Basic principles of PACS photometer data reduc7on

Transcription

1 Basic principles of PACS photometer data reduc7on B. Altieri, HSC/ESAC November 2012 B. Altieri HSC/ESAC November 2012 Slide 1

2 PACS photometer Two filled bolometer arrays in dual band imaging: 64x32 pixels at 70 & 100µm 32x16 pixels at 160µm Field-of-view: 1.7 x3.5 Build-up of maps by scanning in a boustrophedon fashion. PSF FWHM: 5.5, 6.7 and Hz readout frequency on-board co-adding, to stay within allocated 130kb/s rate downloaded frequency : 10Hz (prime) and 5 Hz in parallel in the blue channel) B. Altieri HSC/ESAC November 2012 Slide 2

3 PACS photometer pipeline (SPG) q HSA Levels : q q Level0 : raw data cubes Level 1 : data cubes calibrated in Jy/pixel, with associated pointing and masks now properly deglitched end safe point for automated pipeline q q Level 2 : maps created by projecting high-pass filtered data cube on a grid using pointing information Level 2.5 (extended pipeline): combining pairs of obsid s q q MADmap (extended pipeline) starting from L1 Mosaic of L2 maps. q Other mappers (MADMap, Scanamorphos, ) can start from Level 1. B. Altieri HSC/ESAC November 2012 Slide 3

4 PACS photometer interactive scripts yes Is it faint? (< few 100 mjy) yes no Deep Survey maps Bright pointsource SPG L2.5 HPPPMOS multiple_obsid_scanmap_deep_survey_miniscan_pointsource.py Source is point-like? no yes multiple_obsid_scanmap_pointsource.py Extended source (HPF+photProject) (with caution) Is it extended? (few x FWHM) multiple_obsid_scanmap_extended_emission_photproject.py no Extended source (MADmap) SPG L2.5 HPPPCOM L25_scanMapMadMap.py B. Altieri HSC/ESAC November 2012 Slide 4

5 Documentation Detailed step-by-step description of the ipipe scripts on : PACS Data Reduction guide NHSC tutorials: The effect of the high-pass filter data reduction technique on the PACS Photometer PSF, point-source photometry and noise has been investigated in depth in this technical note B. Altieri HSC/ESAC November 2012 Slide 5

6 Some convenience scripts ü Integrated script for a full masked high-pass filtering, optimized to not restart from level 0 in the second pass, posted on : ftp://ftp.sciops.esa.int/pub/baltieri/pacs/webinarnov2012/miniscanmap_pointsources.py ü Extract a submap from a large field, with less memory ftp://ftp.sciops.esa.int/pub/baltieri/pacs/webinarnov2012/subareafroml1.py ü To be optimized (still loading the full L1 cube) and made available as a plug-in asap. B. Altieri HSC/ESAC November 2012 Slide 6

7 mapping possibilities inside HIPE photproject combined with a masked highpass filtering of the timelines is intended for points sources only or slightly extended source up to the size of the masked area. MADmap (HCSS) preserves extended emission at any scale. It works now very well in latest versions of Hipe: baseline drift corrections improved point-source artifacts on bright sources can now be corrected See tutorial on NHSC page, and/or PDRG (PACS Data Reduction Guide) Some MADmap maps in HSA (SPG8.3) still affected by background artifacts, but large field maps on galactic fields are spectacular and can be used for science Point-source photometry consistent between HPF+photProject and MADmap but MADmap sensitivity for faint point sources typically worse by 40%. B. Altieri HSC/ESAC November 2012 Slide 7

8 Somewhere in the galactic plane with MADmap B. Altieri HSC/ESAC November 2012 Slide 8

9 Somewhere on the galactic plane with MADmap HiGAL OT KP (S. Molinari) 160um 2x2 degrees B. Altieri HSC/ESAC November 2012 Slide 9

10 MADmap versus photproject on deep fields HSA SPG8.3 L2.5 MADmap B. Altieri HSC/ESAC November 2012 Slide 10 HSA SPG8.3 L2.5 photproject

11 mapping possibilities outside HIPE Scanamorphos is widely used by various groups/projects Preserves extended emission just like MADmap works on IDL, easy to use. the export of the level 1 frames done with the task convertl1toscanam very few parameters to tune Documented and carefully maintained by Hélène Roussel at: Photometry of point-like and extended sources is highly consistent between madmap. B. Altieri HSC/ESAC November 2012 Slide 11

12 Other map-makers Public map-makers but not supported by HSC: ü ü Unimap: Generalised Least Square (GLS) approach to obtain the final map, similar to MADmap, used by HiGAL SANEpic: Maximum likelihood map-making ü TAMASIS (v4.2): + Pierre Chanial (CEA) ü JPScanam: Scanamorphos porting into Java (Hipe), under development, by Michael Wetzstein (MPE) B. Altieri HSC/ESAC November 2012 Slide 12

13 Some considerations PACS photometer maps from any map-maker are defined to an arbitrary offset level (telescope background) Background=0 for highpass filter (HPF) processing Error maps: Still not reliable in HPF+photProject processing (correlated noise from 1/f noise + projection on different grid) Not available for MADmap, possibly in the future. B. Altieri HSC/ESAC November 2012 Slide 13

14 New error map based on coverage in SPG10 SPG 8.3 SPG 10.x B. Altieri HSC/ESAC November 2012 Slide 14 New task: photcoverage2noise

15 Herschel map-making workshop B. Altieri HSC/ESAC November 2012 Slide 15

16 Source detection on PACS maps ü Daophot and Sussextractor implemented in Hipe ü Sussextractor requires an error map ü PACS deep survey use so far - mostly: StarFinder (IDL): PSF fitting Sextractor (contiguous pixels above noise threshold) ü Galactic fields: another story getsources: multi-scale, multi-wavelength source extraction algorithm (public) bolosource (reference?) NHSC recommendations: B. Altieri HSC/ESAC November 2012 Slide 16

17 B. Altieri HSC/ESAC November 2012 Slide 17 DEEP PACS imaging

18 Convolving PACS & SPIRE maps ü Plug-in available but for Hipe 10 track only. ü Make use of the Aniano circularized kernels ü Jython scripts also available, allows to load and rotate other kernels. ftp://ftp.sciops.esa.int/pub/baltieri/pacs/webinarnov2012/ B. Altieri HSC/ESAC November 2012 Slide 18

19 B. Altieri HSC/ESAC November 2012 Slide 19