SCIENCEe
Luca ConversiHerschel Science Centre
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Luca Conversion behalf of the SPIRE ICC
Extracting PhotometryFrom SPIRE Maps
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/20122
• Map Calibration and Units– How Calibration is Applied– Units and Beam Areas
• Bright Point Source–Which technique to use?– Estimating uncertainties– Peak fitting demo
• Source Extraction– Inside HIPE: DAOPhot & Sussextractor– Correction factors
• Extended Emission– Spectral response for extended sources– Caveats
Presentation Outline
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Calibration And Units - 1
• SPIRE Calibration is defined in a straightforward way
• When a detector is scanned directly over a point source, the peak deflection of the signal timeline equals the brightness of the source.
Scan of detector PSWE8 over
Neptune,obsid 1342187440
3
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Calibration And Units - 2
The timeline-based definition leads to some accounting corrections
• Units are in Jy/beam• Peak value is nominally the brightness
• Calibration applies to timelines• Map-making lowers the peaks
• Calibration is for ν*Fν = constant• Color correction for other spectral shapes
• Calibration is for point sources• Correction for extended sources
4
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
• Pixelization corrections are described in the SPIRE Observer’s Manual: see Sec 5.2.11
• For nominal pixel sizes,divide by factors of:
– 250µm: 0.941
– 350µm: 0.917
– 500µm: 0.903
250 350
500
Pixelization Correction
5
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
• Colour corrections are described in the SPIRE Observer’s Manual: see Sec 5.2.8
• Multiply by the value appropriate for your source
• Different corrections for point and extended sources
• For specific cases (e.g., grey body), compute your own colour correction.Filters RSRF are available in the calibration tree:– rsrf = obs.calibration.phot.rsrf
6
Colour Correction
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Unit Conversion
• Use the fine-scale beam areas to convert Jy/beam to other units
• The fine-scale beam areas were measured from Neptune maps with 1-arcsec pixels
– 250µm: 423 arcsec2
– 350µm: 751 arcsec2
– 500µm: 1587 arcsec2
• However, these values may be updated in the upcoming weeks. Have a look at the SPIRE Twiki http://herschel.esac.esa.int/twiki/bin/view/Public/SpireCalibrationWeb?template=viewprint
• The convertImageUnit task will change maps to Jy/pixel or MJy/sr
7
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/20128
• Map Calibration and Units– How Calibration is Applied– Units and Beam Areas
• Bright Point Source–Which technique to use?– Estimating uncertainties– Peak fitting demo
• Source Extraction– Inside HIPE: DAOPhot & Sussextractor– Correction factors
• Extended Emission– Spectral response for extended sources– Caveats
Presentation Outline
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Gaussian Fitting
Peak-fitting is recommended for point sources
• Gaussian-fitting on maps gives good results for bright sources– Gaussian is good approximation for radii up to (22, 30, 42) arcsec for (PSW, PMW, PLW)
respectively
• Fitting Gaussians to timelines is under development by ICC– Better results for 100mJy+ sources– Prototype available as script on the HSC public TWiki:
http://herschel.esac.esa.int/twiki/pub/Public/SpireCalibrationWeb/bendoSourceFit_v0_9.py.txt
• Aperture photometry is possible– Official corrections not yet available– Derive your own
• Beam profiles are available in the calibration tree:– beamPSW = obs.calibration.phot.getBeamProf('PSW')– beamPMW = obs.calibration.phot.getBeamProf('PMW')– beamPLW = obs.calibration.phot.getBeamProf('PLW')
9
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Flux Uncertainties
Combine these in quadrature to estimate the uncertainty
• Uncertainty in the fitted amplitude– Includes the instrument and confusion noise
(minimum of ~5 mJy)
• 2%-3% of flux density for uncertainty in pixelization correction
• 7% of flux density for calibration uncertainty– 2% relative (band-to-band) uncertainty– 5% is Neptune model (does not apply to SPIRE colours)
10
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/201211
• Map Calibration and Units– How Calibration is Applied– Units and Beam Areas
• Bright Point Source–Which technique to use?– Estimating uncertainties– Peak fitting demo
• Source Extraction– Inside HIPE: DAOPhot & Sussextractor– Correction factors
• Extended Emission– Spectral response for extended sources– Caveats
Presentation Outline
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Source Extraction in HIPETwo source extractors are available within HIPE:
1. sourceExtractorSussextractor implements the SUSSEXtractor algorithm, as described by Savage & Oliver (2007), ApJ, 661, 1339. The image is smoothed with a convolution kernel, derived from the point response function, and the resulting smoothed image is searched for peaks, which are taken to be the positions of the point sources. The intensity in the smoothed image at the position of a point source is taken as the estimate of that source's flux density. Note that "NaN" pixels near point sources in the image or error extensions of the input image may prevent sourceExtractorSussextractor from finding that source
2. sourceExtractorDaophot implements the DAOPHOT (classic) algorithm, following the "FIND" and "APER" procedures in the IDL Astronomy User's Library. The image is smoothed with the DAOPHOT convolution kernel to find the source positions, as in the sourceExtractorSussextractor task. Then the source flux densities are estimated using aperture photometry, as in the aperturePhotometry task
• Both runnable as Tasks
• Both can accept input source lists
• You must compute corrections yourself
12
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Correction Factors
The simplest operation of extractors uses the FWHM of each band
• Averages for nominal pixels:– 250µm: 17.5 arcsec– 350µm: 23.9 arcsec– 500µm: 35.2 arcsec
• You can supply your own PRF image
• The fine-scale beam areas are needed (or the images must be converted)– (423, 751, 1587) arcsec2
Corrections are still needed for extractor outputs
• Need an aperture correction appropriate to FWHM– About 25% - 30%
• Still need color correction
13
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Presentation Outline
14
• Map Calibration and Units– How Calibration is Applied– Units and Beam Areas
• Bright Point Source–Which technique to use?– Estimating uncertainties– Peak fitting demo
• Source Extraction– Inside HIPE: DAOPhot & Sussextractor– Correction factors
• Extended Emission– Spectral response for extended sources– Caveats
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
• The extended spectral response differs from that for point sources
• Coupling for extended source that uniformly fills the beam (Observer’s Manual)
15
Extended Emission RSRF
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Extended Emission Caveats
• No absolute level– Could get from Planck
• Map-‐making technique can affect extended emission– Baseline removal vs destiper: note different flux scale on the images
• CIB fluctuations at a faint level
16
Median Baseline Removal Destriper
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Ext. Emission Photometry
• Units correction and extended correction must be applied
• Correct units e.g. to Jy/pixel
– convertImageUnit task
– Fine beam areas
• Multiply by these factors for extended relative spectral response function
– 250µm 0.9939/1.0113
– 350µm 0.9898/1.0060
– 500µm 0.9773/1.0065
17
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
To summarize...
18
Point Source Photometry
Do you know where your sources are?
Yes, but they can have any flux
Yes, and they are brighter than ~20mJy in the SPIRE bands
No
Annular Sky Aperture
Timeline Fitter
Sussex or DAOPhotSource Extractor
Extended Source Photometry
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Sussex Source Extractor
19
Sussex Source Extractor
Input: FWHM, beam area,threshold
Output: source catalogue Pixelization correction
Timeline Fitter Annular Sky Aperture
Loop on [RA, Dec] Colour correction[optional]
DONE!
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
DAOPhot Source Extractor
20
DAOPhot Source Extractor
Input: FWHM, beam area,threshold, aperture
Output: source catalogue Pixelization correction
Timeline Fitter Annular Sky Aperture
Loop on [RA, Dec]
Aperture correction
DONE!
Colour correction[optional]
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Timeline Fitter
21
Timeline Fitter
Output: fitted [RA, Dec],flux, error, χ2
Colour Correction[optional]
DONE!
Input: source [RA, Dec]
Single point sourceor catalogue
Note: if you have an external source catalogue (i.e., not derived on SPIRE), it is better the first run a source extractor.Otherwise, it is likely that the timeline fitter will give wrong results for the faintest (below ~20mJy) sources
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Annular Sky Aperture
22
Colour correction[optional for point sources;
mandatory for extended emission]
Annular Sky Aperture
convertImageUnit taskInput: beam area
Convert map in Jy/pixel
Output: flux, backgroundestimation, error, etc.
Point or extended sourceor catalogue
Input: source [RA, Dec], annuli radii
Pixelization correction
Aperture correction DONE!
Wednesday, 22 February 12
Luca ConversiHerschel Science Centre
SCIENCEe
OT2 DP WorkshopESAC, Madrid, 20-24/02/2012
Demo Time!
• Run the script OT2_Workshop_SPHOT_Photometry.py
• Target: Gamma Dra
– Model fluxes: (254, 130, 64) mJy
• Methods used to extract photometry
– Timeline fitter (bendoSourceFit) on a specific source (Gamma Dra)
– SussexExtractor on the PSW map + timeline fitter on the extracted sources
– Aperture photometry of Gamma Dra on PSW map
23
Wednesday, 22 February 12