Quasar Tutorial

Look for high-redshift quasar candidates in selected fields (developed by E. Hatziminaoglou)

Uses VODesktopTOPCATVO services VOSEDVOSpec and


During this exercise we will identify high-probabilty quasar candidates based on their UV-optical colours. Quasars at different redshifts occupy distinct places in the optical colour space. We will first study the colour distribution of known, spectroscopically confirmed SDSS quasars from an early data release. We will identify the redshift region we are interested in. We will then try to identify quasar candidates on a more recent SDSS photometric data release catalogue. The selected colour regions will always be contaminated by the presence of other type of astrophysical objects; we therefore have to apply additional selection criteria based on what we know about quasars, e.g. high-z quasars are point-like, all quasars emit in the X-rays etc Once the selection has been narrowed down, the high-probability candidates will be checked againsta larger quasar catalogues for confirmation.

    1. launch VODesktop and TOPCAT and make sure they "see" eachother
    2. in VODesktop create a New Smart List -> Any main field contains SDSS + any main field contains quasars, then Create
    3. explore VODesktop: try to filter the resources list; check the various capabilities of each resources; check the "Information" and "Table Metadata" tabs at the bottom of the main window; see how the available "Actions" (left column of the main window) change with the selected resource
    4. select the first SDSS quasar catalogue: Schneider et al. (2002); have a look at the table "Information", also check the "Table Metadata"
    5. send the table to TOPCAT using the suggested action (under "Actions", lower left part of the VODesktop main window)
    6. the table we are interested in is the third one in the TOPCAT list, with 3814 entries; selected and click on "Display column metadata" to check the table contents
    7. make various colour-colour plots, then add the redshift as a third column. The should look like that:


    1. decide on the redshift region you are interested in and find a colour-colour plot that separates quasars in this redshift bin as clearly as possible from the rest; note down the colour region
    2. back in VODesktop, select a later photometric SDSS data release (DR5, DR6 or DR7) and "Query" around a position you know it has been covered by that particular SDSS data release; keeping in mind that the SDSS quasar density is about 11/sq deg, try to keep your queries reasonable, i.e. do not give a search radius larger than one deg; you can check for the covergare in the VO Services website:
  • click on "Footprint Services" under VO Services
  • click on "Search" and type SDSS
  • click on the corresponding thumbnail; then "Plot"; "Next"

  1. send the result of your query back to TOPCAT
  2. create the same colour-colour plot as before and identify the objects that fall into the high-z quasar candidates region, by overplotting them on the previous catalogue
  3. select the candidates by drawing a region on the photometric catalogue; save the newly created subset
  4. high redshift quasars (in fact anything with roughly z > 0.3) will be point-like on the SDSS images; filter out the list of candidates based on the morphology (columns type_* contain this information; *=u, g, r, i, z); 3 stands for extended, 6 for point-like
  5. from VO Desktop again (or from within TOPCAT) select the latest SDSS quasar catalogue (from DR5, Schneider et al. 2007; but various other catalogues are available), send it to TOPCAT and cross-correlate with your list of quasar candidates; how many did you find? try changing the selection criteria to see how the completness of the selection changes
  6. you can now create and visualise the SEDs of the objects using VOSED and VOSpec
  • if their number is small, you can try the Single Object Search; unmark the Spectroscopic data; select 2MASS and SDSS from the Photometric Data, select a small search radius (typically 1-2") andSubmit query
  • you can then Display SED; this will launch a new window of VOSpec; click on the list at the bottom right of the main VOSpec window and Retrieve
  • in case the list of candidates is long select the Multi Object Search (VOSED main window) and repeat the previous steps; the results will come in a .zip file that will contain an xml file per object; you can visualise the individual SEDs by loading them one by one in VOSpec
  • using any of the VO tools, find multi-wavelength couterparts for your objects, e.g. X-ray, IR or radio; you can try this also with Aladin