Color Selection of Quasars

1
Color Selection of Quasars

• Jonathan Jerke
• Yale meeting Dec 4 2007

2
Outline of talk

• Quality of B,R,and I filters
• GALEX database
• Discuss variety of quasar selection cuts made
  with GALEX and QUEST
• Present lensing candidates drawn from quasar
  sample
• Discuss criterion for followup of lensing
  candidates

3
Area of Quality analysis

• SDSS stars only
• Single measurements
• In area between 120 and 240 degrees RA and -4 to
  20 degrees DEC.

4
Quality of Filters in selections

• Considering individual measurements of SDSS stars
• For all candidates, Blt20.
• B-R 0.5 for quasars with sigma of 0.2
• R-I -0.2 for quasars with sigma of 0.2
• Inclusive measurements Blt20, Rlt19.5,and Ilt19.7

5
Statistical quantities considered

• SDSS-QUEST
• SIGMA
• CALIBRATION_ERROR
• (SDSS-QUEST)/(CALIBRATION_ERROR2
  QUEST_ERROR2SDSS_ERROR2) has unit sigma
  distribution
• A general calibration quality added in quadrature
  to the pipeline error to estimate true error

6
Quality of filters

• Blt20Rlt19.5Ilt19.7
• R SIGMA 10
• R Calibration Error 7
• I SIGMA 15
• I Calibration Error 9
• B SIGMA 16
• B Calibration Error 10

7
U filter quality

• The U filter has an ERROR of about 30
• Impossible to properly separate quasars by
  UV-Excess using U
• Need to use another UV catalog

8
GALEX database

• Two filters
• FUV between 1350A and 1750A
• NUV between 1750A and 2800A
• Using All Sky Survey
• NUV 21 at 30 errors
• FUV 20 at 30 errors

9
GALEX coverage

• GR1 used for Nov 06 candidates
• 2200 square degrees
• GR3 used for Jun 07 candidates
• 9000 square degrees
• Most Northern Galactic Hemisphere data in SDSS
  coverage

10
Color selections

• Quasar selections made for QUEST-GALEX overlap
• Test of selections made by considering
  QUEST-GALEX measurements of SDSS objects
• Covers all errors in QUEST measuring
• Divide relevant color space into a fine grid
• Place a unit normalized gaussian with width of
  pipeline error per measurement
• Result is a continous distribution of a species
  in color space
• Calculate theoretical quasar probability per
  color
• Make cut at selected theoretical purity
• Defines a region in the color space

11
NUV-BR plot in SDSS overlap
12
Most Robust Selection

• Using only NUV, R, and I one should have the
  greatest precision in color measurement
• However, only 20 purity at best
• Because B-R separates stars and galaxies from
  quasars by 1 sigma each

13
NUV-RI purity verses Quasars per square degree
14
Why we need B

• The center of the Quasar B-R is one sigma from
  both the Galaxy and Star B-R distributions

15
NUV-B selection strong
16
NUV-BR
17
November 2006 selection

• Using NUV-UBR
• Supposed to be the most pure selection using
  GALEX release 1
• U filter reduces galaxy signal
• 2537 objects selected

18
Statistics on Quality of Nov 06 selections
19
NUV-BRI

• Implimented for GALEX release 3
• I band helps alittle
• Goes deeper in B without U
• Higher completeness because U measurement not
  required

20
Statistics on June 2007 all sky quasar catalog 1
21
Statistics on June 2007 all sky quasar catalog 2
22
Quality of NUV-BRI dependent on number of
observations

• Required at least 2 measurements per B,R,and I
• Data set 2 from the FALL catalog had about twice
  the completeness
• The average number of combined measurements is
  about 5
• Commonly 2 measurements in other parts of the sky
• On a equitorial region

23
NUV-BRI one measurement statistics
24
FUV-NUV-BRI

• Adding FUV reduces completeness by a factor of 4
• Increases purity to 80 for B lt 20
• A subset of NUV-BRI

25
Statistics on FUV-NUV-BRI
26
Summary of Selections
27
Feed to Swiss

• Nov,2006. all NUV-BR selected objects
• 2500 sent
• June,2007. list of NUV-BRI created
• Sent brightest and NED(Xray,Radio,Qso)
• About 1000 objects sent
• December,2007 all NUV-BRI candidates
• 25000 objects

28
Swiss processing

• Using single filter (R or I) images
• Examine image for extended structure

29
Swiss lensing candidates

• Nov 06
• 52 candidates
• Aug 07
• 171 candidates
• Sep 07
• 86 candidates
• (there are a few double entries which I removed)

30
Lensing candidates compiled

• R
• Priority
• SDSS (S/G)
• NED(Q/G/X/R)
• 2Mass(Q/G)
• Variability(0,1,2)
• FUV selection
• Imaging
• Spectroscopy
• Status

31
(No Transcript)
32
Discussion of Catagories

• Lens
• Not enough information
• Not lens
• Follow up image

33
What do we look at next?

• Consider past experience
• Consider our flags

34
Two spectra taken last year

• L2.8
• PQL037
• R17
• Clear separation

• L3.14
• PQL017
• R19
• Separation seen at limit

35
17th in R data
36
19th in R data
37
Expected 200in performance

• Cannot take a spectra below 19th magnitude
• Suggests that we should not go below a primary
  magnitude of 18
• Separations greater than 2

38
Short List

• Rlt18
• Not SDSS star
• Not NED Galaxy
• Not previously identified
• Total of 84 across sky
• 33 are observable in January
• Quest_src/Jan08.observing.candidates.xls

39
Conclusion

• Quality of QUEST B,R,and I
• Variety of selections, mostly concerned with the
  NUV-BRI now
• We have about 33 hot candidates for imaging in
  January