F. Maccagni; R. Morganti; T. Oosterloo; K. Geréb; N. Maddox, J. Allison THE LAST SURVEY OF THE OLD WSRT: TOOLS AND RESULTS FOR THE FUTURE HI ABSORPTION SURVEYS
A SURVEY BEFORE THE BLIND SURVEYS During my PhD, WSRT was upgraded to Apertif: over time antennae went offline To detect HI absorption we don t need complete uv-coverage Great opportunity for HI absorption studies Observe as many sources as possible before WSRT observations stop 4/6 hrs observation with variable number of antennae average noise in the spectra ~ 1 mjy Set strategy and tools for the blind surveys: SHARP, MALS, FLASH What is the detection rate of HI in the local Universe? What features of the HI lines relate to the properties of the radio sources?
THE LAST SURVEY OF THE OLD WESTERBORK 248 sources.2 < z <.25 SDSS spectroscopy S Cont 3 mjy mostly AGN in ETG log 1 P1.4GHz [W Hz 1 ] 26 25 24 23 22 21 Jarrett et al. 213 Dust-poor source 12µm bright source 4.6µm bright source Interacting source A3D Dust-poor source A3D 12µm source A3D 4.6µm source This SAMPLE 2 1. 11 sources; S Cont 5 mjy Stacking experiment [Geréb et al., 214] 19 18 ATLAS 3D 39 4 41 42 43 44 45 46 log 1 L 22µm [erg s 1 ] Analysis of the detections [Geréb, Maccagni, et al., 215] 2. All 248 sources Mid-InfraRed [22 µm] - Radio Power Relation This Talk [Maccagni et al., 217]
STRATEGY OF THE SURVEY Similar to APERTIF observations Flagging and automatic calibration Data Cube + Continuum Image Continuum source finder: location continuum sources (down to 1mJy/ 5mJy or even lower) Extract the spectra @ location continuum sources Identify HI absorption detections: beam Bayesian line-finder [Allison et al 212] 4 2 Flux [mjy] 5 Flux [mjy] Noise [mjy] 1 Detection 1.6 1.4 1.2 1..8 1.326 1.328 1.33 1.332 1.334 1.336 1.338 1.34 Frequency [ 1 9 Hz] Noise [mjy] 2 Non Detection 4 1.8 1.6 1.4 1.2 1..8 1.36 1.38 1.31 1.312 1.314 1.316 1.318 1.32 Frequency [ 1 9 Hz]
STRATEGY FOR HI ABSORPTION SURVEYS 4 2 Flux [mjy] 5 Flux [mjy] 1 2 Noise [mjy] Detection 1.6 1.4 1.2 1..8 1.326 1.328 1.33 1.332 1.334 1.336 1.338 1.34 Frequency [ 1 9 Hz] Noise [mjy] Non Detection 4 1.8 1.6 1.4 1.2 1..8 1.36 1.38 1.31 1.312 1.314 1.316 1.318 1.32 Frequency [ 1 9 Hz] Intervening or associated? cross-correlation with spectral surveys Characterisation absorption (width, centre, asymmetry etc. using e.g. busy function) cross-correlation with LOFAR fields & other archives (SDSS, WISE) DATABASE Properties of the HI lines Properties of radio continuum (extent, spectral index, )
HI ABSORPTION LINES 66 detections line features measured with the BusyFunction [Westmeier, et al. 214] 3 < FWHM < 57 km/s 7 < FW2 < 64 km/s 3 main groups: Narrow lines: FWHM < 1 km/s Medium width lines: 1 km/s <FWHM < 2 < km/s Broad lines: FWHM > 2 km/s
CHARACTERISATION OF THE SAMPLE Compact sources (red): unresolved by FIRST. often radio-jets on sub-galactic scales. many compact sources are young AGN. Extended sources (blue): resolved by FIRST. radio-jets on super-galactic scales. usually older AGN than compact sources. NVSS major / minor axis 2. 1.8 1.6 1.4 1.2 Extended non-detection Compact non-detection Extended detection Compact detection Interacting detection Interacting non-detection 1..2.3.4.5.6.7.8.9 1. FIRST peak / integrated flux
CHARACTERISATION OF THE SAMPLE WISE MIR colours dust in the host galaxy Dust-poor sources (green) 12 µm-bright sources (orange) Emission from PAHs and heated dust. 4.6 µm-bright sources (black) W1 W2 [3.4-4.6 µm] 1.5 1..5. Dust-poor non-detection 12µm bright non-detection 4.6µm bright non-detection Dust-poor detection 12µm bright detection 4.6µm bright detection Interacting non-detection Interacting detection The central AGN heats the surrounding circumnuclear dust.. 1. 2. 3. 4. 5. W2 W3 [4.6-12 µm]
DETECTING HI ABSORPTION 248 sources / 66 Detections 27 % ± 5.5 % detection rate Count 25 2 15 1 All sources Detections Constant in redshift and radio power Det. Rate [%] 5 8 6 4 2.2.5.8.11.14.17.2.23 z Compact sources and MIR bright HI often detected (~4%). Extended sources & dust-poor sources HI is rarely detected (~13%). Count Det. Rate [%] 35 3 25 2 15 1 5 8 6 4 2 23 24 25 26 log 1 P 1.4GHz [W Hz 1 ] All sources Detections
KINEMATICS OF THE HI 8 Interacting sources Extended sources Compact sources P 1.4GHz < 1 24 W Hz -1 widths rotational velocity HI likely in a rotating disk. P 1.4GHz > 1 24 W Hz -1 broad asymmetric lines. FW2 [km s 1 ] 6 4 2 22 23 24 25 26 log 1 P 1.4GHz [W Hz 1 ] Sources with broad lines are: Compact, i.e. jets within the galaxy. MIR bright, i.e. rich in heated dust. FW2 [km s 1 ] 8 6 4 2 Interacting sources Dust-poor sources 12µm bright sources 4.6µm bright sources 22 23 24 25 26 log 1 P 1.4GHz [W Hz 1 ]
KINEMATICS OF THE HI P 1.4GHz < 1 24 W Hz -1 lines centred at systemic velocity P 1.4GHz > 1 24 W Hz -1 lines offset w.r.t. systemic velocity offset is blue-shifted. vcentroid -vsystemic [km s 1 ] 6 4 2 2 4 6 Interacting sources Extended sources Compact sources 22 23 24 25 26 log 1 P 1.4GHz [W Hz 1 ] Broad, asymmetric, shifted absorption line Unsettled kinematics Powerful radio sources Compact, i.e. jets within the galaxy. MIR bright, i.e. rich in heated dust. vcentroid -vsystemic [km s 1 ] 6 4 2 2 4 Interacting sources Dust-poor sources 12µm bright sources 4.6µm bright sources 6 22 23 24 25 26 log 1 P 1.4GHz [W Hz 1 ]
STACKING EXPERIMENT Non-detections are important!!!!.3.2 Non Detections Stacking of 17 non-detections NO LINE is detected at ~.15 (3σ).1..1 Stacking of sub-groups of sources NO LINE is detected at ~.3 (3σ).2 15 1 5 5 1 15 Velocity [km s 1 ] Not even in compact sources or MIR.6.4 Compact non-detections Extended non-detections.6.4 12µm &4.6µm-bright non-detections Dust-poor non-detections.2.2 bright sources....2.2 15 1 5 5 1 15 Velocity [km s 1 ] 15 1 5 5 1 15 Velocity [km s 1 ]
STACKING THE ATLAS 3D NON-DETECTIONS 6 4 81 ATLAS3D sources HI is not detected in the centre. STACKING: 3σ detection of HI emission N(HI) [ 1 17 cm 2 beam 1 ] 2 2 4 N(HI) ~ 3.5 x 1 17 (T spin /c f ) cm -2 N(HI) converted in optical depth (T spin ~1 K, c f =1) τ~.6 <<.15 we need to stack more to detect this gas in absorption. 6 1 5 5 1 Velocity [km s 1 ] The HI stacking ATLAS 3D is warm? T spin τ ; T spin τ Stacking in absorption even more difficult
INTERPRETING HI ABSORPTION Understand the overall distribution of the HI traced by the absorption line What to can we infer from only the integrated line and the continuum image? Model the rotating HI disk in front of the radio continuum: 3C 35 Observation Model Plane of the sky : x,y Side view: z,y From above: x,z
INTERPRETING HI ABSORPTION 3C 35 Optical Image (SDSS or other): i, PA of the stellar body Continuum image: against which radio lobe there is absorption? i [,18 ] PA [18, 36 ]
INTERPRETING HI ABSORPTION 3C 35 Optical Image (SDSS or other): i, PA of the stellar body Continuum image: against which radio lobe there is absorption? i [,18 ] PA [18, 36 ] 1484 1237 989 MCMC algorithm Count 742 495 find combination of parameters that best fits the observed line the line i = 45 ; PA = 27 PA [ ] 247 36 33 3 27 24 21 1959 1679 1399 1119 84 56 28 Count 18 15 3 45 6 75 I[] 18 21 24 27 3 33 36 PA [ ]
INTERPRETING HI ABSORPTION The bulk of the absorption generated by a rotating disk: i = 45 ; PA = 27 Blue-shifted wing not reproduced by the model.5 Spectrum. Flux [mjy].5 Flux [mjy].1.4.2..2 1 5 5 1 Velocity [km s 1 ] observation model resitudals.4-125 -1-75 -5-25 25 5 75 1 125 Plane of the sky View from above View from the side 4. 4. 4. 2. 2. 2. y [kpc]. z [kpc]. y [kpc]. -2. -2. -2. -4. -4. -4. -4. -2.. 2. 4. x [kpc] -4. -2.. 2. 4. x [kpc] -4. -2.. 2. 4. z [kpc]
INTERPRETING HI ABSORPTION 1923 162 Less information on the source? i [,18 ] PA [, 36 ] Count 1282 962 641 32 36 3 746. 639.4 Flux [mjy] Flux [mjy].1.5..5.1.15.2.5..5 Spectrum 1 5 5 1 Velocity [km s 1 ] observation model.1 resitudals -125-1 -75-5 -25 25 5 75 1 125 PA [ ] 24 18 12 6 2 4 6 8 1 12 14 16 18 I[] 6 12 18 24 3 36 PA [ ] Best-fit solution: i = 55 ; PA = 24 532.9 426.3 319.7 213.1 16.6 Count 8 4. Plane of the sky 8 4. View from above 8 4. View from the side VLBI high resolution continuum y [pc]. -4. z [pc]. -4. y [pc]. -4. Likely we can improve the fit -8. -8. -4.. 4. 8 x [pc] -8. -8. -4.. 4. 8 x [pc] -8. -8. -4.. 4. 8 z [pc]
CONCLUSIONS 27%±5.5% detection rate of HI in absorption HI detected at all redshifts (.2 < z <.23) and radio powers: promising for SHARP, MALS, FLASH Narrow lines HI mainly in a rotating disk P 1.4GHz < 1 24 W Hz -1, Extended sources, dust poor sources Broad asymmetric shifted lines: HI has unsettled kinematics P 1.4GHz> 1 24 W Hz -1, Compact sources (i.e. often young AGN), MIR bright sources Stacking experiments: low optical depth HI is present in the centre of ETGs, warmer HI (T spin >1 K)?
BLIND SURVEYS: AUTOMATIC SEARCH FOR HI ABSORPTION Extract the spectra @ location of every continuum source in the FOV Line-finder: which one? Flux [mjy] Noise [mjy] 5 1 Detection 1.6 1.4 1.2 1..8 1.326 1.328 1.33 1.332 1.334 1.336 1.338 1.34 Frequency [ 1 9 Hz] Flux [mjy] Noise [mjy] 4 2 2 Non Detection 4 1.8 1.6 1.4 1.2 1..8 1.36 1.38 1.31 1.312 1.314 1.316 1.318 1.32 Frequency [ 1 9 Hz] Intervening or associated? cross-correlation with spectral surveys Characterisation absorption (width, centre, asymmetry etc. e.g. busy function) DATABASE Properties of the HI lines Properties of radio continuum (size, spectral index, ) cross-correlation with LOFAR fields & other archives (SDSS, WISE) Comparison with models Stacking other ideas...