The SAMI Galaxy Survey: extraplanar gas, galactic winds, and their association with star formation history

8.5 MB·

by  I-Ting Ho

MNRAS000,1–25(2015) Preprint26January2016 CompiledusingMNRASLATEXstylefilev3.0
The SAMI Galaxy Survey: extraplanar gas, galactic winds,
and their association with star formation history
I-Ting Ho (何宜庭)1,2(cid:63), Anne M. Medling2, Joss Bland-Hawthorn3, Brent Groves2,
Lisa J. Kewley1,2, Chiaki Kobayashi4, Michael A. Dopita2,5, Sarah K. Leslie2,
Rob Sharp2,6, James T. Allen3,6, Nathan Bourne7, Julia J. Bryant3,6,8,
Luca Cortese9,10, Scott M. Croom3,6, Loretta Dunne7, L. M. R. Fogarty3,6,
6
1 Michael Goodwin8, Andy W. Green8, Iraklis S. Konstantopoulos8,11,
0
Jon S. Lawrence8, Nuria P. F. Lorente8, Matt S. Owers8,12, Samuel Richards3,6,8,
2
n Sarah M. Sweet2, Edoardo Tescari6,13 and Elisabetta Valiante14
a 1Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822, USA
J 2Research School of Astronomy and Astrophysics, Australian National University, Canberra, ACT 2611, Australia

3 3Sydney Institute for Astronomy, School of Physics, University of Sydney, NSW 2006, Australia
2 4Centre for Astrophysics Research, Science and Technology Research Institute, University of Hertfordshire, Hertfordshire AL10 9AB, UK
5Astronomy Department, King Abdulaziz University, P.O. Box 80203, Jeddah, Saudi Arabia

] 6ARC Centre of Excellence for All-sky Astrophysics (CAASTRO)
A 7Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh, EH9 3HJ, UK
G 8Australian Astronomical Observatory, PO Box 915, North Ryde, NSW 1670, Australia
9Centre for Astrophysics & Supercomputing, Swinburne University of Technology, Victoria, Australia
.
h 10International Centre for Radio Astronomy Research, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
p 11Envizi Group Suite 213, National Innovation Centre, Australian Technology Park, 4 Cornwallis Street, Eveleigh, NSW 2015, Australia
- 12Department of Physics and Astronomy, Macquarie University, NSW 2109, Australia
o 13School of Physics, University of Melbourne, Parkville, VIC 3010, Australia
r 14School of Physics and Astronomy, Cardiff University, Queen’s Buildings, Cardiff, CF24 3AA, UK
t
s
a
[
Accepted2016January04.Received2016January03;inoriginalform2015September22

2
v
2 ABSTRACT
2 We investigate a sample of 40 local, main-sequence, edge-on disc galaxies using inte-
0
gral field spectroscopy with the Sydney-AAO Multi-object Integral field spectrograph
2
(SAMI) Galaxy Survey to understand the link between properties of the extraplanar
0
gas and their host galaxies. The kinematics properties of the extraplanar gas, includ-
.
1 ing velocity asymmetries and increased dispersion, are used to differentiate galaxies
0 hosting large-scale galactic winds from those dominated by the extended diffuse ion-
6
ized gas. We find rather that a spectrum of diffuse gas-dominated to wind dominated
1
galaxiesexist.Thewind-dominatedgalaxiesspanawiderangeofstarformationrates
:
v (−1 (cid:46) log(SFR/M(cid:12)yr−1) (cid:46) 0.5) across the whole stellar mass range of the sample
Xi (8.5(cid:46)log(M∗/M(cid:12))(cid:46)11). The wind galaxies also span a wide range in SFR surface
densities(10−3–10−1.5 M(cid:12) yr−1 kpc−2)thatismuchlowerthanthecanonicalthresh-
r
a old of 0.1 M(cid:12) yr−1 kpc−2. The wind galaxies on average have higher SFR surface
densities and higher Hδ values than those without strong wind signatures. The en-
A
hancedHδ indicatesthatburstsofstarformationintherecentpastarenecessaryfor
A
drivinglarge-scalegalacticwinds.WedemonstratewithSloanDigitalSkySurveydata
thatgalaxieswithhighSFRsurfacedensityhaveexperiencedburstsofstarformation
in the recent past. Our results imply that the galactic winds revealed in our study are
indeeddrivenbyburstsofstarformation,andthusprobingstarformationinthetime
domain is crucial for finding and understanding galactic winds.
Keywords: galaxies:evolution–galaxies:starburst–galaxies:kinematics–galaxies:
ISM
(cid:63)(cid:13)cE2-0m1a5ilT:[email protected]
2 I.-T. Ho et al.
1 INTRODUCTION 2002) and NGC1482 (Veilleux & Rupke 2002), open-ended
bipolar structures can extend several kpc from the central
In the standard picture of galaxy formation and evolution,
energy injection zones undergoing concentrated starbursts.
thefeedbackrelatedtoprocessesthatdriveenergyandmo-
Strong supernova feedback drives multiphase gas consist-
mentumintotheinterstellargasservetoregulatetheassem- ing of hot 107−8 K X-ray emitting gas, warm 104 K ion-
blyofbaryonicmatterindarkmatterhaloes.Outflowsfrom
ized gas, and cold neutral atomic and molecular gas, to a
galaxiespreventfurthergasaccretionandejectgas,metals,
speed of a few hundred kilometres per second (see Veilleux
and energy out to many kpc into their haloes. The overy
et al. 2005 for a review). The optical line-emitting gas usu-
of large amounts of gas, metals, and, dust in the circum-
ally presents well-defined conical structures, which are the
galactic medium (at a few hundred kpc radius) has estab-
limb-brightened parts of the expanding X-ray bubbles that
lishedtheimportantrolethatthishalomattermustplayin
also entrain the ambient cold gas. As the bubbles expand
galaxyevolution(M´enardetal.2010;Tumlinsonetal.2011;
supersonically,shockwavesexciteopticalemissionlinesand
Werketal.2013,2014;Peeketal.2015).Thehalogasmay produce the characteristically large [Oi], [Oii], [Nii] and
eventually cool down and fall back to the disc to feed sub- [Sii] to Balmer line ratios. This extraplanar gas in nearby
sequent star formation (i.e. the“galactic fountain”picture;
prototype wind systems represents a very violent form of
Shapiro&Field1976;Bregman1980;deAvillez2000),orit
interactions between discs and haloes.
may be lost to the system through interactions with other
In galaxies lacking spectacular large-scale winds, nar-
galaxies. The interplay between the discs and their haloes
rowband Hα imaging in nearby late-type edge-on galaxies
canstronglyinfluencethedifferentpathwaysgalaxiesevolve
reveals that eDIG is very common, with more than half of
upon over cosmic time.
thegalaxiesshowingextraplanardiffuseemissionsandsome-
Starformationandactivegalacticnuclei(AGN)arethe
times filamentary structures (e.g. Rossa & Dettmar 2000,
two major energy sources capable of ejecting baryonic mat-
2003a,b; Miller & Veilleux 2003a; Rossa et al. 2004). The
ter from the disc into the halo. Different energy input rates
averagedistancesoftheextendedemissionabovethegalac-
and their time-scales dictate the form of interactions be-
tic midplane can range from 1–2 to 4 kpc or more. Kine-
tween galaxies and their haloes.
maticstudiessuggestthattheeDIGco-rotateswiththehost
At the very energetic end, powerful AGNs can deposit galaxy and usually presents a slight lag in the azimuthal
energy and momentum from small to large scales, ejecting velocity, which increases with increasing off-plane distance
multiphasegasatvelocitiesofmorethanathousandkilome- (Healdetal.2006b,a,2007).O-starsinHiiregionsarelikely
trespersecond(e.g.Tremontietal.2007;Rupke&Veilleux to contribute significantly to the excitation of eDIG (Miller
2013; Veilleux et al. 2013). At intermediate energy, galactic & Cox 1993; Dove & Shull 1994). Escaping Lyman contin-
winds driven by star formation can typically reach a speed uumphotonsfromO-starHiiregionstravellinginalowden-
of a few hundred kilometres per second and are known to sity,fractalmediumorsuperbubblechimneycouldreachkpc
be very common at high redshifts (z > 1) where the cos- scalesandexcitetheextraplanargas.Thispicturenaturally
micstarformationrates(SFRs)arehigh(e.g.Weineretal. explains the observed correlations between the far infrared
2009; Steidel et al. 2010). In the local Universe, starburst- luminosityperunitareaandextraplanarionizedmass(and
driven winds are ubiquitous in galaxies with high enough also the presence of extraplanar emission Miller & Veilleux
SFR surface densities (Σ > 0.1M(cid:12) yr−1 kpc−2; Heckman 2003a; Rossa & Dettmar 2003a). This picture also explains
2002). Although present-day normal galaxies do not have the strong spatial correlation between strong diffuse emis-
such high SFR surface densities, the prevalence of galactic sion and Hii regions in the disc (Zurita et al. 2000, 2002).
winds is still implied statistically by stacking analysis with However, Miller & Veilleux (2003b) modelled the forbidden
the Nad absorption lines (Chen et al. 2010). At intermedi- line([Oiii]λλ4959,5007,[Oi]λ6300,and[Sii]λλ6716,6731)
ate redshifts (0.3 < z < 1), the ubiquity of galactic-scale to Balmer line ratios of the extended filaments which typi-
outflows has also been inferred by absorption line studies cally increase with increasing height off the disc plane (e.g.
in individual systems using 10-m class telescopes (e.g. Ru- Otteetal.2001,2002).Theyfoundthatinmostcasespho-
bin et al. 2014; Schroetter et al. 2015; see also Sato et al. toionizationisunlikelytobetheonlyexcitationmechanism,
2009;Martinetal.2012;Korneietal.2012).Finallyatlow with hybrid models combining photoionization and turbu-
energy, the extended diffuse ionized gas (eDIG), as part of lent mixing layers or shocks better explaining the optical
thewarmionizedmediumseenbothinourMilkyWay(also line ratios (see also Reynolds et al. 1999; Collins & Rand
known as the Reynolds Layer; Reynolds et al. 1973) and 2001; Seon 2009; Barnes et al. 2014).
external galaxies represents the interface between the hot Understanding the underlying physical processes and
haloes and the cold discs (see Haffner et al. 2009 for an ex- galaxy properties that drive the different types of disc-halo
tensive review). interactions is critical for building a more comprehensive
Theinteractionsbetweendiscactivitiesandhaloes(the picture of the cycle of baryonic matter in galaxy evolution.
so-called“disc-halo interactions”) can be effectively probed Extraplanar gas can shed light on the different strengths
bystudyoftheextraplanaremissionsattheinterfacewhere andformsoftheinteractions.Observingsuchinteractionsin
the interactions occur. Edge-on galaxies provide the best largenumbersofgalaxieshasbeenmadepossiblerecentlyby
viewing angle for the investigation of the structure, excita- virtue of the growth of integral field spectroscopy. The on-
tion, and dynamics of the extraplanar gas. going integral field spectroscopy (IFS) surveys, such as the
In nearby edge-on galaxies that host well-known star- theCalarAltoLegacyIntegralFieldAreaSurvey(CALIFA;
burstdrivenwinds,suchasM82(Axon&Taylor1978;Bland Sa´nchez et al. 2012), the Sydney-AAO Multi-object Inte-
& Tully 1988; Leroy et al. 2015), NGC253 (Westmoquette gral field spectrograph (SAMI) Survey (Croom et al. 2012;
etal.2011),NGC3079(Veilleuxetal.1994;Ceciletal.2001, Bryant et al. 2015), and the Mapping Nearby Galaxies at
MNRAS000,1–25(2015)
SAMI: extraplanar gas and galactic winds 3
Apache Point Observatory (MaNGA) Survey (Bundy et al. 16
2015), are delivering three dimensional datacubes of hun- al12
g 8
dreds to tens of thousands of galaxies and starting to revo- N
4
lutionize the way galaxies are studied. In this work, we will 0
12
explore extraplanar emissions in edge-on discs using data
10
from the SAMI Galaxy Survey and investigate the connec- SAMIFOV
c] 8 (r=7.5arcsec)
tion between properties of the extraplanar gas and physical e
s
properties of the galaxies. arc 6
The paper is structured as follows. Section 2 describes r[e 4
our sample selection. We uss the analysis of our IFS dat- 2
medianPSF(FWHM=2.1arcsec)
acubes in Section 3, and introduce in Section 4 a novel ap- 0
8 9 10 11 0 5 10 15 20
proach of empirically identifying wind-dominated galaxies.
InSection5,wedetailthederivationofhostgalaxyproper- 12
tiesincludingstarformationrateandstarformationhistory. 10
In Section 6, we investigate common physical properties of c] 8
p
wind-dominatedgalaxies.Finally,weusstheimplicationsof [k 6
ourresultsinSection7andprovideasummaryandconclu- re 4
sion in Section 8.
2
Through out this paper, we assume the concordance Λ
0
cold dark matter cosmology with H = 70 km s−1 Mpc−1, 8 9 10 11 0 5 10152025
0
Ω = 0.3 and Ω = 0.7. We infer cosmological distances 0.3
M Λ
from redshifts that have been corrected for local and large-
scale flows using the flow model by Tonry et al. (2000). 0.2
a
/
b
0.1
2 SAMPLE SELECTION
Weselectedge-onstar-formingdiscgalaxiesfromtheSAMI 0.0 8 9 10 11 0 5 10152025
GalaxySurvey.Theon-goingSAMIGalaxySurveyobserves
low redshift galaxies (z ≈ 0.05) using deployable imaging 0.08 AGN
fiber bundles (“hexabundles”) to obtain spatially-resolved
opticalspectraoveracircularfieldofview(FOV)of15arc- hift0.06
s
secindiameter(Croometal.2012).Thesurveyisconducted ed0.04
R
on the 3.9-m Anglo-Australian Telescope using the flexi-
ble AAOmega dual-beam spectrograph (Smith et al. 2004; 0.02
Saunders et al. 2004; Sharp et al. 2006). For a comprehen- 8 9 10 11 0 51015202530
sive ussion of the sample selection, the reader is directed to log(M /M ) Ngal
∗ (cid:12)
the survey sample selection paper by Bryant et al. (2015).
Figure1.Distributionsofourgalaxiesinstellarmass,r-bandef-
In brief, the SAMI sample is selected from the Galaxy And
fectiveradius,minor-to-majoraxisratio,andredshift.Thecircles
Mass Assembly project (GAMA; Driver et al. 2009) using
and stars are the 82 galaxies that satisfy our geometrical (incli-
stellar-mass cut-offs in redshift bins up to z < 0.12. The
nation angle >80 degrees [b/a <0.26] and re <12 arcsec) and
GAMA sample covers broad ranges in stellar mass and en- morphological(notmajormerger)selections.Theopencirclesare
vironment.SAMIalsoincludesanadditionofeightclusters galaxieslackingsufficientHαdetectionsoutside1re (Ngal =33)
toprobethehigherdensityenvironments.Ouredge-ondiscs and the open stars are AGNs (N =9). These galaxies are ex-
gal
areselectedfromtheGAMAsamplethatcontainsabout830 cludedfromthiswork.Ourfinalsamplecorrespondstothefilled
galaxies at the present stage of the survey (June 2015). blue circles (Ngal =40). The blue and black histograms are the
Weimposecriteriaonthegalaxyinclinationanglesand distributionsforourfinalsampleandallthe82galaxies,respec-
effective radii to ensure that the selected galaxies are edge- tively.Greyscaleinthere [kpc]versusstellarmasspanelshows
themass-sizerelationofgalaxiesintheGAMAsurveywithincli-
ondiscswell-coveredbytheSAMIhexabundles.Werequire
nationangles>80degrees(b/a<0.26).
minor-to-majoraxisratio(b/a)inther-bandtobelessthan
0.26,whichcorrespondstoalowerlimitoninclinationangle
of80degrees,assumingtheclassicalHubbleformula1 (Hub-
Kelvin et al. 2012). We reject major mergers showing clear
ble 1926). We also require the r-band effective radii (r ; or
e tidal features because their complex large-scale winds are
half-lightradii)tobelessthan12arcsec,suchthattheSAMI
notthesubjectofthisstudy(e.g.Wildetal.2014).Thesege-
hexabundles cover more than 50% of the total light for the
ometricalandmorphologicalselectioncriteriayield82galax-
majorityofthesample.Theeffectiveradiiweremeasuredon
ies.Fromthissample,wefurtherrejectAGNsasthosegalax-
Sloan Digital Sky Survey (SDSS; York et al. 2000) r-band ieswhoselineratios[Nii]λ6583/Hαand[Oiii]λ5007/Hβat
imagesbytheGAMAteamusinggalfit(Pengetal.2010;
thecentralspaxelsarebelowthetheoreticalmaximumstar-
burst line by Kewley et al. (2001). Galaxies lacking enough
spaxels outside 1r detected in Hα are also not considered
1 cos2i= (b/1a−)2q−2q02,whereiistheinclinationangleandq0 isa because their datae do not yield strong constraints on the
0
constantof0.2(i=90◦ forb/a<q0). extraplanar gas properties (see Section 4). These galaxies
MNRAS000,1–25(2015)
4 I.-T. Ho et al.
SDSS Hα Vgas σgas
15 3 60
2m] 30 60
sec] 105 1216[10/ergs/s/c− −030[km/s] 2400[km/s]
c
[ar 0 log([NII]/Hα) log([SII]/Hα) −60 log([OI]/Hα) 0
c. -0.5 -0.1 -0.8
e
D 5 5
δ − -0.2 -1.0
-0.6
0
10 -1.2
− -0.3
5 -0.7
− -1.4
15
− 15 10 5 0 -5 -10 -15 -0.4
5 0 -5
δ R.A.[arcsec]
Figure 2. Examples of 2D maps reconstructed from the IFS datacubes of GAMAJ115927.23-010918.4 (CATAID: 31452). The dashed
redcircleintheSDSS3-colourimage(rightpanel)showsapproximatelytheFOVandpointingofSAMI(15arcsecindiameter).Note
the increase in velocity dispersion and line ratios off the disc plane and the disturbed velocity field are all indicative of galactic winds.
WeadoptS/Ncutsof3inthesemaps.Northisupandeastisleft.
without good detections on average have lower star forma- covers≈6300–7425˚Awithaspectralsamplingof0.57˚Aand
tion rates and lower star formation surface densities than spectral resolution of R ≈ 4500 or FWHM ≈ 65 km s−1.
those with good detections. After imposing these two crite- The angular resolution of the data is in the range of 1.4 –
ria, we work with 40 galaxies, representing approximately 3.0 arcsec, with a median value of 2.1 arcsec. The median
5% of the current SAMI GAMA sample. angular resolution corresponds to about 0.8 to 2.8 kpc at
Figure 1 shows the stellar mass, redshift, b/a, and r the distances of our sample.
e
distributions of our sample. The filled circles are our final
sample(N =40).Theopensymbols(circlesandstars)are
gal
those satisfying our geometric and morphological selections
but either have insufficient detections of extraplanar emis-
3.2 Emission line fitting
sions (N =33) or are AGN hosts (open stars; N =9).
gal gal
We adopt the photometrically-derived stellar masses from Weapplyspectralfittingtoextractemissionlinefluxesand
the GAMA survey (Taylor et al. 2011). The strong correla- kinematicinformationfromthedatacubes.Ouremissionline
tion between redshift and stellar mass is a direct result of fitting toolkit lzifu (Ho et al. in preparation; see also Ho
theselectionfunctionofSAMI(seeBryantetal.2015forde- et al. 2014) is employed to construct 2-dimensional maps.
tails).Thelackofcorrelationbetweenreinarcsecandstellar lzifuadoptsthepenalizedpixel-fittingroutine(ppxf;Cap-
mass suggests that our resolution normalized to galaxy size pellari & Emsellem 2004) to model (and subsequently sub-
is on average the same for galaxies of different masses . tract) the stellar continuum and the Levenberg-Marquardt
least-squaresmethod(Markwardt2009)tofitGaussianpro-
filestotheuser-assignedemissionlines.Forfittingthestellar
3 DATA ANALYSIS continuum, we adopt the theoretical simple stellar popula-
tion (SSP) models assuming Padova isochrones of 24 ages3
3.1 Data reduction and 3 metallicities4 from Gonza´lez Delgado et al. (2005).
The integral field spectroscopic data are reduced using the We model 11 emission lines simultaneously as simple Gaus-
automatic data reduction pipeline described in Sharp et al. sians requiring them to share the same velocity and veloc-
(2015)andAllenetal.(2015).Wefirstreducethespectrum ity dispersion. The 11 emission lines are [Oii] λλ3726,3729,
of each fiber following the standard data processing proce- Hβ,[Oiii]λλ4959,5007,[Oi]λ6300,[Nii]λλ6548,6583,Hα,
duresforopticalspectroscopyusingthe2dfdr2 datareduc- and[Sii]λλ6716,6731.Wefixtheratios[Oiii]λλ4959/5007
tionpipeline(Croometal.2004),andsubsequentlyresample and [Nii] λλ6548/6583 to their theoretical values given by
the spectra on to a rectangular grid of 0.5 arcsec using the quantum mechanics. Hereafter, we omit the wavelength no-
drizzletechnique(Fruchter&Hook2002;Sharpetal.2015). tation when appropriate, i.e., [Oiii]≡ [Oiii] λ5007, [Nii]≡
The reduced data products comprise two datacubes for ev- [Nii]λ6583,[Oi]≡[Oi]λ6300,and[Sii]≡[Sii]λλ6716,6731
ery galaxy, one for each of the red and blue arms. The blue
datacube covers ≈3750–5800˚A with a spectral sampling of
1.04˚A and spectral resolution of R ≈ 1750 or full-width at
half-maximum(FWHM)of≈170kms−1.Thereddatacube
3 equallyspacedonalogarithmicscalebetween0.004and11.220
Gyr
2 http://www.aao.gov.au/science/software/2dfdr 4 Z=0.004,0.008,and0.019
MNRAS000,1–25(2015)
SAMI: extraplanar gas and galactic winds 5
log([NII]/Hα) log([SII]/Hα) log([OI]/Hα)
SDSS Pixel-to-pixel Pixel-to-pixel Pixel-to-pixel
15 -0.3 0.0 -0.8
-0.4 -0.1
-1.0
10 -0.5 -0.2
-0.6 -0.3 -1.2
sec] 5 -0.7 -0.4 -1.4
c
[ar 0 z -binning -0.8 z -binning -0.5 z -binning
c. | | -0.3 | | 0.0 | | -0.8
e
D 5
δ −5 -0.4 -0.1 -1.0
0
-0.5 -0.2
10 -1.2
− 5
− -0.6 -0.3
-1.4
15
− 15 10 5 0 -5 -10 -15 -0.7 -0.4
5 0 -5
δ R.A.[arcsec]
Figure 3. Examples of 2D line ratio maps reconstructed from the IFS datacubes of GAMAJ120221.91-012714.1 (CATAID: 185510).
The SAMI FOV and pointing position are shown as the red dashed circle in the SDSS 3-colour image (right panel). We create the
line ratio maps in two different methods: pixel-to-pixel (upper panels) and adaptive |z|-binning (lower panels; see Section 3.2). The
adaptive|z|-binningbinssymmetricallyoneithersideofthedisc.The|z|-binningmapstracetheemissionlineratiosoftheionizedgas
unambiguouslyouttolarger|z|thanthepixel-to-pixelmaps.WeadoptS/Ncutsof5inthesemaps.
3.2.1 Pixel-to-pixel maps there are no strong indications of on-going interactions in
GAMA J115927.23-010918.4 (z = 0.0202), the presence of
Figure 2 shows the emission line flux, line ratio, and kine-
nearby galaxy GAMA J115923.94-010915.4 (about 50 arc-
matic maps of GAMA J115927.23-010918.4 (CATAID5:
sec or 23 kpc away) at a similar redshift (z = 0.0204) im-
31452) and demonstrates the immediate products coming
plies that the environment might play a role in triggering
out of our lzifu pipeline. These maps are useful for inves-
the winds (e.g. Rich et al. 2010; Vogt et al. 2013). Higher
tigating the properties of the extraplanar gas, as we will
spatialresolutionandmulti-wavelengthobservationsarere-
elaborateoninlatersections.Inthisparticulargalaxy,sev-
quired to determine the precise geometry and cause of the
eral features are immediately obvious and imply the pres-
galacticwindsinGAMAJ115927.23-010918.4;nevertheless,
ence of large-scale galactic winds. The [Nii]/Hα, [Sii]/Hα
thepixel-to-pixelmapsreconstructedfromourIFSdataare
and [Oi]/Hα line ratio maps show that the line ratios in-
sufficient to reveal the existence of strong disc-halo interac-
crease as a function of height off the disc plane, and sug-
tions through galactic winds.
gest that the physical conditions of the ionized gas change
dramatically off the disc plane. The enhanced line ratios
are likely due to shocks embedded in galactic winds. The
gasinpartiallyionized,post-shockregionsemitsstrongfor- 3.2.2 Adaptive |z|-binning
bidden lines from low ionization species and hydrogen re-
combination lines (Allen et al. 2008). The elevated velocity GAMA J115927.23-010918.4 in Figure 2 is one of the best
dispersion seen in the σ map is also consistent with the casesinoursamplewhereclearwindsignaturesaredetected.
gas
broad kinematic component seen in IFS data of other wind However, since extraplanar line emission is typically much
galaxies(e.g.Richetal.2010,2011;Fogartyetal.2012;Ho fainter than that of the disc, we spatially bin the datacube
et al. 2014). Two additional interesting features stand out parallel to the disc plane (|z|-binning). This improves our
fromthesemaps.Inthe[Nii]/Hαmap,anelevated[Nii]/Hα detection limits at the cost of losing some spatial and kine-
cone-shapedregionpointingsouth-westcouldbetracingan matic information of the extraplanar gas.
ionization cone emerging from the energy injection zone at In the analysis we adopt“adaptive |z|-binning”to en-
the centre of the galaxy. We do not observe the other side sure adequate signal-to-noise ratio (S/N) in the following
of the putative bipolar structure presumably due to dust way.First,weproduceabinreferencemapwheretheSAMI
obscuration or intrinsic asymmetry of the winds. This fea- FOV is divided into 1-arcsec wide slices parallel to the disc
tureisverysimilartothebipolarstructuresthathavebeen (half the typical spatial resolution). We then bin together
observed in the nearby starburst galaxies NGC 253 (West- therequirednumberofslicessymmetricallyuntiltheS/Nof
moquetteetal.2011)andNGC3079(Ceciletal.2001).The [Sii]λ6716reaches5.Closetotheplaneofthedisc,theS/N
velocity map of the gas, v , shows that the overall veloc- may be already sufficient, leading to a bin containing only
gas
ity gradient is mis-aligned with the galaxy disc, which is the first 1-arcsec-wide slice above and below the disc plane.
caused by the complex velocity field of the bipolar winds Asfluxdropsoffatlargerdistancesfromthediscplane,we
superimposedonthediscrotation(seeSection4).Although adoptlargerbinstoobtaintherequiredS/N(i.e.containing
severalmatchedslicesfromaboveandbelowthedisc).This
symmetric binning assumes that the ionized gas above and
5 CATAIDisthegalaxyidentifierofGAMA below the disc share the same physical properties. Spaxels
MNRAS000,1–25(2015)
6 I.-T. Ho et al.
v asymmetry v asymmetry
gas gas
v v v v v v v v
gas gas,flipped gas− gas,flipped 60 gas gas,flipped gas− gas,flipped 60
ec] 5 30 ec] 5 30
[arcs 0 0 m/s] [arcs 0 0 m/s]
δDec.−5 −30[k δDec.−5 −30[k
60 60
5 0 -5 − 5 0 -5 −
δR.A.[arcsec] δR.A.[arcsec]
Figure 4. The velocity asymmetry across the major axis of Figure 5.SameasFigure4butforGAMAJ120221.91-012714.1
GAMAJ115927.23-010918.4 (CATAID: 31452 ; Figure 2). The (CATAID: 185510 ; Figure 3). The small residuals in the right
middle panel (vgas,flipped) is the velocity field (vgas) in the panelandthelineratiogradientsinFigure3implythepresence
left panel flipped over the galaxy major axis. The difference be- ofeDIG.
tween the left and middle panels is shown in the right panel
e(vqguaasll−ysvpgaacse,fdliipnpe1d5,komrst−he1vinetloercvitaylsr.eTsihdeudalasmhaedp)b.lCacokntlionuersinadrie- imately 10 to 30 km s−1 kpc−1 has been found in nearby
cates the major axis of the galaxy. The grey dashed circles indi- galaxies where high quality optical or Hi data are avail-
cateconcentricellipsesof1re (inner)and1r˜e (outer)centeredat able (e.g. Fraternali et al. 2005; Heald et al. 2006b,a, 2007;
thegalaxyopticalcentre.Theouterellipsesareapproximately1 Zschaechner et al. 2015). There is no strong evidence of
arcseclargerthantheinnerellipses.Thestrongresidualsoutside pronounced asymmetry in the vertical gradient, i.e. the gas
1r˜e areclearsignsofgalacticwinds. aboveandbelowthediscslowsdownequallywith|z|;how-
ever, data of high enough quality to constrain the levels of
asymmetryremainscarce.Inthecaseofstrongdisc-haloin-
insidethefinalbinsaresummedandthebinnedspectraare
teractionsthroughgalacticwinds,thebipolaroutflowscom-
re-analysed by lzifu.
monly present opposite velocity fields on either side of the
As a technical note, we take into account non-zero co-
disc. The flow axis is approximately perpendicular to the
variance between neighbouring spaxels while binning the
discplanebuteveninextremeedge-onsystemstheoutflow
datacubes. The cubing algorithm in our data reduction
velocity fields are never perfectly symmetric across the ma-
pipelineresamplestheobservedrowstackedspectraofmul-
joraxesofthehosts(e.g.Shopbell&Bland-Hawthorn1998;
tiple dithers on to a rectangular grid, which results in cor-
Cecil et al. 2001; Sharp & Bland-Hawthorn 2010; Westmo-
relatednoisebetweenneighbouringspaxels(seeSharpetal.
quette et al. 2011). The asymmetry of the velocity fields is
2015andAllenetal.2015).Theappropriatewaytopreserve
presumably due to a combination of the outflow axis not
the variance when binning the data is to bin the datacubes
lying exactly on the plane of the sky, the clumpy, dusty in-
and refit the binned spectra, rather than binning the pixel-
terstellar medium, the highly structural nature of the wind
to-pixel emission line maps.
filaments, and the wide opening angle of the bipolar cones
In Figure 3, we show examples of some line ratio maps
(typically 40 to 60 degrees).
fromfittingtheun-binneddatacubesandbinneddatacubes
Thus, to determine whether the extraplanar emissions
ofGAMAJ120221.91-012714.1(CATAID:185510).Spaxels
trace winds or eDIG, we quantify the velocity asymmetry
or spatial bins with S/N< 5 on any of the lines associated
across the galaxy major axis on the velocity map. We first
withtherelevantlineratiosaremaskedout.Figure3demon-
flip the pixel-to-pixel velocity map over the galaxy major
strates the advantage of the adaptive |z|-binning approach
axis,v ,andsubtracttheflippedmapfromtheorig-
overthetraditionalpixel-to-pixelapproach.The|z|-binning gas,flipped
inal map (Figures 4 and 5). We define an“asymmetry pa-
mapscantracetheionizedgasunambiguouslyouttolarger
rameter”,ξ,thatquantifiestheflatnessofv −v
|z|whilesinglespaxels atsuchlarge distancesmaynotsur- gas gas,flipped
(i.e. the velocity asymmetry map):
vive the S/N cuts and thus provide less of a constraint on
the physical conditions of the gas. ξ˜ +ξ˜
ξ= + −, (1)
2
where
4 GAS KINEMATICS (cid:16) v −v (cid:17)
ξ˜ = std gas gas,flipped . (2)
+/− (cid:112)
Whileopticallineratiosshedlightontheunderlyingexcita- r+/−>r˜e Err(vgas)2+Err(vgas,flipped)2
tionsourcesoftheextraplanargas,thevelocityandvelocity Here, Err(v ) is the 1σ map from lzifu, and
gas
dispersion of the gas provide additional constraints on the Err(v ) is the corresponding map for v .
gas,flipped gas,flipped
form and strength of disc-halo interactions. The ξ parameter measures the standard deviation of the
S/N of the velocity asymmetry map (i.e. residuals) using
only pixels outside the elliptical aperture of r˜ 6. This char-
e
4.1 Asymmetry of the extraplanar velocity field
acteristicradiusr˜ isther-bandeffectiveradiusincreasedby
e
We empirically quantify the strength of disc-halo interac-
tionsbytheregularityand(a)symmetryofthevelocityfield.
6 WedefinetheopticalcentreinourIFSdatausingthecontin-
Inthecasesofweak(orno)disc-halointeractions(e.g.eDIG
uumimagesummedoverthereddatacube.WefitaS´ersicprofile
with no gas flows), the extraplanar gas co-rotates with the usinggalfittotheIFScontinuumimagewhilefixingtheshape
disc and usually presents a slight lag in the azimuthal ve- parameterstothosedeterminedfromtheGAMAteamfittothe
locity. A vertical gradient in azimuthal velocity of approx- SDSSr-bandimage.
MNRAS000,1–25(2015)
SAMI: extraplanar gas and galactic winds 7
approximately 1 arcsec to reduce the effect of beam smear- etal.(2006b,a,2007)foundextraplanarσ (cid:46)50kms−1in
gas
ing.OnlyspaxelswithS/N(Hα)>5areconsidered,andall NGC 891, NGC 5775, and NGC 4302. These three edge-on
40ofourgalaxieshavemorethan150spaxelsoutsider˜ .We discshaveverysimilarionizedgasrotationvelocities(v ≈
e rot
adopt the mean value of the standard deviations on either 170–200 km s−1) and stellar masses (log(M∗/M(cid:12))≈10.7).
side of the disc (i.e. ξ˜ and ξ˜ ) as ξ7. As a technical note, Their findings imply that σ /v (cid:46) 0.3 for eDIG. In the
+ − gas rot
to obtain the flipped maps, we first reflect the pixel centres case of strong interactions, the velocity dispersion of the
overthegalaxymajoraxisandthenlinearlyinterpolatethe extraplanar gas is broadened by both the turbulent mo-
v and Err(v ) maps around the reflected positions. tionoftheoutflowinggasandlinesplittingcausedbyemis-
gas gas
A perfectly symmetric velocity field would yield ξ of sionsfromtheapproachingandrecedingsidesoftheoutflow
roughly unity as a result of random noise fluctuation in the cones. The magnitude of the line splitting depends on out-
velocity asymmetry map. The asymmetry parameter ξ be- flow velocities, topology and inclination angle. In classical
comesmuchlargerthanunitywhenthevelocityasymmetry wind galaxies, line splitting of about 100 km s−1 is com-
map is not flat due to asymmetry in the velocity field. mon (e.g. Heckman et al. 1990; Sharp & Bland-Hawthorn
Figure 4 illustrates our methodology applied to the 2010). In the extreme case of M82, line splitting as high as
galaxy in Figure 2, which has the largest ξ in our sam- 300 km s−1can be easily resolved by optical spectrographs
ple (ξ = 11.9). The strong residuals outside 1r˜ are clear (Axon&Taylor1978;Shopbell&Bland-Hawthorn1998).A
e
evidence of galactic winds, consistent with the line ratio, verybroadcomponent(FWHMof300kms−1)fromthehalo
structural, and kinematic signatures seen in Figure 2. The gashasalsobeenobservedinM82(Bland&Tully1988).In
flat residuals inside 1r˜ indicate the good symmetry of the less extreme winds, the line splitting could be too narrow
e
prominent disc rotation. Figure 5 shows the same applica- for SAMI to resolve, especially if viewing the outflow cones
tiononthegalaxyinFigure3.Nostrongresidualsarevisible from the side where the projected velocities are small. In
onthevelocityasymmetrymap(v −v ;ξ=1.3) fact, we do not see clear indications of line splitting in the
gas gas,flipped
despite the clear line ratio gradients seen in Figure 3, im- majority of our wind candidates given the S/N of our data.
plyingthepresenceofeDIGratherthanshockedoutflowing Nevertheless, the increase in velocity dispersion as a result
gas.Thesesetsoffiguresdemonstratetheimportanceofus- of unresolved line splitting is still strong evidence of the
ing both kinematics and line ratios to investigate galactic presence of outflows.
winds and eDIG . To quantify the extraplanar velocity dispersion, we in-
TheparameterξcanbeusedtotracewindsoreDIGun- voke a simple parameter η , the“velocity dispersion to ro-
50
der the assumptions that our edge-on discs are regular and tation ratio”parameter, defined as
not heavily warped. Mergers are excluded from our sam-
η =σ /v , (3)
ple(Section2),andallourgalaxiespresentregularrotation 50 50 rot
on their discs, consistent with not undergoing major inter- where σ is the median velocity dispersion of all spaxels
50
actions (see below Figures 7 and A1). The SDSS images outside 1r˜ with S/N(Hα) > 5. As before, all 40 of our
e
also suggest that all the galaxies have regular discs without galaxieshavemorethan150spaxelsmeetingourS/Ncrite-
strong warp signatures. Although warping of stellar discs is rion outside 1r˜ . The rotation velocity is measured on our
e
common, the warps are only important at large radii (3–6 velocity map from the spaxels along the optical major axis
times the disc scalelength or 2–4r ; Saha et al. 2009), typi- with the maximum velocity. This method is applied only
e
callybeyondtheSAMIFOVforthemajorityofoursample when our FOV extends to 1.4r (< R ≈ 7.5 arcsec).
e SAMI
(90%). Thus, our velocity asymmetry parameter predomi- For an exponential disc, this radial distance corresponds to
natelyreflectshowtheextraplanargasiseffectedbywinds. about 2.4 times the disc scale length at which typical rota-
It is worth pointing out that our approach of quantify- tion curves already reach their maximum velocities (Sofue
ing velocity asymmetry does not depend on any kinematic &Rubin2001;Ceciletal.2015).Forgalaxieswithoutsuffi-
modeling,whichcanbenon-trivialduetotheedge-onview- cientlylargespatialcoverage,weusethestellarmassTully-
ingangle.Futureworkshoulddevelop3-dimensionalmodels Fisherrelationtoinferv (Bell&deJong2001).Wenote
rot
that include the effects of beam smearing (such as the ap- thatforourpurposev shouldideallybethemaximumro-
rot
proach by Jo´zsa et al. 2007 and Di Teodoro & Fraternali tationvelocityoftheionizedgas,andweapproximateitus-
2015) and radiative transfer to reproduce the emission line ing the Tully-Fisher relation. The stellar mass Tully-Fisher
data cubes. relation unfortunately exhibits a scatter of about 0.5 dex,
which directly translates to a factor of 3 in the systematic
uncertainty of η .
50
4.2 Elevated extraplanar velocity dispersion
We also make use of the velocity dispersion of the extra- 4.3 An empirical identification of wind-dominated
planar gas to assess the form and strength of disc-halo in- galaxies
teractions. In the case of weak (or no) interactions, the ve-
Based on the two different empirical parameters for quan-
locity dispersion of eDIG arises predominately from line-of-
tifying the strength of disc-halo interactions, it should be
sight projection of the gas corotation with the disc. Heald
possible to distinguish galactic winds from eDIG. Galactic
winds are expected to show both high ξ and η , whereas
50
7 We note that ξ˜+ and ξ˜− in principle contain redundant infor- eDIGhaslowξandη50.Thatis,galacticwindsbothdisturb
mationbutduetointerpolationandfinitegridsizethetwovalues the symmetry of the extraplanar velocity field and increase
are not exactly the same. We take the mean value to minimise the extraplanar emission line widths, whereas eDIG should
suchanartefact. still follow the velocity field of the galaxy. If winds are the
MNRAS000,1–25(2015)
8 I.-T. Ho et al.
r /R r /R r /R
e SAMI e SAMI e SAMI
0.40 0.48 0.56 0.64 0.8 1.0 1.2 1.4 0.6 0.9 1.2 1.5
vrot measured vrot fromT-F vrot measured
ξ[]101 vrot fromT-F
er
t
e
m
a
ar
p
y
tr
e
m
m
y
s
A
100
0.2 0.3 0.4 0.5 0.6 0.2 0.3 0.4 0.5 0.6 0.2 0.3 0.4 0.5 0.6
Velocitydispersionto Velocitydispersionto Velocitydispersionto
rotationratio[η ] rotationratio[η ] rotationratio[η ]
50 50 50
Figure 6.Theasymmetryparameterofthevelocityfieldξversusthevelocitydispersiontorotationratioparameterη50 ofoursample.
The left panel presents galaxies with vrot measured from the SAMI data, the middle panel presents galaxies with vrot inferred from
the Tully-Fisher relation, and the right panel shows all the galaxies. The three high-ξ low-η50 galaxies marked by the grey crosses are
ussedinSection4.3.TheredcrossintherightpanelcorrespondstotheexamplegalaxyshowninFigures2and4;andthebluecrossin
Figures3and5.Galaxiesfallingintheshadedregionintherightpanelareidentifiedaswind-dominatedgalaxies(seeFigure7).
onlymechanismdisruptingtheextraplanargas,thenatrend with similar kinematic properties of their extraplanar gas;
between ξ and η is expected. Figure 6 shows our sample instead,thedistributioncouldimplyacontinuous“strength
50
in the ξ versus η parameter space. We present separately sequence”of disc-halo interaction, connecting the very qui-
50
galaxies with v measured (left panel) and v inferred escent, non-interacting (or weakly interacting) eDIG to the
rot rot
from the Tully-Fisher relation (middle panel). very violent,stronglyinteractingclassicalgalacticwinds.It
is likely that in galaxies not classified as wind-dominated,
Forgalaxieswithdirectlymeasuredv ,itisinconclu-
rot therecouldstillbelocalizedfeedbackfromHiiregionsdriv-
sive whether there is a trend between ξ and η . There are
50
no galaxies in the lower right corner (η (cid:38)0.4 and ξ (cid:46)3), ing gas away from the discs, disturbing the extraplanar gas
50
kinematics at small scales, and causing a slight increase in
however,therearethreegalaxies(markedwithcrosses)with
ξ and/or η . In fact, some galaxies (e.g. CATAID: 496966,
larger ξ than the other galaxies of similar η . The Spear- 50
50
228066,348116,517594,595027;FigureA1)indeedshowin-
man rank correlation test implies no significant correlation
dicationsoflocalisedvelocityasymmetrypossiblydrivenby
for the galaxies in the left panel (ρ of 0.21 and p-value of
off-centre starbursts, consistent with the galactic chimney
0.46), unless the three high-ξ low-η galaxies are excluded
50
picture (Norman & Ikeuchi 1989). It is also likely that in
from the test (ρ of 0.80 and p-value of 0.0016). We specu-
some of the wind-dominated galaxies, a portion of the pre-
late that there might be additional mechanisms responsible
existing eDIG remains co-rotating with the discs while the
for causing the asymmetry of the extraplanar gas (see Sec-
outflowinggasinteractswiththerestoftheeDIG.Givenour
tion 7.3). A trend between ξ and η can not be confirmed
50
sample size, it is nonetheless convenient to invoke a binary
by our medium-sized sample.
classification scheme in order to investigate the underlying
Forgalaxieswithv inferredfromtheTully-Fisherre-
rot physicalprocessesgoverningthestrengthofdisc-halointer-
lation, there is also no clear relationship between ξ and η
50 actions.
and the Spearman rank correlation test suggests no signifi-
cantcorrelation(ρof−0.19andp-valueof0.35).Thelackof
The wind-dominated galaxies in Figure 7, by selec-
correlationisexpectedgiventhatthescatterofTully-Fisher
tion, show strong velocity asymmetry and elevated extra-
relation (0.5 dex) yields a factor of 3 in the systematic un-
planarvelocitydispersion.Classicalgalacticwindsignatures
certainty of η .
50 can be easily identified in these maps. Some galaxies (e.g.
Weconservativelyselectgalaxieswhoseextraplanargas CATAID:228432,574200)presentionizationconesignatures
is affected by galactic winds from the upper right corner of ontheirvelocitydispersionmapsandtheirvelocityasymme-
Figure 6. We select galaxies with η >0.3 and ξ >1.8, as trymaps,withstrongvelocityresidualstracingthelimbsof
50
definedbytheshadedregionintherightpanel.Thesecrite- the putative bipolar cones. In others (e.g. CATAID: 31452,
riayield15galaxies.Forthepurposeofpresentation,wewill 239249,417678,106389),thevelocityasymmetrymapscor-
henceforth refer these galaxies as “wind-dominated galax- relate spatially with the velocity dispersion maps and usu-
ies”. Figure 7 presents their velocity, velocity asymmetry, ally present positive/negative residuals on either sides of
velocity dispersion, and [Sii]/Hα line ratio maps. The re- thediscs.Onthecontrary,theremaininggalaxiespresented
maininggalaxiesarepresentedintheappendix(FigureA1). in the appendix do not show these kinematic signatures of
We note that it is, to some extent, unphysical to impose galacticwindsasclearlyasthewind-dominatedgalaxies.In
hard limits to identify wind galaxies. The distribution of all the wind-dominated galaxies, the [Sii]/Hα line ratio in-
our galaxies in the ξ versus η parameter space does not creases with |z|. This is partly due to excitation by shocks
50
favour the idea that there exists distinct groups of galaxies embedded in the outflows (see Section 6.2).
MNRAS000,1–25(2015)
SAMI: extraplanar gas and galactic winds 9
v asymmetry
gas
v v - v σ
log([SII]/Hα) log([SII]/Hα) gas gas gas,flipped gas
SDSS SAMI Hα Pixel-to-Pixel z-binning [km/s] [km/s] [km/s]
| |
543769 -0.48 50 ξ:2.57 30 η50:0.31
25 15 60
-0.54
0 0 40
-0.60
-0.66 −25 −15 20
-0.72 −50 −30 0
93167 -0.16 40 ξ:2.36 30 η50:0.44
-0.24 20 15 60
-0.32 0 0 40
--00..4480 −20 −3105 20
24433 -0.16 −5040 ξ:2.15 −30 η50:0.31 0
-0.24 25 15 60
-0.32
0 0 40
--00..4480 −5205 −3105 20
− − 0
567624 -0.30 2550 ξ:4.11 1350 η50:0.32 60
-0.35
0 0 40
-0.40
-0.45 −−5205 −−3105 020
574200 -0.30 60 ξ:5.31 30 η50:0.40
-0.36 30 15 60
-0.42 0 0 40
-0.48 −30 −15 20
-0.54 60 30
− − 0
228432 -0.10 60 ξ:4.09 30 η50:0.43
-0.20 30 15 60
-0.30 0 0 40
-0.40 −30 −15 20
-0.50 −60 −30 0
239249 -0.12 60 ξ:1.96 30 η50:0.33
-0.18 30 15 60
-0.24 0 0 40
-0.30 −30 −15 20
[arcsec] 150 31452 ----0000....32106468 −0135060 ξ:11.88 −0135030 η50:0.49 04600
ec. -0.32 −15 −15 20
D 15 -0.40 −30 −30 0
δ− 15 0 -15 5 0 -5
δ R.A.[arcsec]
Figure 7.Fromlefttoright:SDSS3-colour,SAMI3-colour(red:6300˚A;green:4800˚A;blue:4000˚A),Hα,[Sii]/Hα(pixel-to-pixeland
adaptive|z|-binning),velocity,velocityasymmetry,andvelocitydispersionmapsofourwind-dominatedgalaxies(η50>0.3andξ>1.8).
The adaptive |z|-binning maps present the average [Sii]/Hα measured in different bins (Section 3.2.2). GAMA CATAIDs and the η50
and ξ values are labelled in the plots. The velocity and velocity dispersion contours are equally spaced in 20km s−1 intervals. The red
dashedcirclesontheSDSSimagesindicateapproximatelytheFOVandpointingsoftheSAMIhexabundles.Thedashedlinesindicate
thegalaxymajoraxes.Theinnerandouterconcentricellipsescorrespondtoellipticalaperturesof1re and1r˜e,respectively.Theouter
ellipsesareapproximately1arcseclargerthantheinnerellipses.
5 HOST GALAXY PROPERTIES 5.1 Star formation rate
To estimate the SFR, we adopt the SED fitting code mag-
physdescribedindetailindaCunhaetal.(2008).magphys
We compute the host galaxy properties before returning to is aBayesian-basedpackage thatusesobservedphotometry
investigate the underlying processes driving the different to constrainstellar population synthesis models by Bruzual
properties of the extraplanar gas. The host galaxy prop- & Charlot (2003) and dust emission of different tempera-
erties we investigate are stellar mass, star formation rate, turesfrommolecularcloudsandthediffuseISM.TheUV-to-
star formation history, and their associated quantities. As IRSEDsassumeexponentialstarformationhistoriesofdif-
before, we adopt the photometrically-derived stellar mass ferentstarformationtime-scaleswithrandomburstsofstar
from the GAMA survey (Taylor et al. 2011). The following formationsuperimposed.Physicalparametersofthemodels
two subsections describe how we measure SFRs using spec- (SFR,starformationhistory,stellarmass,dusttemperature,
tralenergydistribution(SED)fittingandHαflux,andstar etc.) and corresponding errors are constrained by compar-
formation histories using the D (4000) and Hδ indices. ing synthesized and observed photometry under a Bayesian
n A
MNRAS000,1–25(2015)
10 I.-T. Ho et al.
v asymmetry
gas
v v - v σ
log([SII]/Hα) log([SII]/Hα) gas gas gas,flipped gas
SDSS SAMI Hα Pixel-to-Pixel z-binning [km/s] [km/s] [km/s]
238125 | | 00..0100 2550 ξ:1.88 1350 η50:0.41 60
-0.10
0 0 40
-0.20
--00..4300 −−5205 −−3105 020
106616 -0.08 50 ξ:4.58 30 η50:0.35
-0.16 25 15 60
-0.24 0 0 40
--00..4302 −5205 −3105 20
0.00 − − 0
486834 -0.08 80 ξ:2.51 30 η50:0.34
-0.16 40 15 60
-0.24 0 0 40
-0.32 −40 −15 20
-0.40 −80 −30 0
417678 -0.10 80 ξ:3.48 30 η50:0.56
-0.20 40 15 60
-0.30 0 0 40
-0.40 −40 −15 20
80 30
-0.50 −150 − 0
106389 -0.10 100 ξ:5.03 30 η50:0.32
-0.20 50 15 60
-0.30 0 0 40
--00..5400 −−11550000 −−3105 020
593680 0.10 −100 ξ:2.67 30 η50:0.40
-00.0.100 50 15 60
-0.20 0 0 40
-0.30 −50 −15 20
[arcsec] 150 618906 ---000...432024 −061021000 ξ:2.45 −0135030 η50:0.31 04600
ec. -0.40 −60 −15 20
D 15 -0.48 −120 −30 0
δ− 15 0 -15 5 0 -5
δ R.A.[arcsec]
Figure 7–continued
framework. We utilize photometry from the Galaxy Evo- tometryfromtheHerschel-ATLASsurvey.Ofthe12galaxies
lution Explorer (GALEX; FUV, NUV; Martin et al. 2005; without Herschel photometry, two of them fall outside the
Wyder et al. 2005), SDSS (u,g,r,i,z; Adelman-McCarthy Herschel-ATLAS SPIRE footprints and the remainder falls
et al. 2008), UKIRT Infrared Deep Sky Survey (UKIDSS) below the detection limits at 250µm.
Large Area Survey (Y, J, H, K; Lawrence et al. 2007) and In addition to the SED-based SFRs, we also calculate
HerschelSpaceObservatory(PhotodetectingArrayCamera SFRs based on the Hα line measured in our integral field
and Spectrometer: 100µm and 160µm; Spectral and Pho- data.Weconstructglobal,one-dimensionalspectrabysum-
tometric Imaging Receiver or SPIRE: 250µm, 350µm, and ming all spaxels in the datacubes, and fit the spectra using
500µm; Pilbratt et al. 2010; Poglitsch et al. 2010; Griffin thesamemethoddescribedinSection3.2.TheHαfluxesare
etal.2010).TheSDSSandUKIDSSphotometryarer-band extinctioncorrectedusingtheBalmerdecrementmethodas-
aperture-matchedbytheGAMAteam,andarecompiledin suminganintrinsicHαtoHβ lineratioof2.86undercase-B
their data release 2 (Hill etal. 2011; Liske et al. 2015).The recombination of T of 10,000 K and n of 100 cm−3 (Os-
e e
Herschel photometry comes from the Herschel-ATLAS sur- terbrock & Ferland 2006). We adopt the extinction law by
vey (Eales et al. 2010) Phase 1 Internal Data Release. The Cardellietal.(1989)ofR of3.1.Theextinctioncorrected
V
maps and catalogues will be described in Valiante et al (in Hα line fluxes are then converted to SFRs following Mur-
preparation) and have been processed in a similar way to phy et al. (2012). We note that a Chabrier (2003) initial
that described by Ibar et al. (2010), Pascale et al. (2011) mass function (IMF) is assumed in the Bruzual & Charlot
and Rigby et al. (2011). The matching to optical data was (2003)modelsusedinmagphys,andaKroupa(2001)IMF
performedinasimilarwaytothatdescribedinSmithetal. is assumed for the Hα SFRs; however the systematic errors
(2011), and will be presented in Bourne et al. (in prepa- duetoslightdifferencesbetweentheIMFsaresmall(within
ration). The GALEX, SDSS, and UKIDSS photometry are 10%; Bell et al. 2007).
available for almost all galaxies: 38/40, 40/40, and 39/40, Figure 8 compares the SED- and Hα-based SFRs. The
respectively.Morethanhalfofoursample(28/40)haspho- presenceofapositivecorrelationwithoutanappreciableoff-
MNRAS000,1–25(2015)