Finding Extragalactic Point Sources in WMAP Temperature Maps: A CMBindependent Technique

1
Finding Extragalactic Point Sources in WMAP
Temperature MapsA CMB-independent Technique

• Xi Chen (UCLA)
• CMB Component Separation and the Physics of
  Foregrounds,
• Pasadena, July 14 - 18, 2008

2
Outline

• Motivations
• Why CMB-free, why high frequency bands
• Point source search in WMAP V (61 GHz) and W (94
  GHz) bands maps
• Point source search in WMAP Q (41 GHz), V and W
  bands maps
• Flat spectrum source favored
• Conclusions

3
Motivations

• Why we want a CMB-independent technique?
• Avoid confusion of positive CMB excursions with
  point sources
• Improve detection rate with increasing observing
  time
• WMAP technique (Bennett et al. 2003, Hinshaw
  et al. 2007, Wright et al. 2008)
• filter Nobs1/2T by
• Why we apply this method to the high frequency
  bands WMAP maps?
• Only 46.7 and 9.3 of the WMAP three-year
  cataloged sources are 5? detections at V and W
  band respectively.
• Refine the power spectrum model for unresolved
  point sources

4
Point Source Search in V W bands

• Smooth W-band map to match V-band resolution
  (?Rc1º)

5

• Subtract the smoothed W-band map from the V-band
  map

V
WSM
Fig.3. from Bennett et al, 2003
6

• Filter the residual map to get true point source
  flux distribution

Minimizing ?2 gives,
7

• Source identification (99 sources in total)
• Cross-correlate our source list with the WMAP
  5-yr catalog
• (64 are present in the catalog)
• Estimate V-band flux densities of our sources
  (17 more are identified using their SEDs)

with
8

• 81 of our sources have low frequency
  counterparts
• Quasars (58), Galaxies (14), HII regions (9)
• 17 can be associated with extended radio
  sources
• NGC 1499 (2), Ori Nebula (3), Gum Nebula (8),
  Oph Complex (3), Cen A vicinity (1)
• 1 remains uncertain
• 1? position uncertainty 1.4
• A 0.012 0.004 ?K2 sr (Chen Wright, 2008)
• A 0.014 0.003 ?K2 sr (Hinshaw et al.,
  2007)
• A 0.012 0.005 ?K2 sr for l lt 500
  (Huffenberger et al., 2007)
• A 0.015 0.005 ?K2 sr for l gt 500

9
(No Transcript)
10
Strengths and limitations

• No CMB noise gt low detection threshold and
  detection rate increases rapidly with integration
  time
• N tint0.72 from 1 year to 5 years, compared to
  N tint0.39 using the WMAP team method
• For Euclidean source counts the expected scaling
  is t0.75
• Use maps with high resolution and low foreground
  contamination gt better positional accuracy

• Subtraction of maps may cause missing sources.
• for flat spectrum sources, TW 1/2 TV
• V W maps have relatively lower SNR for flat
  spectrum sources.

11
Point Source Search in Q, V W bands

• Smooth V- and W-band maps to match Q-band
  resolution (?Rc1.25º)

• Construct a linear combination map that is
• CMB free,
• optimized for flat-spectrum source,
  
• variance minimized,

12

• Use the same Kp0LMCSMC mask
• as in Wright et al., 2008

13
(No Transcript)
14

• 381 sources are detected in the QVW combination
  map
• 291 WMAP cataloged point sources, 75 newly
  detected sources
• QSO (269), G (61), GPair (1), GTrpl (1), PN (1),
  RadioS (33)
• 15 remain uncertain (could be contaminations
  from extended radio emissions)
• 1? position uncertainty 2.0

15
Conclusions

• We established a CMB-independent method for point
  source search in CMB data.
• This technique is proved to be reliable,
• 99 detections in V W bands maps (99
  identified) 381 detections in Q, V W bands
  maps (94 identified)
• consistent,
• Sources detected in the first year maps are
  repeatedly detected in three-year and five-year
  co-added maps.
• and complementary to the WMAP technique
• new detections, but not complete
• No sign of a significant different class of new
  objects
• Contamination from unresolved point sources at
  WMAP V-band is negligible, Cv (2.4?0.8)10-3
  ?K2 sr