Exercise: Using the OpenSkyQuery web portal perform a brown dwarf search. This portal will provide you with an efficient network framework for performing complex queries and demonstrate flexibility in defining attributes for filtering.
Software needed: www.openskyquery.net, mirage
VO technology employed: VOTable, Conesearch,SIAP, Skynode/OpenSkyQuery
Background: An early scientific example of the NVO standard technology application to astronomical research was demonstrated by the successful discovery of a new L-type brown dwarf, 2MASSI J0104075-005328. In this prototype a web application was designed to extract data from SDSS and 2MASS databases using NVO standard cone search protocols and correlate fields of interest based on scientific constraints to filter brown dwarf candidate sources. Several advances have been made in the NVO since this time, especially with web services and the OpenSkyQuery portal. Validity of the findings need to be confirmed by follow-up data from other photometric and spectroscopic findings.
Scientific Criteria: In order to detect cooler and more distant distant brown dwarfs, knowledge of the 2MASS and SDSS survey wavelength detection limitations suggest the primary constraint:
PART I: Cross Correlation between Survey Archives 1) Begin with opening the OpenSkyQuery (OSQ) web portal at http://www.openskyquery.net/Sky/skysite/Browse.aspx. 2) On the left side listing of the Skynodes, you will see SDSS and TWOMASS surveys listed. You can look at the available node tables by clicking on the nodes. 3) On the right panel you will see samples listed. The second sample listed: X Match/Region 2, is fairly similar to what we will be exploring in this exercise. When you click on this it should appear in the main query box. You can submit this to see what a test run would look like. Note the sample surveys are SDSS and TWOMASS. 4) For performing brown dwarf search, enter the following SQL-like query which includes the scientific contraints (the magnitude difference was relaxed to see a larger set of cross matches) in the central panel of OSQ. You can either type it or copy and paste into the query window. Note SDSS represents the DR1 catalog in this current portal release.z-band in Sloan and J-band in 2MASS search for a 3-mag difference (J brighter than z) and a 2 arcsec spatial coincidence.
SELECT o.objId, o.ra,o.dec, o.type,
t.objId, t.j_m, o.z
FROM SDSS:PhotoPrimary o,
TWOMASS:PhotoPrimary t
WHERE XMATCH(o,t)<2.5
AND Region('Circle J2000 16.031 -0.891 .10')
AND( o.z- t.j_m)>1
5) Choose in the Result pull down the VOTable type (located bottom right of query box).
6) After Submitting the query the Query Status will go through the
Cost Estimates and show completion. Click on View to see the VOTAble output. Click on Save results to save the VOTable to a local directory. These are candidate brown dwarf sources.
7) To View the results in tabular format (at little easier to read), change the Result pull down type back to HTML and rerun the query.
The OpenSkyQuery portal can be programmatically used via web service to search various sky areas and therefore scan larger amounts of sky. An example of the web service client software can be found in the course software development directory under skyportalclient directory.
In the next part of this exercise we will use the Mirage Analysis tool to visualize the source found.
PART II: Brown Dwarf candidate Analysis and Visualization
Use Mirage to perform visual inspection of the source and multiparametric views of the VO data sets.
1) First you need to setup all the appropriate java paths by running:
C:\nvoss\bin>setupjava
2) Then start the Mirage java application:
3) From the pull down Console Menu, select new dataset from VO. Select the Cone Service and SIAP Services from SDSS DR2 as shown below. In order to see the brown dwarf, you must enter the coordinates AS SHOWN BELOW to actually find the source (i.e. remember they are 'cool' and thus faint). (NOTE: the SDSS datasets vary slightly between DR1 and DR2 so if you see any discrepancy between PartI & II of this exercise it may be related to minor calibration variances in selected datasets).C:\nvoss\bin>mirage
* Mirage Brown Dwarf Data Sets:
4) The brown dwarf source will appear in the Cone Search list, Click on the Number '1' at the top of the page to see all the analysis views. Slide the image onto the lower left panel(not on top of the coordinate panel) by dragging the image icon from the bottom of the icon list in the middle.
5) Do a file Open in the Image Panel and select the first SIA image.
6) Scale the image using the arcsinh. You can also do the same for the second image in the list. You then need to use the contrast sliders on the right of the image to adjust the scale to see objects.
7) If you then go to the coordinate object panel and select the broadcast symbol, zooming out of the image panels you can find the object appear (it actually shows up in the first image). You can then zoom in on that panel by clicking on the source to recenter and the +. You will find an extremely faint source that is the brown dwarf object.
8) Now we can do some analysis with the XML data you saved locally from OSQ. In the mirage Console menu, select New Dataset and then browse and select your saved VOTAble file. The Data Set will load and appear automatically in Mirage.
9) We will do a color magnitude plot by typing in the Command box at the bottom of the mirage console:
addcol z - j_m
This adds a column to the end of the table.
(10) If you click on the '1' tab at the top middle of mirage console, several graphics windows will appear. The scatter plot can be reset to plot on the y-axis the z-j_m column, and then on the x-axis the z column. You will now see a distribution that looks something like...
- Color Magnitude plot:
(11) Now you can try to select sources with colors about 3 and then do a 'broadcast' to the other panes. If you then load Dataset from VO Source and choose SDSSDR2-z SIA service, you can also look at the scatter plot, VOTable, and image to see the sources and their attributes in multiparametric views. ***NOTE again that the DR1 and DR2 coordinates vary slightly in the SDSS data sets.