1. HI absorption-line science: exciting opportunities with ASKAP- 12 Elaine Sadler University of Sydney / CAASTRO on behalf of the ASKAP FLASH team 5 August 2013

2. Summary 2 Why is an HI absorption-line survey an ideal ASKAP Early Science project? It opens up a completely new parameter space for HI spectral line studies - no other radio interferometer has a wide-band spectral-line capability at 700-1000 MHz (0.5 < z < 1 for the 21cm HI line). It can deliver unique and important new science results in a modest amount of observing time (days to weeks) with ASKAP-12. It will showcase the exceptional radio-quiet qualities of the ASKAP site at frequencies below 1 GHz. Our proposed survey is very flexible, can use any configuration of ASKAP antennas, and could be carried out commensally with the proposed EMU/POSSUM Early Science continuum survey.

3. TechniqueRedshift range MeasuresNotes HI emission-line surveys 0 < z < 0.2Individual galaxies Detection rate drops with redshift HI emission-line stacking 0 < z < 0.4‘Average’ HI properties Detection rate still drops with redshift, but depends on the amount and quality of optical redshift data HI absorption-line survey 0 < z < 1 with ASKAP Individual galaxies Detection rate independent of redshift A key advantage of absorption surveys is that they tell us what kinds of galaxies (uv-bright? dusty?) dominate in an HI-selected sample at high redshift. Important for designing/interpreting stacking surveys. Observing the 21cm HI line

4. Science goal: gas and galaxy evolution The rate at which new stars form in galaxies has decreased by about a factor of 20 over the past 7-8 billion years (from redshift z~1 to 0). What caused this? A decline in the supply of cold neutral gas in galaxies? We don’t know! (Hopkins & Beacom 2006) The changing cosmic star-formation rate:

5. Neutral hydrogen is the missing link in our current models of galaxy evolution. We know almost nothing about the HI content of individual galaxies in the distant universe A wide range of models and simulations exist, and make diverse predictions about the cosmic HI mass density at z>0. Better data are needed to constrain them. The cosmic HI mass density High-redshift measurements of  HI use observations of the Lyman-  absorption line in QSOs at z > 1.7, results are less reliable at low redshift!

6. (Ellison et al. 2001) Optical: Damped Lyman-  absorbers DLAs: Intervening absorbers with high HI column density (N HI > 2 x 10 20 cm -2 ) can be used to detect and study neutral hydrogen in the very distant universe. Ground-based observations of the Lyman-  line are only possible at redshift z > 1.7 DLA Lyman  forest

7. Are optical DLA surveys biased? Zwaan et al. (2005) Optical QSO DLA surveys do not detect the highest column-density absorbers expected on ~0.1% of sightlines, and “do not trace the majority of star-forming gas in the universe” (Ledoux et al. 2003). Dust obscuration?

8. Radio: Intervening HI absorption Darling et al. (2004) Unlike optical, no redshift limit for detecting radio 21cm absorption lines. But do need many targets, wide bandwidth

9. Unique discovery space for ASKAP- 12 The only radio interferometer with a wide-band capability at 700-1000 MHz – provides unique coverage of the HI line at 0.5 < z <1 Radio-quiet site! TelescopeFrequency rangeNotes JVLA (USA)1 – 50 GHzNRL providing a low-band system at 50-500 MHz WSRT Apertif (NL)1.0 – 1.75 GHz1.0 GHz limit set by RFI GMRT (India)1.0 – 1.45 GHzLow-frequency band at 590- 630 MHz Meerkat (S Africa) [Phase 1, 2016-18] 0.9 – 1.67 GHzNo capability below 900 MHz until at least 2018, possibly later GBT (USA)290 MHz – 100 GHzSingle dish, affected by RFI

10. What’s been done so far? Largest existing survey at 0.5 < z < 1 (Darling et al. 2013). ‘Semi-blind’ survey for intervening HI absorption against a sample of 181 bright background radio sources with z > 1.1. Made ten re- detections of known systems, no new detections. “We attribute the lack of new detections in our large survey to severe and persistent RFI… Optical selection bias also contributes”

11. RFI spectrum at the GBT site (via NRAO web pages) Roughly half the GBT band below 1 GHz is lost to RFI. 700- 1000 MHz is considered one of the better regions!

12. RFI spectrum at the ASKAP site | Measured Spectrum at MRO | A. Chippendale Measured Frequency Occupancy (plot from Aaron Chippendale) What fraction of channels are RFI affected at high sensitivity? (percentage of occupied 27.4 kHz channels in 10 MHz blocks in 2hr spectra ) 700-1000 MHz band is extremely clean!

13. What can ASKAP do? Figure of merit: Search path  z set by Number of sources searched (to a given column density limit) multiplied by the Redshift interval searched for each source. ASKAP-12 can easily outperform any existing telescope in searches for high column-density HI absorbers. Huge multiplex advantage from wide field of view!

14. ATCA: Oosterloo et al., targeted HI, z = 0.01 Radio: associated HI absorption Nearby galaxy NGC 6868, continuum flux density ~120 mJy at 1.4 GHz. Associated HI absorption at or near the redshift of radio galaxies and quasars: seen in 10% to 30% of nearby radio galaxies, redshift evolution unclear. Traces gas kinematics in the central regions,and can reveal jet- driven outflows of gas (Morganti et al. 2003, 2005). Science goal: tracing gas flows and AGN triggering in powerful radio galaxies

15. Associated absorption in HIPASS 4 detections in 210 nearby radio galaxies (z < 0.04) (Allison et al. 2013, in prep.) New absorption Strong associated HI absorption linked to presence of OH/H 2 0 megamasers?

16. HIPASS - ATCA comparison 15 arcmin resolution10 arcsec resolution New New absorber Results show that spatial resolution is not critical for detecting these strong HI absorption lines.

17. Early science with ASKAP-12 A 2hr integration with ASKAP-12 will find the strongest (intervening and associated) HI absorption systems, using continuum sources brighter than ~100 mJy. The most effective strategy is to maximize the survey area, then build up sensitivity later as ASKAP is extended. Lines show the detection limit in HI optical depth  versus continuum flux density for a 2-hr integration with 6, 12, 18 and 36 ASKAP antennas. Some known associated and intervening HI absorbers are also shown as individual points. PKS 1814-637: ATCA real-time display, 270 seconds int. HI line has z=0.064,  = 0.19 HI

18. Early Science plans for ASKAP-12 Minimum requirements for HI absorption-line Early Science: Observations in the 700-1000 MHz band – opens up important new discovery space at 0.5 < z < 1 for the HI line (and is free of RFI). A multi-band survey (700-1000, 1000-1300 and 1300-1600 MHz) could also be carried out if feasible and better suited to commensal observations for EMU/POSSUM. This would increase the observing time, but also broaden the redshift range (to 0 < z < 1 for HI). In ~1 week (60 hours, ~1500 sightlines) – first glimpse into a new parameter space, outperforms all existing telescope in searches for high column-density HI absorbers. [~1000 deg 2 of sky, estimate ~5 intervening, 30 assoc. ] In ~1 month (200 hours, ~5000 sightlines) – first unbiased sample of HI-selected galaxies at z > 0.5, answer question of whether QSO DLA surveys are biased. [~3000 deg 2 of sky, estimate ~15 intervening, 100 assoc. lines ] In ~3 months (600 hours, ~15,000 sightlines) – statistical samples of intervening and associated absorbers, can start studying redshift evolution. [~10,000 deg 2 of sky,, estimate ~50 intervening, 300 assoc. lines ]