Colour Segregation in Abell Catalogue

Background: The Abell Catalogue lists positions of some galaxy clusters. It does not give us colour information about the members of the galaxies. The Sloan Digital Sky Survey (SDSS) provides magnitudes in five bands (u,g,r,i,z). We should be able to match Abell objects with Sloan objects to get some colour information about objects in the Galaxies. Galactic cores should contain redder older stars while the edges should contain bluer younger stars according to current Galactic Morphology theories.

Software Needed: www.openskyquery.net, voplot

Which objects are in Sloan?

First we need to build a candidate list from the Abell catalogue. We can do this easily with www.openskyquery.net. We simply take the catalogue and upload import it to OpenSkyQuery(OSQ). We could search for and download the catalogue as per the HRDiagram example but it is already in the data directory of the summerschool.zip.

To import the catalog:

1. go to http://www.openskyquery.net/Sky/skysite/importtable.aspx
2. Lets put in a table name of Abell.
3. Select Format of VOTable XML, and RA/DEC in degrees .
4. Browse to find the file abell.xml (located in \nvoss\data\abell.xml in the NVO software distribution)
5. Hit import

To match this with SDSS and get a candidate list

1. go to http://www.openskyquery.net/Sky/skysite/Browse.aspx by clicking the Query button at the top of the web page.
2. click the (+) symbol by MyData (at the bottom)
3. click the (+) symbol by SDSS (in green (optical band) at the top)

You should now see a query like

SELECT a.unique_id, b.objid 
FROM  MyData:abell a, 
 SDSS:photoprimary b 
WHERE XMATCH(a, b) < 3.5

Lets modify this to

SELECT  a.name, a.redshift,b.objid ,b.ra,b.dec, b.specobjid,c.z as sredshift
FROM  MyData:abell a, 
 SDSS:photoprimary b,
 SDSS:photoz c 
WHERE XMATCH(a, b) < 3.5
and c.objid=b.objid

so we get the ra and dec from the sloan object and the objectid as well a redshift from photoz and a specobjid if it exists.

1. now select format CSV
2. hit submit
3. When done hit save on the right panel save to a file called AbellInSDSS.csv .

Plans are underway to make footprint serviceswhich would be more effective for this type of question.

Examining a cluster

We need to now get data from Sloan for the members of the cluster. But how much data should we get. The first cluster we have is

name,redshift,ra,dec,specobjid,sredshift,objid,xmatch_a,xmatch_ax,xmatch_ay,xmatch_az,pk,id,b_objid
ABELL91,0,10.8004689013586,-10.6191063173061,0,0.237198,582379153896637040,1,0.965461729159397,0.184171674688734,-0.184294448562333,3002,0,582379153896637040

We could assume anything within 1Mpc of the center would be in the cluster. But for that to work we need a distance to the cluster - luckily we have redshifts in the Abell we uploaded, unluckily they are for the most part zero which means unknown in that catalogue.

We could also ask OpenSkyQuery for the redshift from Sloan but our query shows us we have no spectra for those objects which we matched (specobjid =0). We do have a redshift from photoz called sredshift in the table.

The redshift we may calculate the distance - there is a WebService which can help here on http://voservices.net/Cosmology/ws_v1_0/Distance.asmx?op=AngularDiameter . Here we may plug in

z = 0.237198 – from above query
hubble=0.7   - from an expert !
omega=0.3    - from an expert !
lambda=0.7   - from an expert !

This gives us a distance in Mpc of 775.6411. The angular size then of our 1Mpc radius at a distance of 775.6411 which is 1/775.6411 or 0.001 radians. This is about 0.07 degrees or 4.4 arcmins.

Visually if we may of course have a look at http://skyserver.sdss.org/dr2/en/tools/chart/navi.asp for 10.8,-10.619.

Utilizing that we may return to OpenSkyQuery and do a circle search in this region for galaxies. We can also getthe redshift from the photz table. We utilize the following statement to get all galaxies in a 4.4 arcmin radius of 10.8, -10.619, type 3 in SDSS means galaxy - select output type as VOTable :

SELECT a.objid ,a.ra,a.dec,a.g,a.r,b.z
FROM  SDSS:photoprimary a , sdss:photoz b 
WHERE  a.type=3 and a.objid = b.objid and 
 Region('Circle J2000 10.8 -10.619 4.4')

Save this VOTABLE as Abell91.xml.

Plotting

Open the VOTable in Voplot with

voplot 

Use File-open to load abel91.xml.

We need a blueness index in SDSS g-r will give us a good indication so we add a column by selecting Functions-> Create New Column. Create blueness as $4 - $5 , you may do this by clicking on the relevant columns,

We also need distance from the center of the cluster .. this is a more complex formula and ideally we would have used the function in the database to do this. The new ADQL0,.8 will allow us to call functions in the target nodes. For this cluster the center is 10.8, -10.619 the formula for distance then would be

2 * todegrees( asin( sqrt( 
      pow((cos(-10.619*0.017)*cos(10.8*0.017))-(cos($3*0.017)*cos($2*0.017)),2) 
         + pow((cos(-10.619*0.017)*sin(10.8*0.017))-(cos($3*0.017)*sin($2*0.017)),2) 
         + pow((sin(-10.619*0.017))-(sin($3*0.017)),2)
      )/2 ) ) 

Now select blueness for Y and distance for X and hit plot (all in the right panel.

Filtering

Unfortunately this does not yield the type of plot one might expect. Of course we may have source other than cluster members in our data – so we probably want to trim to some redshift. This would be best done by histogramming the redshifts and picking the mean.

To make a histogram of z in voplot choose z for x axis hit histgram. Under view->plot properties, set the bin to 0.005. From this we may visually deduce that the cluster is between 0.26 and 0.4.

Under the functions menu we may create a filter called cluster members to do this with the formula:

$5 < 0.4  && $5 > 0.26

Select this filter in the right on voplot also reselct distance and blueness for the x and y axis and click plot. This is decidedly better.

Further work

To do a complete job on this would require some programming see ColourSegProgramming.