IRAM 300, rue de la Piscine38406 ST. MARTIN d’HERES (France)Fax: (33/0) 476 42 54 69

PROPOSAL FOR 30M TELESCOPEDeadline: 12 Mar 2009 Period: 01 Jun 2009 — 30 Nov 2009

For IRAM use

Registration N◦:

Date:

TITLERenaissance in the Green Valley? CO in HI-rich Galaxies

Type: Solar system: continuum i lines i other i Extragalactic: continuum i CO lines y other iGalactic: continuum i lines i circumstel. env. i young stel. obj. i cloud struct. i chem. i other i

ABSTRACTThe GASS survey is a large targeted survey at Arecibo designed to measure the HI gas content of 1000 massivegalaxies uniformly selected from the SDSS spectroscopic and GALEX imaging surveys. GASS has identified apopulation of red, early-type galaxies with a significant fraction of atomic gas. We wish to test the hypothesisthat this population has accreted gas from the external environment and is currently transitioning from the redto the blue sequence. We propose IRAM 30m CO observations of 15 massive galaxies with HI mass fractionsgreater than 10%, spanning a range in global NUV-r colour from 2 to 6. We will derive total molecular gasmasses and profiles and thereby constrain whether the redder HI-rich galaxies correspond to galaxies caughtearlier in the accretion stage, when more of the gas is in a phase and a location where it is unable to form stars.

Is this a resubmission of a previous proposal ? no y yes i – proposal number(s): . . . . . . . . . . . . . . . . . . . . . . .Is this a continuation of (a) previous proposal(s) ? no y yes i – proposal number(s): . . . . . . . . . . . . . . . . . . . . . . .

Hours requested for this period:

70LST range(s): from: 09h to: 19h number of intervals: 5

from: 19h to: 05h number of intervals: 2

Special requirements: Large Program i pooled obs i service obs i remote obs i polarimeter i

Scheduling constraints: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .We prefer scheduling in the first half of the semester.

Receivers: EMIR y HERA i Bolometer i Other i

List of Objects (give most common names)

( for additional sources which do not fit hereuse the \extendedsourcelist macro )

Epoch: J2000.0Source RA DEC VLSR

GASS3505 01:17:46.76 13:19:24.5 0.04791GASS4216 01:55:51.99 14:56:24.9 0.04383GASS29596 10:47:02.82 14:15:48.1 0.0324GASS23685 11:23:11.63 13:07:03.8 0.04696GASS24149 12:13:56.25 13:40:35.3 0.04219GASS24426 12:36:22.72 13:36:10.3 0.03053GASS28168 12:40:54.96 08:03:23.2 0.04783GASS35981 13:53:08.36 35:42:50.5 0.04113GASS9483 14:20:56.55 03:52:17.5 0.03541GASS9551 14:27:32.38 04:49:17.9 0.02688GASS9948 14:44:14.75 04:13:06.8 0.0256GASS38717 14:49:29.32 09:04:45.2 0.0406GASS42013 15:16:04.47 06:50:51.5 0.03681GASS11112 23:02:40.31 13:19:44.7 0.02765GASS1977 23:18:15.67 00:15:40.2 0.0298

Principal Investigator:

Guinevere KauffmannMPAKarl Schwarzschildstr. 185748 Garching (Germany)Tel: (+49) 89 30000-2013 Fax: (+49)89 30000-2235Email: gamk@mpa-garching.mpg.de

Other Investigators (name, institution):

Barbara Catinella (MPA – Germany); Silvia Fabello(MPA – Germany); Riccardo Giovanelli (Cornell –USA); Javier Gracia-Carpio (MPE – Germany); MarthaHaynes (Cornell – USA); Timothy Heckman (JHU –USA); Cheng Li (MPA – Germany); Amelie Saintonge(University of Zurich – Switzerland); David Schimi-novich (Columbia University – USA); Linda Tacconi(MPE – Germany);

Expected observer(s) Gracia-Carpio,Catinella,Fabello

join to this form: scientific aims ≤ 2 typed pages (≤ 4 pages for Large Programs) and ≤ 2 pages Figs., Tabs., and Refs.

Technical Summary

? EMIRNote that up to 4 IF signals can be recorded and up to 2 EMIR (always dual polarization) bands can be combined in one EMIR

setup. For a summary of EMIR connectivity consult the IRAM Granada home page or the Call for Proposals

TransitionsT ∗A = expected line antenna temperature; ∆v = required velocity resolution.

setup band species transition frequency T ∗A rms ∆v backend a)GHz mK mK km s−1

1 E0 CO 1-0 110 5–10 1–2 30–40 W2 E1 CO 2-1 222 5–10 60 W

a) V: VESPA, W: WILMA, 4: 4 MHz filterbank, 1: 1 MHZ filterbank

Observing parametersmap size in arcmin; T = requested telescope telescope time per setup

setup map size mapping switching T remarkNo. ∆x×∆y mode a) mode b) [h]

1 × none WSw 70 —Total EMIR time requested: 70

a) none, OTF (on–the–fly), R: Rasterb) PSw: position switching, FSw: frequency switching, Wsw: wobbler sw.

Renaissance in the Green Valley? CO in HI-rich Galaxies1

Scientific background: Perhaps the most fundamental new discovery of the large imaging and spec-troscopic surveys of galaxies that have been carried out over the past decade is that there are two verydistinct populations of galaxies: a ”red sequence” consisting of dense, massive galaxies with little or noongoing star formation, and a ”blue cloud” consisting of low mass, low surface density galaxies that arestill forming stars (e.g. Strateva et al 2001; Kauffmann et al 2003; Baldry et al 2004). The integratedstellar mass density in the red sequence increases by a factor ∼ 2 from redshift 1 to zero (e.g. Bell et al2004, Faber et al 2007), leading to the idea that there is a net migration of galaxies from the blue cloudto the red sequence as a function of cosmic time. In between the red sequence and the blue cloud, thenumber density of galaxies reaches a local minimum. Because UV/optical colours are sensitive to evena trace amount of unobscured star formation in the galaxy, UV/optical colour-magnitude diagrams havelarge dynamic range and are extremely effective at picking out a so-called ”green valley” population thatlies between the red and the blue clouds (e.g. Schiminovich et al 2007, see Figure 1).

The green valley is hypothesized to consist of galaxies that are in the process of transitioning betweenthe red and blue sequences. Because there is net growth in the mass contained in red galaxies from redshift1 to 0, considerable attention has been given to physical processes that cause star formation to shut downin galaxies, so called quenching mechanisms. Examples include supernova feedback, gas heating by radio-loud AGN in clusters, galaxy-wide winds generated from material falling into black holes in accretiondisks, ram-pressure stripping of gas, ”harrassment” effects due to galaxy encounters in rich environments,”starvation” as galaxies run out of fuel etc. Observations of atomic and molecular gas in galaxies play animportant role in constraining some of these processes. For example, CO observations of spiral galaxies inthe Virgo cluster revealed that although these systems were frequently deficient in HI, they had normal COluminosities and extents relative to optical properties, lending support to processes such as ram-pressurestripping that selectively remove low density gas in the galaxies (Kenney & Young 1989)

Although observations show that there is a net migration of galaxies to the red sequence, we stillexpect that galaxies might transition through the green valley because they accrete new gas from theexternal environment (e.g. Kauffmann et al 2007). Microwave background experiments such as WMAPnow constrain the fraction of baryonic matter in the Universe to be 4.5% of the critical density (Spergel etal 2007). Remarkably, only 3.5% of all the baryons are currently in stars, the rest must be in the form ofgas. One of the major difficulties faced by current galaxy formation models is to explain why the efficiencywith which gas cools and condenses onto galaxies is so low. AGN feedback has been proposed as a possiblesolution, but for galaxies in lower-mass halos, the fraction of radio-loud AGN drops sharply (Best et al2005). It is thus reasonable to suppose that outside massive clusters, galaxies may still undergo cycles ofrebirth through gas accretion. One way to determine observationally if rebirth is occurring, would be tocatch objects early in the accretion stage when they have a lot of gas in a phase (HI) and location whereit is not yet forming stars.

The GASS HI Survey: The GALEX Arecibo SDSS Survey (GASS) is an ongoing large targetedsurvey at Arecibo, home to the world’s largest single-dish radio telescope. GASS is designed to measurethe neutral hydrogen content of a representative sample of 1000 galaxies uniformly selected from the SDSSspectroscopic and GALEX imaging surveys, with masses in the range 1010 − 1011.5M¯ and redshifts inthe range 0.025 < z < 0.05. Integration times are set so that we should detect all galaxies with HImass fractions of 1.5% or more. GASS will produce the first statistically significant sample of massivetransition galaxies with homogeneously measured stellar masses, star formation rates and gas properties.Observations started in March 2008, and are expected to be completed over a period of 3 years. Moreinformation is available at http://www.mpa-garching.mpg.de/GASS.

One of the earliest results to emerge from GASS is that there is indeed a population of red, early-typegalaxies with significant amounts of HI gas. These galaxies have global UV/optical colours that placethem on the red sequence, early-type morphologies, and SDSS spectra that indicate old stellar populationsand no trace of emission lines, but they also have HI mass fractions (evaluated within the 3.5 arcminuteArecibo beam) that reach values in excess of 40%. We have ruled out the presence of any close companionsthat may confuse our measurements. Interestingly, for some of these systems, the GALEX FUV and NUVimages show evidence for extended UV structures. Figure 3 shows SDSS and GALEX images of the 15targets in this proposal (see below). As can be seen, there are extended UV structures in the 2 reddest

1These data will form part of the PhD thesis of Silvia Fabello, who has just started a PhD at MPA. She is supervised byGuinevere Kauffmann.

galaxies, GASS 24149 and 3505, in the form of rings and ripples. Both these galaxies have NUV-r coloursthat place them firmly in the red sequence, but the HI mass fractions are 0.56 and 0.44!

The fact that 15-20% of all elliptical galaxies have significant HI gas content has been known for manyyears (Knapp, Turner & Cuniffe 1985). However, these galaxies were studied as a separate class of objects,in isolation from the rest of the more “normal” galaxy population. Our aim is to understand if and howour anomalous HI-rich early-type galaxies relate to the population of normal spirals with similar stellarmasses.

Proposed observations: We propose IRAM 30m CO observations of 15 galaxies with stellar massesin the range 1010−1011M¯ selected from the GASS parent sample (a larger super-set of galaxies that meetthe survey selection criteria). We require that the galaxies already have HI detections from either GASS,the ALFALFA survey (Giovanelli et al 2005), or the Cornell archive (Springob et al 2005), and that theHI mass fraction be greater than 10%. Example Arecibo spectra are shown in Figure 2. In order span thefull transition regime, our sample is selected to have a large range in global NUV-r colour from 2 ( edge ofthe blue sequence) to 6 (oldest galaxies on the red sequence), and to have i < 65 deg, so that the colourestimates are not severely affected by dust attenuation. SDSS and GALEX images of our 15 target galaxiesare shown in Figure 3. We have checked that none of these objects have existing CO measurements in theliterature. As can be seen, the bluest galaxies in our sample look like ordinary star-forming spirals, but asNUV-r increases, fewer signs of star formation such as HII regions and spiral arms are apparent, and thereddest galaxies look for the most part like normal early-type galaxies, even though they contain as muchatomic gas as the bluest objects (right hand panel of Figure 1). The CO observations we propose willallow us to test the hypothesis that the 15 galaxies in our sample constitute an evolutionary sequence, withthe reddest objects corresponding to galaxies that have accreted their gas most recently. This proposalis a pilot study to test the above scenario. If successful, we will submit a larger, comprehensive proposaltargeting a selection of GASS galaxies for the fall 2009 deadline.

Observing strategy: We ask for 5 pointings per galaxy: one at the center, and the others spreadaround the galaxy. Where the GALEX images show evidence for extended structures, we will ensure thatone or more of these outer pointings coincide with the UV detections. We will compute the total moleculargas content of each galaxy and compare it with the HI mass. The extra pointings away from the centers willallow us to derive a rough molecular gas density profile and specify whether the gas is centrally concentrated,as is usually found in spiral galaxies, and whether the degree of central concentration changes systematicallyas a function of NUV-r colour. The aim is to use the CO observations to constrain the evolutionary stateof the gas and to quantify how the anomalous HI-rich early-type galaxies differ when compared to ordinarystar-forming spirals with similar stellar masses and atomic gas fractions. If we are truly seeing galaxiesthat are migrating from the red sequence to the blue cloud, we might hope to uncover ”mirror images” ofthe HI-deficient spirals studied by Kenney & Young, that is HI-rich early type galaxies that are stronglydeficient in molecular gas, or with larger molecular extents compared to optical properties.

These straightforward 3mm CO 1-0 observations require average observing conditions; they are ideal forthe summer/fall observing period. Assuming molecular masses comparable to the HI masses, and centrallyconcentrated CO emission, we will need to integrate to an rms of ∼1-2 mK per point. This is the limitto detect ∼109 M¯ per pointing (for a Galactic X factor), or ∼10% molecular gas fraction for our lowestmass galaxies. Using the EMIR time estimator (which yields very conservative integration estimates)this is achievable in 30–60 minutes of telescope time per pointing, including all overheads, depending onweather conditions, and after smoothing to spectral resolutions of 30–40 km/s. The redshifts of thesetargets range from z=0.025–0.047 or cz= 7500–14,000 km/s. Thus, all can be observed with one tuning ofEMIR, assuming 4 GHz bandwidth per polarization available with the current IF transport and backendcapabilities of the 30-m. In case of excellent weather, we will observe simultaneously the CO 2-1 line, butthis is not a driver for the proposal. We request a total allocation of 70 hours to complete this pilot study.

ReferencesBaldry, I. et al 2004, ApJ, 600, 681; Bell, E. et al 2004, ApJ, 608, 752; Best, P. et al 2005, MNRAS, 362,25; Faber, S. et al 2007, ApJ, 665, 265; Giovanelli, R. et al 2005, AJ, 130, 2598; Kauffmann, G, et al 2003,MNRAS, 341, 54; Kauffmann, G. et al 2007, ApJS, 73, 173, 357; Kenney, J. & Young, J. 1989, ApJ, 344,171; Knapp, G., Turner, E. & Cuniffe, P.E. 1985,AJ, 90, 454; Morganti, R. et al 2006, MNRAS, 371, 157;Schiminovich, D. et al 2007, ApJS, 173, 315; Springob, C. et al 2005, ApJS, 160, 149; Spergel, D. et al.2007, ApJS, 170, 377; Strateva, I. et al 2001, AJ, 122, 1861.

Figure 1: Left: UV/optical colour-magnitude diagram showing the blue cloud, red sequence and greenvalley. Right: The location of our 15 proposed targets are shown as black points in the plane of atomic gasfraction versus NUV-r. The location of blue sequence spirals is shown by the blue points (from ALFALFA).The location of typical ”red and dead” ellipticals is shown as the red ellipse (Morganti et al. 2006).

Figure 2: HI spectra from Arecibo for two of the galaxies in our sample.

GASS No. RA DEC z iso rad inclin NUV-r log M* MHI/M*35981 208.28484 35.714027 0.04113 75.8 64 2.15 10.297 1.0424426 189.09468 13.602871 0.03053 52.2 36 2.57 10.125 0.614216 28.96662 14.940255 0.04383 48.3 24 2.65 10.737 0.3238717 222.3722 9.079212 0.0406 52.6 56 2.97 10.415 0.199483 215.23561 3.871523 0.03541 41.6 16 3.00 10.713 0.1023685 170.79846 13.117716 0.04696 41.6 51 3.31 10.704 0.2828168 190.22902 8.056454 0.04783 20.9 31 3.32 10.181 0.461977 349.56528 0.261179 0.0298 81.7 53 3.56 10.781 0.199948 221.06145 4.218545 0.0256 96.4 37 3.75 10.696 0.259551 216.8849 4.821626 0.02688 77.6 59 4.09 10.91 0.1011112 345.66794 13.329071 0.02765 66.6 48 4.30 10.815 0.1642013 229.01863 6.847628 0.03681 53.5 51 4.34 10.771 0.2529596 161.76176 14.26335 0.0324 54.4 37 5.26 10.676 0.1024149 183.48438 13.676459 0.04219 42.2 28 5.40 10.327 0.563505 19.44485 13.323483 0.04791 26.8 21 5.93 10.211 0.44

Table of the target galaxies: iso rad is the SDSS isophotal r-band diameter of the galaxy in arcseconds,inclin is the inclination in degrees.

35981 24426 4216 38717 9483

23685 28168 1977 9948 9551

11112 42013 29596 24149 3505

Figure 3: SDSS g,r,i colour images (top) and GALEX FUV/NUV(bottom) images of our 15 target galaxies.The galaxies are arranged in order of increasing NUV-r colour. Images are 2×2 arcminutes for both theoptical and UV images (but note that GALEX has a resolution of 5 arcsec). There are extended UVstructures in the 2 reddest galaxies, GASS 24149 and 3505. Extended UV structures are also visible in23685, 1977, 9948.