by Ivan Zolotukhin (IRAP, Toulouse / Sternberg Astronomical Institute) and Igor Chilingarian (Smithsonian Astrophysical Observatory / Sternberg Astronomical Institute)
0. Download jar files of VOSpec and TOPCAT if you do not have them. One can easily find them using Google. To be on the safe side, here is the links:
1. Launch them on Linux (in terminal): java -jar topcat-lite.jar. On Windows and Mac they can be launched by double-click on a jar-file. Alternatively, launch both applications using Java WebStart, if your browser supports this method (most of browsers do): click on the link http://esavo.esac.esa.int/webstart/VOSpec.jnlp to launch VOSpec, on http://aladin.u-strasbg.fr/java/nph-aladin.pl?frame=get&id=aladin.jnlp to launch Aladin, and on http://www.star.bris.ac.uk/~mbt/topcat/topcat-lite.jnlp to launch TOPCAT.
2. Load Chilingarian & Zolotukhin catalog. In TOPCAT do: File → Load table → DataSources → Table Access Protocol (TAP) Query → TAP URL: http://gal-02.sai.msu.ru:8080/__system__/tap/run/tap → Enter Query:
SELECT
objid, mjd, plate, fiberid, ra, dec, age_ssp, age_exp, met_ssp, met_err_ssp, dispvel_ssp, dispvel_err_ssp, z, corrmag_nuv, corrmag_g, corrmag_r, corrmag_z, corrmag_k, kcorr_nuv, kcorr_g, kcorr_r, kcorr_z, kcorr_k, petror50_r
FROM
specphot.catalog_czkm
As a first exercise, we propose visualisation of a Cosmic Web, the largest scale structure of the Universe. For that purpose do a 3D plot (Graphics → 3D) in TOPCAT with the following axes:
X: cos(ra / 57.3) * cos(dec / 57.3) * z * 3e5 / 72
Y: sin(ra / 57.3) * cos(dec / 57.3) * z * 3e5 / 72
Z: sin(dec / 57.3) * z * 3e5 / 72
color: corrmag_k - kcorr_k - 5 * log10(z * 3e5 / 72) - 25
For better results set up limits on color axis to the range -24 … -17.
Hint: it is actually much easier to do a 3D radial spherical plot (Graphics → Sky → Plot points with radial …) and just add z * 3e5 / 72 as a radial coordinate to get essentially the same distribution.
3. Add new column, gr_fit, using this expression:
0.0008569*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),0)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),0) +0.4145246*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),0)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),1) -0.3126628*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),0)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),2) +0.1915254*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),0)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),3) -0.0604829*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),0)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),4) +0.0100710*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),0)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),5) -0.0008631*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),0)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),6) +0.0000304*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),0)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),7) +0.1037934*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),1)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),0) -0.2982120*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),1)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),1) +0.2527798*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),1)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),2) -0.1029656*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),1)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),3) +0.0219900*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),1)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),4) -0.0023795*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),1)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),5) +0.0001031*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),1)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),6) -0.0146987*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),2)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),0) +0.0487196*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),2)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),1) -0.0351197*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),2)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),2) +0.0100174*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),2)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),3) -0.0012482*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),2)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),4) +0.0000568*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),2)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),5) +0.0003963*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),3)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),0) -0.0029259*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),3)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),1) +0.0021275*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),3)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),2) -0.0004860*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),3)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),3) +0.0000349*pow((corrmag_z-kcorr_z-25-5*log10(luminosityDistance(z,72.0,0.3,0.7))+21.6665),3)*pow((corrmag_NUV-corrmag_r-kcorr_NUV+kcorr_r),4)
4. Create 3D plot using following axes which represent (Mz, NUV-r, g-r) parameter space respectively:
X: corrmag_z - kcorr_z - 25 - 5 * log10(z * 3e5 / 72.0)
Y: corrmag_NUV - corrmag_r - kcorr_NUV + kcorr_r
Z: corrmag_g - corrmag_r - kcorr_g + kcorr_r
After inserting these expressions for the coordinates, please set up the following axes limits for proper plot appearance:
X: -25 … -16
Y: 0 … 7
Z: 0 … 1
5. Add new 3D plot to the previous one using gr_fit column as Z axis keeping same X and Y axes. This plot represents universal optical-ultraviolet color-color-magnitude relation (surface) for normal galaxies (see http://adsabs.harvard.edu/abs/2012MNRAS.419.1727C for the original paper and coefficients of the surface that we have entered when creating gr_fit column at step #3.
6. Select continuous region of cE (compact elliptical galaxies) outliers which reside above the fitted surface at low luminosity part. You can draw it by hand but for more formal criterion we suggest you to make the following filter in TOPCAT:
(corrmag_NUV - corrmag_r - kcorr_NUV + kcorr_r) > 4.0 &&
(corrmag_g - kcorr_g - 25 - 5 * log10(luminosityDistance(z, 72.0, 0.3, 0.7))) > -18.7 && (petror50_r < 2.0 || petror50_r / 206.265 * luminosityDistance(z, 72.0, 0.3, 0.7) < 0.7) && (corrmag_g - corrmag_r - kcorr_g + kcorr_r - gr_fit) > 0.03 &&
age_ssp > 4000.0 && dispvel_ssp > 60.0
We also found it useful to change the last condition to (dispvel_ssp > 60.0 || dispvel_exp > 60.0)
7. Explore the outliers in the SDSS data: 1) setup “Transmit coordinates” activation action in TOPCAT and 2) in Aladin do File → All sky → Image → Optical → SDSS colored to load all sky SDSS image and zoom to the desired level of details. Once you click on a row in TOPCAT, Aladin will re-point to an object’s coordinates in SDSS imagery.
8. Filter cE candidates by their angular size and/or emission line properties. For this purpose one needs to do TAP table upload cross-match query with the main catalog (table catalog_czkm), emission lines catalog (table catalog_czkm_lines) and the cE candidate outliers selected at step #6. It is also possible to do a cross-match with SDSS DR7 TAP service (can be found from the TOPCAT query registry interface) to restrict the sample on other parameters not present in Chilingarian & Zolotukhin catalog. Cross-match with emission lines table can be achieved with this ADQL query:
SELECT
TOP 1000 *
FROM
specphot.catalog_czkm_lines AS db
JOIN
TAP_UPLOAD.t2 AS tc
ON
db.mjd = tc.mjd
AND db.plate = tc.plate
AND db.fiberid = tc.fiberid
For selecting objects based on their emission line properties, you may want to use following coordinates (or their logarithm):
X: f6565_h_alpha_sm / (f6550_nii_sm + f6585_nii_sm)
Y: f4863_h_beta_sm / f5008_oiii_sm
9. Verify candidates by spectra downloaded from our SSAP service (endpoint URL is http://gal-02.sai.msu.ru:8080/specphot/ssap.q/ssa/ssap.xml?request=queryData&) and store their name, RA and Dec in space-separated CSV file. You may want to do Joins → Multiple SSA query in TOPCAT to download spectral metadata for all candidates at once and explore their spectra further in VOSpec.
10. Check HST archive at CADC for better imaging data using ObsTAP access and caom.py script.
11. Load value-added catalog of Groups and clusters of galaxies in the SDSS DR8 by Tempel et al. (2012) either from VizieR or from the disk and cross-match it with Chilingarian & Zolotukhin catalog using 3 arcsec sky match. You may want to restart TOPCAT before doing cross-match of half a million rows tables using the following command in the terminal to give it more memory: java -Xmx400M -jar topcat-lite.jar. Filter resulting table by meaningful absolute z magnitude (corrmag_z - kcorr_z - 25 - 5 * log10(luminosityDistance(z_1, 72.0, 0.3, 0.7))) and make it a default subset.
12. Create 3D plot using following axes which represent evolution of luminosity-metallicity relation depending on galaxy group size:
X: met_ssp
Y: Nrich (and tick flip axis for convenience)
Z: corrmag_z - kcorr_z - 25 - 5 * log10(luminosityDistance(z_1, 72.0, 0.3, 0.7))
When you explore this figure, it becomes evident, that luminosity-metallicity relation gets much wider for the field galaxies and it only exists essentially in groups.