1. EPoS: The Earliest Phases of Star formation Mapping and analysing extended emission and point sources EPoS: The Earliest Phases of Star formation Mapping and analysing extended emission and point sources Markus Nielbock (MPIA, PACS ICC, Heidelberg) on behalf of the EPoS group: O. Krause (PI), Z. Balog, H. Beuther, T. Henning, J. Kainulainen, R. Launhardt, H. Linz, N. Lippok, J. Pitann, S. Ragan, M. Schmalzl, A. Schmiedeke, J. Steinacker, A. Stutz, J. Tackenberg, T. Vasyunina Markus Nielbock (MPIA, PACS ICC, Heidelberg) on behalf of the EPoS group: O. Krause (PI), Z. Balog, H. Beuther, T. Henning, J. Kainulainen, R. Launhardt, H. Linz, N. Lippok, J. Pitann, S. Ragan, M. Schmalzl, A. Schmiedeke, J. Steinacker, A. Stutz, J. Tackenberg, T. Vasyunina

2. EPoS: The Earliest Phases of Star formation Herschel guaranteed time key programme (112 hours, PI: O. Krause, MPIA)Herschel guaranteed time key programme (112 hours, PI: O. Krause, MPIA) to map well studied cloud cores across the entire mass range (no blind survey)to map well studied cloud cores across the entire mass range (no blind survey) separated into a low (12 cores) and a high-mass (45 regions) partseparated into a low (12 cores) and a high-mass (45 regions) part to determine the dust temperature and density distribution of 12 near and isolated low-mass cores (e.g. Stutz et al. 2010, Nielbock et al. 2012, Launhardt et al. 2013)to determine the dust temperature and density distribution of 12 near and isolated low-mass cores (e.g. Stutz et al. 2010, Nielbock et al. 2012, Launhardt et al. 2013) to characterise the embedded core populations of 45 high-mass SF regions and IRDCs (e.g. Beuther et al. 2010, 2012, Henning et al. 2010, Linz et al. 2010, Ragan et al. 2013, Pitann et al. 2013)to characterise the embedded core populations of 45 high-mass SF regions and IRDCs (e.g. Beuther et al. 2010, 2012, Henning et al. 2010, Linz et al. 2010, Ragan et al. 2013, Pitann et al. 2013) used PACS and SPIRE bolometers at 70, 100, 160, 250, 350, and 500 µmused PACS and SPIRE bolometers at 70, 100, 160, 250, 350, and 500 µm small (< 10’) and deep (6 and 30 reps per map) to be sensitive for extended emission, PACS and SPIRE in prime observing mode (opposite to Gould Belt, HiGal surveys)small (< 10’) and deep (6 and 30 reps per map) to be sensitive for extended emission, PACS and SPIRE in prime observing mode (opposite to Gould Belt, HiGal surveys) added ground-based (sub)mm and NIR extinction dataadded ground-based (sub)mm and NIR extinction datahttp://www.mpia.de/IRSPACE/herschel/epos

3. Observational setup no parallel modeno parallel mode spatial distortion of point sources (scanning speed, on-board averaging)spatial distortion of point sources (scanning speed, on-board averaging) digitisation noise of faint extended emission (bit rounding in parallel mode)digitisation noise of faint extended emission (bit rounding in parallel mode) small but deep mapssmall but deep maps just large enough to assess background emission (isolated low-mass cores)just large enough to assess background emission (isolated low-mass cores) small map pixel size at 70 µm (high spatial coverage/sampling)small map pixel size at 70 µm (high spatial coverage/sampling) Gould Belt Survey 100 µm 160 µm EPoS 100 µm 160 µm B 68IRDC 316.72 EPoS 70 µm HiGal 70 µm EPoS – HiGal EPoS / HiGal

4. Data reduction identical for both EPoS branchesidentical for both EPoS branches standard L1 processing (incl. 2 nd level/map deglitching)standard L1 processing (incl. 2 nd level/map deglitching) branched out to Scanamorphos/IDL for processing to L2 (settings: galactic, noglitch, nzdata=‘yes’)branched out to Scanamorphos/IDL for processing to L2 (settings: galactic, noglitch, nzdata=‘yes’) highpass filter + photProject (PACS) unsuitable for extended emissionhighpass filter + photProject (PACS) unsuitable for extended emission Mad Map had issues with point sources; pre-processing of signal drifts not deterministic (much better now)Mad Map had issues with point sources; pre-processing of signal drifts not deterministic (much better now) At that time: Scanamorphos best compromise between recovering faint extended emission and, at the same time, resolving point sources At that time: Scanamorphos best compromise between recovering faint extended emission and, at the same time, resolving point sources Scanamorphos was applied to all PACS and SPIRE data to maintain consistency, but the decision was mainly driven by the PACS data quality.Scanamorphos was applied to all PACS and SPIRE data to maintain consistency, but the decision was mainly driven by the PACS data quality. alternative mappers too late; changing would break consistency with already published resultsalternative mappers too late; changing would break consistency with already published results

5. Data reduction PACS 100 µm PACS 160 µm HPF (100), photProjectScanamorphospresent-day Mad Map B 68

6. Data reduction PACS 70 µm PACS 160 µm Scanamorphospresent-day Mad Map IRDC 316.72 HPF (100), photProject

7. Data analysis – extended emission Final goal: to derive dust temperature and particle density mapsFinal goal: to derive dust temperature and particle density maps convolved data to same resolution; regridded maps to same pixel size using IDL (Aniano et al. 2011, circularised convolution kernels)convolved data to same resolution; regridded maps to same pixel size using IDL (Aniano et al. 2011, circularised convolution kernels) background subtraction using identical area in each mapbackground subtraction using identical area in each map ideally: to model background (zodiacal, cirrus, cosmic), but turned out to be negligibleideally: to model background (zodiacal, cirrus, cosmic), but turned out to be negligible Nielbock et al. (2012)

8. Data analysis – extended emission spatially resolved SED fitting (Launhardt et al. 2013)spatially resolved SED fitting (Launhardt et al. 2013) only correct for single T component along LoS (averaging effect)only correct for single T component along LoS (averaging effect) employed ray-tracing algorithm with functional relationship for radial density distribution using GILDAS (Plummer-like profile, cf. Whitworth & Ward-Thompson 2001)employed ray-tracing algorithm with functional relationship for radial density distribution using GILDAS (Plummer-like profile, cf. Whitworth & Ward-Thompson 2001) RT fit yields central T dust = 8 K, smaller by 2 K than with LoS averaged SED fittingRT fit yields central T dust = 8 K, smaller by 2 K than with LoS averaged SED fitting Nielbock et al. (2012)

9. Data analysis – PSF photometry target regions located in complex environments with high dynamic rangetarget regions located in complex environments with high dynamic range compact emission sourcescompact emission sources  on top of structured extended emission  embedded within extinction structures aperture photometry unreliable  PSF photometry (Starfinder / IDL; easy to use, other tools not available/reliable/validated at that time)aperture photometry unreliable  PSF photometry (Starfinder / IDL; easy to use, other tools not available/reliable/validated at that time) PSF templates reproduced from Vesta calibration observations (not circularised)PSF templates reproduced from Vesta calibration observations (not circularised) PSF template rotated to match position angle of observation (rotation angle = position angle as given in metadata, counter clockwise)PSF template rotated to match position angle of observation (rotation angle = position angle as given in metadata, counter clockwise) Iterative approach:Iterative approach: 1)run starfinder on unsharp-masked version of science data to detect point sources 2)detection list as input for PSF fitting and photometry on original science data

10. Data analysis – PSF photometry original PACS 70 µm maporiginal PACS 70 µm map IRDC with embedded point sourcesIRDC with embedded point sources H II region to the SEH II region to the SE LBV shell to the NWLBV shell to the NW IRDC H II (DR 15) LBV point sources detected with IDL starfinderpoint sources detected with IDL starfinder false detections in LBV shell and H II regionfalse detections in LBV shell and H II region all visible point sources in the IRDC detectedall visible point sources in the IRDC detected altogether 496 confirmed point sources in 45 IRDCsaltogether 496 confirmed point sources in 45 IRDCs (Ragan et al. 2012) (Ragan et al. 2012) background image with point sources removedbackground image with point sources removed G79.31+0.36 (Cygnus Rift)

11. Data analysis – PSF photometry One goal: characterise embedded core populationOne goal: characterise embedded core population Ragan et al. (2012) core sizes: 0.05 – 0.3 pccore sizes: 0.05 – 0.3 pc core luminosities: 0.1 – 10 4 L core luminosities: 0.1 – 10 4 L  core masses: 0.1 – a few 10 3 M core masses: 0.1 – a few 10 3 M  core temperatures: 13 – 30 Kcore temperatures: 13 – 30 K

12. Summary EPoS designed to investigate galactic low/high mass star forming sitesEPoS designed to investigate galactic low/high mass star forming sites rather high S/N and structure/point source detection instead of large fields  PACS/SPIRE individually instead of parallel moderather high S/N and structure/point source detection instead of large fields  PACS/SPIRE individually instead of parallel mode data reduction aimed at doing equally well for extended emission and point sources  standard L1 processing in HIPE + Scanamorphos in IDLdata reduction aimed at doing equally well for extended emission and point sources  standard L1 processing in HIPE + Scanamorphos in IDL resulted in high quality maps suitable for fitting dust properties  resolved dust T and n distributions of isolated low-mass coresresulted in high quality maps suitable for fitting dust properties  resolved dust T and n distributions of isolated low-mass cores resulted in high point source detection rate (469) in 45 complex high-mass SFR IRDCs  catalogue and characterisation of cores embedded in IRDCsresulted in high point source detection rate (469) in 45 complex high-mass SFR IRDCs  catalogue and characterisation of cores embedded in IRDCs Generally, Herschel mapping data are of very high quality for achieving EPoS goals.Generally, Herschel mapping data are of very high quality for achieving EPoS goals. Possible improvements worth reprocessing: better noise handling (faint extended emission, detection of even fainter sources)better noise handling (faint extended emission, detection of even fainter sources) corrections for pointing jitter and improved PACS FPG characterisation (smaller FWHM, consolidation of flux calibration, revise source associations)corrections for pointing jitter and improved PACS FPG characterisation (smaller FWHM, consolidation of flux calibration, revise source associations)