The Arecibo Legacy Fast ALFA Extragalactic HI Survey:

(The Search for Low-Mass, Gas-Rich Halos)

Martha Haynes (Cornell University) for the ALFALFA team

ALFALFA:

ALFA is not a car…

Arecibo L-band Feed Array

It is a radio “camera”

..on Arecibo 305m telescope

• An extragalactic spectral line survey • Covers 7000 sq deg of high galactic latitude sky• 1345-1435 MHz (-2000 to +17500 km/s for HI line) • 5 km/s resolution• 2-pass, drift mode (total int. time per beam ~ 40

sec)• 1.5-2 mJy rms• 4000 hrs of telescope time, 6-7 years• started Feb 2005; as of end of 2006, 1/3 complete

• One of several major surveys currently ongoing at Arecibo, exploiting its new multibeam capability

http://egg.astro.cornell.edu/alfalfa

ALFALFA as a Legacy Survey

Peebles (2000)

Halo mass function in voids : Gottlöber et al (2003)

•Cosmic voids are filled with low mass dark matter haloes

• ~1000 haloes with M < 109M and vrot< 20 km/s in a 20 h-1

Mpc void are predicted

vrot>55km/s

The Void Phenomenon

Void galaxies in the SDSS : Hoyle et al (2005)

• Void LF has a faint M* but a similar faint-end slope, compared to the overall LF

• Void galaxies are blue, disk-like and have high H equivalent width

Luminosity function of void galaxies

The “Missing Satellite Problem”

• Models/simulations predict large numbers of satellites

• Kauffmann et al. (1993)• Klypin et al. (1999)

• The current census does not count them.

• Faint end slope of the optical LF• Faint end slope of the HIMF• Willman et al. (2005) suggest that

more than half of the MW satellites have not been identified because of extinction, surface brightness limits

• Nothing to detect?• Baryon loss during reionization ( e.g.,

Efstathiou 1992; Barkana & Loeb 1999; Shaviv & Dekel 2003)

• Can they (ever) form stars? (Verde et al. 2002)

Dwarf galaxies

• dE, dSph, dIrr• Low mass: detected only nearby• Dark matter dominated• Low abundances• Stellar mass: 106-108 M

• Blue Luminosity: 106-108 L (MB> -15)• Dynamical mass: 107-109 M

Where they are gas-rich:• HI mass: 106-108 M

•Sometimes, extensive HI•Evidence for dark matter

“Dwarf galaxies of the Local Group” Mateo 1998 ARAA

Leo A

Sagittarius

LGS3

Carina

IZw18

SBS0335-052

Substructure in the Local Group

Diagram from Grebel 1999

Giant spiralsdSph (+dEll)dIrrdIrr/dSph

•Galaxies mainly clustered around the two principal galaxies MW & M31

•Morphological segregation evident

•dE/dSph near large galaxies

•dI at larger distances

The Search for Low Mass Halos

• Do large numbers of low mass “halos” exist?

• If so, do they contain baryons?

• If so, could they be “starless” but gas-rich?=> Extragalactic HI survey

• If so, could they be found preferentially in some environments but not in others?

=> “Fair sample”

ALFALFA is designed to detect low mass gas-rich halos

Parkes HIPASS survey: Zwaan et al. 2003

?

The HI Mass Function

•Previous surveys have included few (if any) objects with HI masses less than 108 M.

•At lowest masses, differ by 10X:

Rosenberg & Schneider (2000)

versus

Zwaan et al. (1997)

Environment & the HIMF

Previous studies based only on Virgo have suggested that the HIMF in Virgo is shallower than in the field

• Only a single cluster• Very small number statistics• Is this just HI deficiency?• Watch out for morphological biases

Springob, Haynes & Giovanelli (2005)• Much larger sample, but optically targeted• Used PSCz density field out to 6000 km/s• Low mass end of HIMF in high density regions

flatter and M* lower• Cannot be explained simply by morphology or HI

deficiency.

Zwaan et al. (2005): HIPASS seems inconsistent

ALFALFA will provide needed statistics and be HI blind.

Previous surveys for HI in voids

for example:

• Weinberg et al 1991 (PPS & foreground void)

• Szoromu et al 1994

• Szomoru et al 1993 (HI in Bootes void galaxy)

• Szomoru et al 1996a (Bootes void)• Szomoru et al 1996b

• Pustilnik et al 2002 (Blue compact galaxies)

HI in “famous” voidsVLA surveys:

•Szomoru et al (1996): Bootes void

HI in “famous” voidsVLA surveys:

•Szomoru et al (1996): Bootes void

HI in “famous” voidsVLA surveys:

•Szomoru et al (1994): PPS & foreground void

Void PPS

# fields 30 12

Int. time/field 40 min 210 min

rms/beam 1 mJy 0.4 mJy

# detections 0 17

Serious limitations:• Relatively small volume sampled• Bandwidth only 6.25 MHz (1200 km/s)• Velocity resolution ~42 km/s (too wide)

ALFALFA Survey

• Commensal with TOGS HI

• Does not compete with galactic plane surveys

High galactic latitude

sky visible from AO

Supergalactic plane

Virgo Leo

1. Determination of the faint end of the HI Mass Function and the abundance of low mass gas rich halos

2. Environmental variation in the HI Mass Function

3. Blind survey for HI tidal remnants

4. Determination of the HI Diameter Function

5. The low HI column density environment of galaxies

6. The nature of HVC’s around the MW (and beyond?)

7. HI absorbers and the link to Ly absorbers

8. OH Megamasers at intermediate redshift 0.16 < z < 0.25

ALFALFA Science Goals

ALFALFA as a Blind HI Survey

• HI mass and distribution (for extended objects)• Normal, star-forming disks• Potential for future star

formation (HI content)• HI deficiency in clusters• History of tidal events• Low mass, LSB dwarfs

• HI absorption: optical depth• Link to Ly-α absorbers

• Redshifts• Rotational velocities

• Dark matter• Distances via Tully-Fisher

relation

HI in M31

Credit: R. Braun

∫ SdV HI mass V Distance ∆V Mass

ALFALFA as a Blind HI Survey

• HI mass and distribution (for extended objects)• Normal, star-forming disks• Potential for future star

formation (HI content)• HI deficiency in clusters• History of tidal events• Low mass, LSB dwarfs

• HI absorption: optical depth• Link to Ly-α absorbers

• Redshifts• Rotational velocities

• Dark matter• Distances via Tully-Fisher

relationCredit: A. Chung

ALFALFA as a Blind HI Survey

• HI mass and distribution (large objects)• Normal, star-forming disks• Potential for future star

formation (HI content)• HI deficiency in clusters• History of tidal events• Low mass, LSB dwarfs

• HI absorption: optical depth• Link to Ly-α absorbers

• Redshifts• Rotational velocities

• Dark matter• Distances via Tully-Fisher

relation

Durrell & DeCesar;

+Yun 1994

ALFALFA as a Blind HI Survey

• HI mass and distribution (large objects)• Normal, star-forming disks• Potential for future star

formation (HI content)• HI deficiency in clusters• History of tidal events• Low mass, LSB dwarfs

• HI absorption: optical depth• Link to Ly-α absorbers

• Redshifts• Rotational velocities

• Dark matter• Distances via Tully-Fisher

relation I Zw 18van Zee et al.

1998

Carignan & Beaulieu 1989

VLA D HI

Arecibo map outer extent [Hoffman et al. 1993]

Extent ofOptical image

MH = 2.5 x 108 M

Mstars = 5.0 x 107 M MDyn = 3.0 x 109 M

DDO154

Are there totally “dark” galaxies?

Optical galaxy

Giovanelli, Williams & Haynes 1989

HI 1225+01

ML

> 200

•ALFALFA is a collaboration of >50 people, from 34 institutions in 13 countries.

•ALFALFA is an open collaboration: anybody with a valid scientific interest and a hardworking disposition can join.

The ALFALFA Team

•Legacy survey on a national telescope used not only for radio astronomy but also radar studies of Solar System objects and the Earth’s upper atmosphere

•Heavy student involvement•Broad institutional

representation

•Cornell University•NAIC•NRAO•Lafayette College•Union College•Wesleyan University•Harvard/Smithsonian CfA•U. of Michigan•Indiana U.•Rutgers U.•U. of Minnesota•U. of Wisconsin•St. Lawrence U.•Humboldt St. U.•Colgate U.•Georgia Southern U.•NRL•NASA/GSFC

•U. of Crete•U. of Tel-Aviv•U. of Rome•U. of Milan II•Obs. Of Brera•U. of Provence•Arcetri Astrophys. Obs.•Lab. D’Astrophysique/Marseille•Cardiff U.•U. de Barcelona•U. de Cordoba (Argentina)•NCRA/GMRT (India)•P. U. Cat. de Chile•Special Astrophys. Obs. (Russia)•U. of Kiev

Participating Institutions

Beam dilutionBeam dilution To sample more deeply, you have to

integrate A LOT longer!

ALFALFA: a wide area survey

For low masses: larger solid

angle

Giovanelli et al. 2005, AJ130, 2598

…

ALFA beams are 3.3’x3.8’

“Almost” fixed azimuth drift

mode

2nd pass offset from the first to

give 1.05’ sampling

ALFALFA strategy: Keep it simple!

● Signal extraction in the Fourier domain by Amélie Saintonge – Match filter over a range of

widths of the template● e.g. 10 km/s – 600km/s

– Choose the width for which the convolution is maximized --> position of the signal

– Calculate the amplitude of the signal from the width

Slide: Amelie Saintonge

•Run on 3-D datacubes after completion of “tile”•Once signals identified, further interactive analysis

Automated signal detection

Exploit VO tools during data processing

• Data processing tools developed here at CU are now running at 11 other institutions

• VO tools incorporated to allow access to external datasets during data processing

• DSS, DSS2, Sloan, 2MASS, NVSS images can be fetched

• NED and other on-line catalogs, including our own, can be accessed and overplotted.

VO portal

http://arecibo.tc.cornell.edu/hiarchive

•ALFALFA data & products will be made public as soon as possible.

•Targeted HI survey and precursor data already available

•Remember: we need a complete dataset – both passes – to make a grid.

•Issue is serving data volume/local cpu:

“google ALFALFA”

Current Status (by 15 Dec 2006)

Total # blocks 177 145

Total hours 1070 690

In map region 1010 625

07h30 – 16h30 +04 to +16 deg22h00 – 03h00 +12 to +16 deg +24 to +32 deg• 1/3 of survey covered with 2 passes (by 15Jan07)

• Processing requires complete sky coverage• Riccardo will present only 4% of final survey…

Survey Beam Area rms min MHI Ndet ts Nlos arcmin sq. deg. (mJy @ 18 km/s) @ 10 Mpc sec

AHISS 3.3 13 0.7 2.0x106 65 var 17,000ADBS 3.3 430 3.3 9.6x106 265 12 500,000HIPASS 15. 30,000 13 3.6x107 4315 460 1.9x106

HIJASS 12. (TBD) 13 3.6x107 (?) 3500 (TBD)J-Virgo 12 32 4 1.1x107 31 3500 3200HIDEEP 15 32 3.2 8.8x106 129 9000 2000

ALFALFA 3.5 7,000 1.7 4.4x106 20,000+ 30 7x106

ALFALFA will be ~ 1 order of magnitude more sensitive than HIPASS with 4X better

angular resolution, 3X better spectral resolution, and 1.6X total spectral

bandwidth

Comparison of blind HI surveys

Giovanelli et al 2005 AstronJ 130, 2598 & 2613 * Aug-Sep 2004 * Candidate Detections Confirmation Run Jan-Feb 2005 * 36 hours of ALFA data

166 confirmed HI sources : - 25 with HI mass > 1010 M

- 4 with HI mass < 107 M (twice as many as all of HIPASS) - high positional accuracy => optical counterparts ID’d - slightly better detection rate than expected (high side), i.e. our ability to reliably dig in low S/N territory is high - system hardware performance, “hands-off” bandpass calibration and baselining (IDL processing pipeline) yield EXCELLENT data quality

ALFALFA Precursor

F

Integrated Flux of 1 Jy km/s

HIPASS Completeness Limit

HIPASS Detection Limit

HIPASS would have detected only a handful; and none of the low mass ones.

Precursor (ALFA commissioning phase) results

In 36 hours, we detected 4X more lowest mass objects than all of HIPASS

First ALFALFA catalog

Giovanelli et al (2006, Astron. J. submitted)

Northern Virgo cluster region11h44m < R.A. < 14h00m

+12º < DecJ < +16º716 detections of good/excellent qualityMedian redshift 7000 km/s

In the same sky region, HIPASS detected only 40 objects.

HIJASS Virgo survey: region of maximum sensitivity:ALFALFA: 193 detections

HIJASS: 13

ALFALFA: Hunting in/around voids

•HI mass function to low masses, and its environmental dependence

•HI detections + HST distances => shapes and kinematics of voids

Tikhonov & Karachentsev 2006 astro-ph/0609109

•Evolutionary history of isolated (never interacting!) galaxies

ALFALFA is designed to hunt for low mass systems efficiently

•High sensitivity (Arecibo = 1/10th SKA!)•Efficient: 97% “open shutter” time•Large solid angle (7000 sq deg)•Moderate angular resolution (~3.5’) => optical i.d. - or

none!•High spectral resolution (5 km/s)•“Minimum intrusion” => high data quality•Dual pol/2 pass coverage (confirmation)•Automated signal detection

Riccardo will talk next, after tea