classifies the clump on the right as "QSO".

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS00002aq. ETG to NE is a confirmed companion (z = 0.0593, 0.0578). GZ classifiers saw "tidal debris."

    enter image description here

    Posted

  • mlpeck by mlpeck

    Having finally gone through the entire sample, here is a quick summary.

    First, in an effort to cast a slightly larger net I repeated the sample selection using the table "SpecObjAll," which is supposed to contain all spectra instead of the "SpecObj" view, which is supposed to contain only primary spectra. I removed repeated observations of neighbors by selecting on reported position. I ended up with 476 unique specObjId's of quench sample objects and 617 total neighbors, which is an increase of 4 and 20 from the second round.

    I counted 70 spectroscopically confirmed mergers and another 18 that I considered iffy. Those include one 3 way merger and one 4 object merger.

    For 20 objects the "neighbor" is in the same galaxy. Not all of these were exceptionally large on the sky or nearby. Here is a list of those 20. I worked entirely in DR10 databases, so these are DR8+ specObjid's. I'll publish a full table, quite possibly with DR7 ids and GZ uid's later.

    specObjid0	z0	ra0	dec0	specobjid	ra	dec
422255656800643072	0.09620748	336.0889	-0.1723252	4729926765563609088	336.0905	-0.1734522
468382364330911744	0.04746294	56.36828	-0.4396844	2972392871556048896	56.36799	-0.4370324
646436185706620928	0.03274847	155.0912	4.888982	647484019683911680	155.092	4.890252
778164548044613632	0.01823853	12.39182	1.120978	443705473888184320	12.39889	1.115243
923302560738076672	0.03221132	253.5886	36.01525	922139277033236480	253.5865	36.01362
1066233574421719040	0.01448957	159.6799	56.55413	1067423246288185344	159.6792	56.5527
1091158678909773824	0.003230308	182.8028	50.4946	1092240048109152256	182.8068	50.49707
1454830675944302592	0.008819454	230.4938	41.12542	1889293636382779392	230.4967	41.12713
1638346594245961728	0.02885401	192.6636	47.93427	1640532698692872192	192.666	47.93341
1690083300179208192	0.01842817	21.06192	1.046928	783750617037301760	21.06221	1.048933
1817236037093582848	0.005579995	185.8337	11.37571	1817258851959859200	185.8128	11.36378
1830885655224281088	0.02436449	187.3494	8.025373	1831965925415348224	187.3488	8.02658
1887025613974300672	0.0472241	225.5617	44.23249	6808527945217867776	225.5592	44.23323
2341998306273552384	0.04699261	118.6338	16.8042	2161728426493372416	118.6324	16.8073
2364476178326120448	0.04275434	169.9761	33.09011	2372279688451090432	169.9766	33.09029
2474857803345324032	0.03095886	243.9801	15.2965	4587015271801683968	243.9805	15.29598
2840728554797295616	0.04623846	238.4678	10.31506	2837298903353681920	238.4672	10.31602
2981408042454640640	0.02340478	191.277	21.17125	2943212932583942144	191.2773	21.16957
3118777727113521152	0.01685593	209.5045	19.68366	3103139648076015616	209.5062	19.68576
3323763528992581632	0.1273873	228.6403	2.957841	4518499478755868672	228.6409	2.958282

    Posted

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

    For 20 objects the "neighbor" is in the same galaxy. Not all of these were exceptionally large on the sky or nearby. Here is a list of those 20. I worked entirely in DR10 databases, so these are DR8+ specObjid's. I'll publish a full table, quite possibly with DR7 ids and GZ uid's later.

    In most cases there is actually no "neighbor" in the same galaxy.: Spectra was taken from different parts (nuclear and non-nuclear) of a galaxy.

    e.g. AGS000005p "neighbor" with spectrum in the same galaxy
    e.g. AGS00000iz just one galaxy with 2 spectra

    Posted

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

    Was I not clear about something? There are 20 quench sample objects with multiple spectra at different positions within the same galaxy. Or perhaps it's 21 depending on how you interpret the off-center clump in AGS00001wf. I called that a merging pair.

    Edit: That's 20 or 21 in addition to our friends AGS0000080 and AGS00000j6, which are of course both quench objects that inhabit the same galaxy. That galaxy also has a number of spectroscopically confirmed neighbors.

    I will in due course produce a spreadsheet with cross-ids and ancillary data on the putative neighbors.

    Posted

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

    Was I not clear about something? There are 20 quench sample objects with multiple spectra at different positions within the same galaxy.

    I know, that those listed 20 QS galaxies have multiple spectra at different positions.
    I just wanted to remark, that most aren't mergers, but single galaxies, as you know. (There are no overlaps regarding the redshifts of the listed specobjds.) BTW, there is the related topic "QS objects whose spectra are not nuclear", as you know.: http://quenchtalk.galaxyzoo.org/#/boards/BGS0000008/discussions/DGS0000220

    Or perhaps it's 21 depending on how you interpret the off-center clump in AGS00001wf. I called that a merging pair.

    AGS00001wf is also like the other 20 cases. (Image had been posted by you in this topic on page 4.)
    Curiously, DR9 and DR10 differ.: According to DR9 the off-center clump has the specclass star, but according to DR10 the specclass is galaxy.
    If DR10 is right, it is a merger, because the QS galaxy and the off-center clump have similiar redshifts in DR10. It might be even a triple merger.: There is a 3rd galaxy, but it has no redshift in SDSS. The GZ classification result is "Merging", as mentioned by you on page 4.
    I know, that there are some cases, where there are redshift discrepancies between DR7 and DR8., e.g. You had started a related topic, but the QS spectra aren't concerned.
    I am astonished about the above case, where there is a redshift discrepancy between DR9 and DR10. It is a BOSS spectrum and the date of the taken spectrum is identical in DR9 and DR10. Well, the date of the taken spectrum in the other cases (DR7/DR8 redshift discrepancies) is also identical, but I am somehow astonished, that DR9 and DR10 state a different redshift at a BOSS spectrum.

    EDIT:

    I will in due course produce a spreadsheet with cross-ids and ancillary data on the putative neighbors.

    Could you please consider to post first a list, which contains the 70 spectroscopically confirmed mergers, which you found, and the 18 iffy mergers? You had informed about these in your previous post.

    Posted

  • mlpeck by mlpeck

    Curiously, DR9 and DR10 differ.: According to DR9 the off-center clump
    has the specclass star, but according to DR10 the specclass is galaxy.
    If DR10 is right, it is a merger, because the QS galaxy and the
    off-center clump have similiar redshifts in DR10. It might be even a
    triple merger.:

    I am astonished about the above case, where there is a redshift
    discrepancy between DR9 and DR10. It is a BOSS spectrum and the date
    of the taken spectrum is identical in DR9 and DR10. Well, the date of
    the taken spectrum in the other cases (DR7/DR8 redshift discrepancies)
    is also

    This isn't really too surprising or unusual. The SDSS spectro pipeline fits stellar, galaxy, and QSO templates to every spectrum with a range of assumed redshifts. The best fitting combination of templates "wins." The SMALL DELTA CHISQ flag in the spectrum indicates that at least two template fits worked nearly as well. There's at least one other example like this in either the quench or control sample.

    It's possible that "clump" is really a foreground star that's contaminated the spectrum enough that the spectrum was classified a star in one run and a galaxy in the next. Or the star classification could just have been an error due to the low signal to noise of the spectrum.

    Posted

  • mlpeck by mlpeck

    There's one more noteworthy pair of galaxies that haven't previously gotten much attention. AGS000004s and AGS000004t are both on the outskirts of the cluster Abell 2255 and fairly close to each other (~ 320 kpc transverse separation). The two are NW and SE of center in this DR10 finder chart image:

    enter image description here

    Posted

  • mlpeck by mlpeck

    I've posted a table of quench sample neighbors to my dropbox account in both CSV and open document spreadsheet formats. The ODS file has a second sheet with a key briefly explaining the contents of each column. Links are:

    https://www.dropbox.com/s/38deecdjfp0pfet/qndb.csv

    https://www.dropbox.com/s/sgfb13izctjygq6/qndb.ods

    I would not recommend using Excel on these. It seems to insist on converting ID strings to numbers.

    Posted

  • zutopian by zutopian

    AGS0000026 ra: 201.704, dec: 1.223, z=0.081 (H-alpha line is missing.)
    AGS0000027 ra: 201.809, dec: 1.237, z=0.081

    Posted

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

    AGS00000z6: GZ classifiers correctly called this a merger. The program object is the small galaxy. The larger galaxy is a BOSS target and therefore new to DR9 or DR10.

    It is a special merger, because it has a Dual AGN according to DR10. Besides there is bluish stuff, which might be a voorwerpje.

    PS: This merger is in the QS 778 sample.

    Posted

  • zutopian by zutopian

    AGS000023v is however missing in the dropbox list, though it has a companion with same redshift z=0.034.

    enter image description here

    Posted

  • zutopian by zutopian

    AGS000017x Dual AGN according to DR10.:

    enter image description here

    Posted

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

    AGS000023v didn't meet my selection criteria. I set an upper limit of 2' for neighbor selection and if I calculated the distance correctly these are 2.24' apart.

    Happy New Year.

    Posted

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

    Thanks for your reply.

    enter image description here
    http://writing.galaxyzoo.org/x4q528/

    Posted

  • zutopian by zutopian

    AGS00000nn is missing in the dropbox list, though the other galaxy has same redshift (z=0.046). I guess, that they also didn't meet your selection criteria, but I don't know, how to calculate the distance.

    enter image description here

    Posted

  • zutopian by zutopian

    AGS00001xa isn't in the dropdox list. Both galaxies have z=0.051.

    enter image description here

    Posted

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

    The separations on the sky for the last two were 2.5 and 2.14 arc-minutes respectively.

    I used the SDSS cross-id tool to compile an initial list of neighbor candidates, and then narrowed it down as described in an early post (later updated somewhat). Assuming the SDSS spectrum database is complete this should have produced a list of all possible observed neighbors that met my admittedly over-simplistic search criteria.

    See http://en.wikipedia.org/wiki/Great_circle_distance for formulae to calculate distance on the sky from sets of coordinates. The "from chord length" version is easy to implement and numerically well behaved.

    Posted

  • JeanTate by JeanTate

    Are you considering continuing with this mini-project?

    Posted

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

    I'm not sure what else to do with it actually. There are much more sophisticated algorithms to select neighbors, but I wouldn't know how to implement them with SQL queries.

    Should the quench project be revived it might be worth noting that at least a small number of objects are confirmed to be members of groups or clusters, and this will help narrow down the possible mechanisms for quenching of star formation. Also, objects like the couple of confirmed multi-way mergers ought to be promising targets for follow-up studies.

    Posted

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

    There are much more sophisticated algorithms to select neighbors, but I wouldn't know how to implement them with SQL queries.

    IIRC, some of these are actually described, in some detail, somewhere in the SDSS website.

    Should the quench project be revived it might be worth noting that at least a small number of objects are confirmed to be members of groups or clusters, and this will help narrow down the possible mechanisms for quenching of star formation.

    True, but trying to get a handle on the selection effects - 55", fiber collisions, etc - would not be easy or straight-forward, I would think.

    objects like the couple of confirmed multi-way mergers ought to be promising targets for follow-up studies.

    Yep ... fancy trying to write a proposal for time on something like the HST or a Gemini? 😉 (that's the 'wink' smilie)

    Posted

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

    Are you considering continuing with this mini-project?

    Here's a possible continuation. I posted a version of this plot on page 1 of this thread:

    enter image description here

    This is a plot of relative velocities (in km/sec) vs. transverse separations (kpc) for all neighbor candidates of the full quench sample. I ran a "model based clustering" algorithm on this with distances assumed to be drawn from a mixture of log-normals. The algorithm likes 3 distinct classes as indicated by the different colored points. The filled circles are neighbors of subset 2 quench sample objects.

    The red points evidently are the most likely actually to be physically interacting (although the classifications are purely statistical), and these seem deserving of special scrutiny. If I counted right there are 55 unique QS objects in the most likely interesting class, with 63 neighbors. A few of the neighbors are probably spectra from the same galaxy, but most are genuinely interacting.

    Posted

  • mlpeck by mlpeck

    Here are the 55 quench sample objects with confirmed neighbors that are most likely actually to be interacting -- remember per the above post that these were selected purely on statistical criteria applied to velocity differences and transverse separations. This list just gives uid, ra, and dec suitable for copying and pasting into the SDSS image list tool.

    Below that, so you don't have to guess, is a list of the 63 neighbors. First column is DR8+ specObjid.

    By my count 6 of these are single objects with multiple spectra. One of them however is in a group and appears to be interacting with other members of the group. One disk galaxy looks slightly warped and disrupted to me, so conceivably that is an actual minor merger. The other 4 are ordinary disk galaxies where a clump in the disk got targeted.

    Yes, I'm aware there's a bright foreground star close to the fiber location in one merging galaxy pair.

    uid,ra,dec
AGS000000d,156.8542,1.241761
AGS000000x,170.2041,0.4592703
AGS000005m,334.2885,-0.4592362
AGS000005p,336.0889,-0.1723252
AGS0000063,355.049,0.1680176
AGS000006f,13.11726,0.3760174
AGS000006m,19.92594,1.131008
AGS000006u,29.56817,-0.5219178
AGS000007y,53.05504,0.2629581
AGS0000080,53.52493,1.094325
AGS0000082,56.36828,-0.4396844
AGS000009r,132.7135,54.82316
AGS00000bw,160.1344,2.490956
AGS00000cd,186.2534,1.733861
AGS00000cq,202.6031,2.390516
AGS00000ep,155.0912,4.888982
AGS00000fs,206.6339,64.39033
AGS00000j0,14.58846,0.05364013
AGS00000n4,234.8665,52.9974
AGS00000ok,177.0525,5.999049
AGS00000ov,191.5578,5.040455
AGS00000un,318.5023,0.5351073
AGS00000vn,175.7858,54.19743
AGS00000wi,332.6292,-0.8442407
AGS00000wl,337.5737,1.18082
AGS00000xz,222.9339,52.58641
AGS00000yk,141.3519,6.633405
AGS00000yv,118.3462,24.55016
AGS00000z6,165.0749,10.04904
AGS0000103,127.2719,27.08174
AGS0000107,130.3277,29.48968
AGS000010e,135.0251,33.07647
AGS0000141,167.0173,44.57936
AGS0000148,176.7858,45.53121
AGS000016t,172.229,47.21678
AGS00001cj,245.9379,23.1299
AGS00001d0,225.5617,44.23249
AGS00001d5,226.1044,43.39985
AGS00001dm,255.0287,23.13151
AGS00001gv,155.4283,13.11559
AGS00001hk,179.2943,13.47501
AGS00001oi,171.4863,31.88707
AGS00001qk,159.3087,31.14991
AGS00001rc,199.2168,38.71033
AGS00001rm,118.6338,16.8042
AGS00001s7,181.6669,31.548
AGS00001wf,243.9412,14.87368
AGS00001x1,172.1449,26.70269
AGS00001y5,200.7495,27.11643
AGS0000230,175.3076,21.93948
AGS0000242,238.4678,10.31506
AGS0000254,149.4083,16.96779
AGS000029a,207.3166,22.00754
AGS000029n,220.5848,18.31668
AGS00002a9,224.5741,18.31847

    Neighbors:

    specobjid,ra,dec
567479705305901056,156.8729,1.246717
315360852611983360,170.2,0.4686064
422297163364591616,334.2957,-0.4605812
4729926765563609088,336.0905,-0.1734522
769086435742476288,355.0436,0.1735857
4743629979081244672,355.0436,0.1735697
779239870450132992,13.11898,0.3757897
782666773332781056,19.92329,1.127731
789319917806577664,29.5673,-0.522525
802895038102136832,53.05302,0.2620322
906515494247360512,53.05566,0.2637156
907578997372643328,53.05324,0.264555
1198129892563118080,53.05565,0.2637
1301656884851271680,53.05325,0.2645792
802910431264925696,53.52546,1.094434
2972392871556048896,56.36799,-0.4370324
502322364369889280,132.7216,54.82705
568751015642294272,160.1367,2.489359
5350401246728028160,186.2589,1.734156
594562046155057152,202.5994,2.384546
647484019683911680,155.092,4.890252
680190088537204736,206.6626,64.40588
444861884756158464,14.59564,0.0557142
780358349542680576,14.58601,0.05766214
693614581846665216,234.8672,52.99814
944786467182372864,177.0502,5.99451
5353891646507843584,191.5659,5.042348
1714880022692521984,318.4992,0.540243
1143994615989299200,175.7927,54.20095
419963444029057024,332.6315,-0.8417053
423449726478739456,337.5761,1.179809
1495338334417872896,222.9442,52.58628
5483151887149039616,141.3491,6.62845
1044837080946993152,118.345,24.55069
6036086389991800832,165.0796,10.04742
1785793300138059776,127.2753,27.08199
1428803314941716480,130.3323,29.4874
1431177161166841856,135.0302,33.07581
1534765451479902208,167.0155,44.57134
1540328431080728576,176.7818,45.53112
1623668663116654592,172.2279,47.21181
4710873328683319296,245.9415,23.13145
6808527945217867776,225.5592,44.23323
6808355871648120832,226.0999,43.39759
1899461642136086528,255.0292,23.12753
6010264085022113792,155.427,13.11497
1812844038299084800,179.2962,13.47375
2228314028000700416,171.4757,31.89678
2217006374737438720,159.3077,31.15042
2225991310425221120,199.2073,38.71795
2161728426493372416,118.6324,16.8073
2162894458624960512,118.6342,16.80904
2256383740357928960,181.6637,31.5436
5191592184419713024,181.6693,31.54331
4587012248144707584,243.9405,14.87435
2498454970433562624,172.1381,26.70344
2526646997973231616,200.7552,27.09969
2819429083828480000,175.3173,21.93939
2837298903353681920,238.4672,10.31602
2908366653844645888,149.4058,16.97179
3133446037096654848,207.3133,22.00909
6157796280946393088,220.5835,18.31528
3127901749965252608,224.5763,18.31469

    Posted

  • mlpeck by mlpeck

    I wonder if perhaps there's an actual discovery here. I stumbled across a paper by Liu et al. 2011 containing a catalog of AGN pairs selected from SDSS DR7. They used similar -- slightly looser -- criteria to select potentially physically associated pairs, and examined each pair for morphological evidence of interaction. It turns out that 5 QS objects with neighbors were in their most interesting class, which were pairs with tidal features.

    Here is the list, followed by finder chart images from DR10. A couple of these are old friends. I've marked the objects with AGN and in some of the charts all objects with spectra. Note that AGS0000080 is in a small group -- 5 of the objects with spectra in the finder chart image are in the same redshift range. All 4 objects neighboring AGS000007y are at the ~same redshift.

    Object of the day anyone?

    uid,ra,dec
AGS000005q,334.6664,-0.4005234
AGS000006u,29.56817,-0.5219178
AGS000007y,53.05504,0.2629581
AGS0000080,53.52493,1.094325
AGS00000yv,118.3462,24.55016

    enter image description here

    enter image description here

    enter image description here

    enter image description here

    enter image description here

    Posted

  • mlpeck by mlpeck

    AGS000029a is another QS object that may belong to the same class, although it's not in their catalog. Its companion to the right is classified a QSO in the SDSS pipeline.

    enter image description here

    Posted

  • mlpeck by mlpeck

    And yet another dual AGN not in the Liu et al. catalog is AGS00000z6. The neighbor spectrum is from BOSS and doesn't show up in navigate. It didn't show up because I inadvertently went to DR9 Navigate and the spectrum is new to DR10. I've marked its position on the finder chart image. The QS object is the small galaxy at center that appears to be falling into the disk.

    enter image description here

    Posted

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

    Object of the day anyone?

    Sure! 😄 I take this to mean you'd like to become an OotD author, over in Galaxy Zoo's Object of the Day section, right? I'll make that happen later tonight (my time).

    An interesting follow-on: what about QC objects? Among "the 1196", there's at least one which has two spectra (two objects close together on the sky, with spectra whose redshifts are ~the same), and looks awfully like a 'messy merger', AGS00002vf. Comparing the prevalence of these sorts of mergers - QS vs QC - will be tricky, not least because, as we have found (if not yet quantified), mergers are more common in QS

    Posted