Stellar Populations of Lyman-alpha Emitters at z=4.86: A Comparison to z∼5 LBGs
Suraphong Yuma, Kouji Ohta, Kiyoto Yabe, Kazuhiro Shimasaku, Makiko Yoshida, Masami Ouchi, Ikuru Iwata, Marcin Sawicki
aa r X i v : . [ a s t r o - ph . C O ] J u l A CCEPTED FOR PUBLICATION IN A P J Preprint typeset using L A TEX style emulateapj v. 5/25/10
STELLAR POPULATIONS OF LYMAN-ALPHA EMITTERS AT Z = 4 . Z ∼ S URAPHONG Y UMA , K OUJI O HTA , K IYOTO Y ABE , K AZUHIRO S HIMASAKU , M AKIKO Y OSHIDA ,M ASAMI O UCHI , I KURU I WATA , AND M ARCIN S AWICKI Accepted for publication in ApJ
ABSTRACTWe present a study of stellar population of Lyman Alpha Emitters (LAEs) at z = 4 .
86 in the Great Obser-vatories Origins Deep Survey (GOODS) North field and its flanking field. The LAEs are selected based onoptical narrowband (NB711) and broadband ( V , I c , and z ′ ) observations by Suprime-Cam attached on SubaruTelescope. With the publicly available IRAC data in GOODS-N and further IRAC observations in the flankingfields, we select five LAEs which are not contaminated by neighboring objects in IRAC images and constructtheir observed spectral energy distributions (SEDs) with I c , z ′ , IRAC 3.6 µ m, and 4.5 µ m band photometry. TheSEDs cover the rest-frame UV to optical wavelengths. We derive stellar masses, ages, color excesses, and starformation rates of five LAEs using SED fitting method. Assuming the constant star formation history, we findthat the stellar masses range from 10 to 10 M ⊙ with the median value of 2 . × M ⊙ . The derived agesrange from very young ages (7.4 Myr) to 437 Myr with a median age of 25 Myr. The color excess E ( B - V )are between 0 . - . - M ⊙ yr - . A comparison of the stellar populationsis made between three LAEs and 88 LBGs selected at the same redshift, in the same observed field, and downto the same limit of the rest-frame UV luminosity. These three LAEs are the brightest and reddest samplesamong the whole LAE samples at z = 4 .
86. The LAEs distribute at the relatively faint part of UV-luminositydistribution of LBGs. Deriving the stellar properties of the LBGs by fitting their SEDs with the same modelensures that model difference does not affect the comparison. It is found that the stellar properties of the LAEslie on distributions of those of LBGs. On average, the LAEs show less dust extinction, and lower star formationrates than LBGs, while the stellar mass of LAEs nearly lies in the middle part of the mass distribution of LBGs.However, the stellar properties of LAEs and LBGs are similar at the fixed UV or optical luminosity. We alsoexamine the relations between the output properties from the SED fitting and the rest-frame Ly α equivalentwidth, but cannot find any significant correlation. Subject headings: galaxies: evolution — galaxies: formation — galaxies: high-redshift INTRODUCTION
There are two popular techniques for isolating galaxiesat high redshift ( z & α emission lines which fall [email protected] Based on data collected at Subaru Telescope, which is operated by theNational Astronomical Observatory of Japan. Department of Astronomy, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan Department of Astronomy, School of Science, University of Tokyo,Tokyo 113-0033, Japan Observatories of the Carnegie Institution of Washington, 813 SantaBarbara St., Pasadena, CA 91101 Okayama Astrophysical Observatory, National Astronomical Obser-vatory of Japan, Okayama 719-0232, Japan Department of Astronomy and Physics, Saint Mary’s University, 923Robie Street, Halifax, Nova Scotia, B3H 3C3, Canada into a narrowband filter. This method is useful for select-ing high-redshift galaxies that have a strong Ly α emissionline. These objects are called Lyman Alpha Emitters (here-after LAEs). Partridge & Peebles (1967) proposed that pri-mordial galaxies in the early stage of their formation shouldshow a strong Ly α emission line. LAEs are thus expected tobe young galaxies with low metallicity. Many surveys havebeen made to seach for galaxies with a strong Ly α emis-sion at various redshifts ranging from 2.1 to 6.6 or even more(e.g., Hu & McMahon 1996; Cowie & Hu 1998; Hu et al.1998, 2004; Rhoads et al. 2000, 2004; Kudritzki et al.2000; Rhoads & Malhotra 2001; Malhotra & Rhoads 2002;Fynbo et al. 2001; Ouchi et al. 2003, 2008; Fujita et al.2003; Shimasaku et al. 2003, 2004, 2006; Kodaira et al.2003; Ajiki et al. 2003, 2004, 2006; Taniguchi et al. 2005;Venemans et al. 2002, 2004; Iye et al. 2006; Ota et al.2007; Nilsson et al. 2009; Guaita et al. 2009).In order to examine the evolutionary stage of LAEsand to reveal what kind of galaxy an LAE is, variousproperties of LAEs have been studied such as luminosityfunctions (e.g., Ouchi et al. 2003, 2008; Malhotra & Rhoads2004; Shimasaku et al. 2006; Kashikawa et al. 2006;Gronwall et al. 2007; Ota et al. 2007) or clustering proper-ties (e.g., Ouchi et al. 2003; Shimasaku et al. 2003, 2004,2006; Kashikawa et al. 2006; Murayama et al. 2007).Revealing the stellar population of LAEs is one of crucialstudies for understanding their physical nature. In order todo that, spectral energy distribution (SED) of a galaxy is Yuma et al.compared with stellar population synthesis models producedby varying the ages, metallicities, amounts of dust extinc-tions, SFRs, etc. The stellar population of a galaxy can beconstrained from the best fit model. This method is knownas SED fitting method. Recent studies show that LAEs arein wide ranges of ages (1Myr – 1Gyr) and stellar masses(10 - M ⊙ ) (e.g., Gawiser et al. 2006, 2007; Lai et al.2007, 2008; Nilsson et al. 2007; Finkelstein et al. 2007,2008, 2009; Pirzkal et al. 2007). Among these studies,stacking analysis shows that LAEs at z ∼ ≤ z ∼ α emission line, while young sub-sample do not. In other words, they found that LBGs withLy α emission line are older than those without Ly α emissionline. Confirming this statement, the recent work with a largersample at the same redshift by Kornei et al. (2009) showsthat objects with rest-frame Ly α equivalent width larger than20 Å seem to be older, lower in star formation rate, and lessdusty than those without Ly α emission line. Pentericci et al.(2007), in contrast, found that at z ∼ α emis-sion are less massive and younger than those with no line.By considering the results from the SED analysis mentionedabove, there may be an evolution of difference between LAEsand LBGs with redshifts (e.g., Shimizu & Umemura 2010).Although there are some SED studies at higher redshift ( z & z = 4 .
86 from theirrest-frame UV-to-optical spectral energy distributions. Downto the same UV luminosity, the derived stellar properties ofLAEs are compared to those of LBGs by Yabe et al. (2009).They selected LBGs at the same redshift range in the samefield and they used the same fitting code and synthesis spectralmodel assumptions. This direct comparison between LAEsand LBGs is expected to reveal their difference and to knowwhat kind of properties make a galaxy an LAE.Data sources and photometry are described in section2. Section 3 explains the selection criteria, and the LAEcandidates. The observed SEDs are constructed for eachLAE candidate and compared to the models in section 5. Thefitting results are shown in Section 6. Section 7 is for com-parisons between LAEs and LBGs. The summary is givenin section 8. Through out this paper, we use AB magnitudesystem (Oke & Gunn 1983) and adopt a cosmology withparameters of Ω m = 0 . Ω Λ = 0 .
7, and H = 70 km s - Mpc - . DATA SOURCES AND PHOTOMETRY
Optical data
Observations and data reduction D e c li na t i on Right AscensionHDF-NIRAC: GOODS-NSuprime-CamGOODS-FF
Figure 1.
Observed fields. The solid line indicates the effective field ob-served with Suprime-Cam, whereas the long and short dashed lines show theGOODS-N and GOODS-FF fields observed with IRAC. Hubble Deep FieldNorth (HDF-N) is shown as a reference at the center of the figure.
Optical data were obtained with Suprime-cam(Miyazaki et al. 2002) attached on the Subaru telescope(Iye et al. 2004). The observed field is toward the Hub-ble Deep Field-North (HDF-N; Williams et al. 1996)[RA(2000)=12 h m . s
4, Dec(2000)=+62 ◦ ′ ′′ ] and isillustrated by solid line in Figure 1. Pixel scale of the CCDwas 0 . ′′
20. We used NB
711 [ λ e f f = 7126, FWHM = 73]and three broadband filters: V , I c , and z ′ to select z = 4 . NB
711 is 4 . ≤ z ≤ .
89. Transmission curves, whichinclude the responses of CCD, prime focus corrector (PFC),mirror, and airmass, of all filters used in this work are shownin Figure 2. The observation with the NB711 filter wasmade on March 16, 2005. Images were taken with ditheringof ∼ ′′ and the exposure time of 1200 seconds for eachframe. With 23 exposures, we covered ∼
750 arcmin fieldof view. The total integration time was 7.7 hours. Theweather condition was not so excellent. Seeing size duringthe observation was averagely 1 . ′′ -
170 non-saturated stars commonin several object frames. The frame alignment and correctionfor flux/count and full width at half maximum (FWHM) weremade based on these star data. A resulting FWHM of pointsources in the mosaicked image is ∼ . ′′ ∼ , . ′′ rms over the image. The magnitude zero pointwas derived based on the imaging data of two spectrophoto-metric standard stars (G191-B2B and HZ44) taken during thesame observing night. As an independent check for photome-try in NB
711 images, we derived NB
711 magnitudes of starsby interpolating their V , I c , and z ′ magnitudes. The derivedmagnitude zero points are in agreement within ∼ .
18 mag.A 3 σ limiting magnitude at 2 . ′′ NB
711 image.The observations and data reductions of broadband imagesare described in detail by Iwata et al. (2007). The imagestellar populations of LAEs at z = 4 .
86 3 E ff e c t i v e T r an s m i ss i on f n ( m Jy ) Observed wavelength ( m m) Figure 2.
Transmission curves of V , NB711, I c , and z ′ bandpasses. CCD re-sponse, transmission of PFC, mirror reflectivity, and airmass (sec z = 1.2) areincluded. Also shown is a model spectrum in the unit of µ Jy (solid line) of astar forming galaxy at z = 4 .
86 including Ly α emission line with a rest-frameequivalent width of 10Å. The model is constructed by Bruzual & Charlot(2003) stellar synthesis code. A constant star formation history and SalpeterIMF are assumed at an age of 12 Myr and at star formation rate of 1 M ⊙ yr - .The attenuation by intergalactic medium is applied with the prescription byMadau (1995). were taken with V , I c , and z ′ filters in February 2001. The typ-ical seeing during the observations was ∼ . ′′
1. In addition,Iwata et al. (2007) also used the imaging data taken fromFebruary to April 2001 by the University of Hawaii (UH)group (Capak et al. 2004), who used the same instrument andfilters. The final effective survey area is 508.5 arcmin . Thelimiting magnitudes of V , I c , and z ′ -band images are 28.1,26.8, and 26.6 mag, respectively (3 σ at 1 . ′′ . ′′ NB
711 image. The 3 σ limit-ing magnitudes of final PSF matched images at 2 . ′′ diameteraperture are 27.3, 26.1, and 25.9 mag for V , I c , and z ′ images,respectively. Photometric catalog
The photometric catalogs of the optical images were madeby using SExtractor version 2.5.2 (Bertin & Arnouts 1996).The positions of objects were extracted from the NB
711 im-age and the photometry was then made at the extracted posi-tions in all optical images via dual-image mode. In order touse the dual-image mode, we first have to make a positionregistration between the broadband images and the NB geomap in IRAF based on positions of the bright butnon-saturated stars detected in the images. Then the registra-tion of the images was performed by gregister . The accuracyin making a position registration is within 0.2 pixels or 0. ′′ GLOBAL back-ground from each registered image individually and homog- IRAF is distributed by the National Optical Astronomy Observatories,which are operated by the Association of Universities for Research in As-tronomy, Inc., under cooperative agreement with the National Science Foun-dation. Based on our simulation with artificial objects, the processes of subtract-ing background before smoothing the images and re-subtracting the
LOCAL background in SExtractor give the better photometry. enized the broadband images so that they all have the sameseeing size as the NB
711 image within 0. ′′
01 accuracy by gauss . The dual-image mode was then performed by usingthe following parameters. The images were filtered with adefault convolution kernel (default.conv). The
LOCAL back-ground was estimated using 64 ×
64 pixel background meshwith 3 × . σ minimum thresh-old. These parameters are found to maximize the number ofdetected objects in the NB
711 image and minimize the de-tected numbers of objects in the negative NB
711 images.In SED fitting process, the total photometry of the objectsis necessary. The total magnitudes were obtained by applyingaperture corrections to 2 . ′′ . ′′
5, which maximizes the signal-to-noise ratio (S/N). Theaperture corrections were determined from the Monte Carlosimulations where artificial objects with PSF shape corre-sponding to the seeing size of 1 . ′′ - . ± .
01 and - . ± . I c and z ′ images, respectively, where the errors arethe PSF uncertainties. Errors of the total magnitudes are thecombination of 1 σ error in making photometry of simulatedobjects and the uncertainties of correction factors. Becausewe do not use V and NB
711 images in SED fitting process,we do not compute the aperture corrections for them.
Mid-infrared data
Mid-infrared images are obtained from deep observationswith Infrared Array Camera (IRAC) on Spitzer Space Tele-scope (SST). We used the publicly available mid-infrareddata in the Great Observatories Origins Deep Survey Northfield (Dickinson et al. 2003, hereafter GOODS-N) centeredat 12 h m s
87, +62 ◦ ′ . ′′ . The field coversan area of approximately 10 ′ × ′ or ∼
160 arcmin as shownin Figure 1. In this work, we used the First Data Release(DR1) and Second Data Release (DR2) of IRAC data consist-ing of imaging data in 3.6, 4.5, 5.8, and 8.0 µ m bandpasses.Pixel scale of all images after drizzled is 0 . ′′
60 pixel - . MeanFWHM of PSF in IRAC GOODS-N images is ∼ ′′
7. 3 σ lim-iting magnitudes of IRAC 3.6, 4.5, 5.8, and 8.0 µ m images at2 . ′′ http: //ssc.spitzer.caltech.edu/legacy/goodshistory.html Yuma et al.and combined together to get an improved mosaic image cov-ering the whole observed area of ∼
300 arcmin . The mosaicimage has a final pixel scale of 0 . ′′
61 pixel - and a PSF size(FWHM) of 1. ′′
7. The 3 σ magnitude limits at 2 . ′′ µ m bands, respectively. Combining the infrareddata obtained in GOODS-N and GOODS-FF fields, we havethe effective area of ∼
400 arcmin covering ∼
80% of Subaruimages as seen in Figure 1.The photometry of IRAC was made based on the positionsof LAEs at z = 4 .
86, which are selected from the criteria de-scribed in the next section. Since the dimensions and the pixelscale of the narrow-band image and IRAC images are differ-ent, we could not use dual-image mode in SExtractor to de-termine aperture photometry in IRAC images.
Phot providedby IRAF was used instead to make aperture photometry at2 . ′′ NB
711 image are used as the reference, we made a registra-tion of IRAC images before making IRAC photometry of theobjects. The position alignment was made globally for eachIRAC image by using positions of point sources detected inboth NB
711 and IRAC images. The estimated errors in de-termining the position shifts are 0. ′′ ′′ ′′ ′′ - .
62 mag and - .
72 mag in 3.6 µ m and 4.5 µ m bandsrespectively. The correction uncertainties due to the uncer-tainties in PSFs are 5% and 6% in 3.6 µ m and 4.5 µ m bandsrespectively. For GOODS-FF images, they are respectively - .
70 mag and - .
73 mag with 6% and 7% uncertainties in3.6 µ m and 4.5 µ m bands. Errors in IRAC photometry are 1 σ standard deviations of sky background and the uncertaintiesof correction factors. Since 5.8 µ m and 8.0 µ m images showtoo low signal-to-noise (S/N) ratio to place a useful upperlimit in SED fitting process, we do not use them in furtheranalysis. SAMPLE SELECTION NB < . . σ limiting magnitude). Figure 3 shows I c - NB
711 colorversus NB711 magnitudes of all objects (black dots). We se-lected NB - excess objects from the following selection cri-teria: I c - NB > .
64 and (1) I c - NB > σ I c - NB , (2)where 3 σ I c - NB is a 3 σ error in measuring I c - NB
711 colorfor a source with a color of I c - NB
711 = - .
28 mag, the aver-age color of all objects with NB711 magnitudes brighter than26.0 mag. Note that the magnitudes and colors of objectsused in the sample selection are the values at 2. ′′ -1-0.5 0 0.5 1 1.5 2 2.5 3 21 22 23 24 25 26 27 I c - N B NB711z’ >= 26.5z’ < 26.5
Figure 3. I c - NB
711 color versus NB
711 2 . ′′ z ′ total < . z ′ total ≥ . I c - NB σ I c limiting magnitude. Note that we used 2 σ I c limiting magni-tude as a lower limit for objects whose I c magnitudes are fainter than that.A diagonal black line indicates 2 σ limiting magnitude of the I c image. Redcurves indicate the distributions of 3 σ errors in measuring I c - NB
711 colorfor a source with a color of I c - NB
711 = - .
28 mag, the average color ofall objects brighter than NB
711 = 26 . NB - excess criterion and NB
711 = 26 . corresponds to a rest-frame equivalent width (EW rest ) of 10Å assuming a flat continuum ( f ν = constant). The EW cutof 10 Å used in this paper is lower than that commonly usedin other LAE studies (20 Å). Other studies (e.g., Ouchi et al.2003; Nilsson et al. 2009) compute the EW by using the av-erage continuum flux density from the continuum both blue-ward and redward of the Ly α line, while the continuum inthis work is extrapolated from the continuum redward of theLy α line. Because the blueward continuum suffers from theattenuation by neutral hydrogen in the intergalactic medium(IGM) at high redshift, the continuum estimated by the lat-ter way is expected to be higher than that in the usual case.Consequently, the EW computed in this paper is smaller thanthat from other studies, even though the object has the sameredward continuum. The Ly α EWs computed in the usualway are also shown in Table 2; most of the LAEs comparedto LBGs show the EWs larger than 20Å when we adopt thecommonly used way.By the above criteria, 667 objects were selected. Amongthem, there are low-redshift interlopers with strong emissionlines such as H α , H β , [OIII] λ λ z ′ band.For candidates with z ′ total < . z ∼ z ′ total < . , (3) V - I c > .
55 and (4) V - I c > . I c - z ′ ) + . , (5)where z ′ total is the total magnitude in z ′ band and colors are de-rived from 2. ′′ z = 4 .
86 5
Figure 4.
Montage images of group I LAEs. Size of each image is 10 ′′ × ′′ with an LAE at the center of the image. North is at the top; east is to theleft. IDs of objects are shown on the left of the figure. Note that the Subarubroadband images displayed here ( V , I c , and z ′ ) are those before smoothing.An asterisk indicates the LAE with the spectroscopic redshift. Figure 5.
Same as Figure 4 but for group II LAEs.
Figure 6.
Same as Figure 4 but for group III LAE. with z ′ total ≥ . V - bandmagnitude ( V > σ limiting magnitude). As a result, 24 ob-jects meet all criteria simultaneously down to NB
711 = 26 . z ′ mag-nitudes brighter than 26.5 mag. Note that z ′ total here is slightlydifferent from z ′ magauto by Iwata et al. (2007) (up to ± . z ′ = 26 . ′′ radius with GALFIT(Peng et al. 2002) and subtracted them from the original im-age. In the GALFIT process, we masked the LAEs out sothat the accurate sky estimation can be obtained. Then wemade aperture photometry of the LAEs from the residual im-ages. We only selected LAEs of which magnitude differencebetween the original images and the residual images are lessthan 0.1 mag as isolated LAEs. Finally, we have 12 isolatedobjects. We divided these 12 LAEs into four groups basedon their brightness in rest-frame UV and optical wavelengths.Group I is for two LAEs which are detected above 2 σ limit- Figure 7.
Same as Figure 4 but for group IV LAEs. ing magnitudes in all bands used in SED fitting process (i.e., I c , z ′ , IRAC 3.6 µ m and 4.5 µ m). One of them, z = 4 .
82. The emissionline does not show a doublet profile in R ∼ I c and z ′ bands but detected in both 3.6 µ m and 4.5 µ m bands.Group IV is for 7 LAEs that are not detected above 2 σ limit-ing magnitudes in three or more bands. In SED fitting process,it is necessary that an LAE should be detected above 2 σ mag-nitude limit in more than two bands; therefore, we can useonly group I–III LAEs. Hereafter we call group I–III LAEsSED fitting sample. The montage images of LAEs are shownin Figure 4, 5, 6, and 7 for group I, II, III and IV, respec-tively. Aperture photometry of the isolated LAEs is summa-rized in Table 1. Neighboring objects are seen around someof group I-II LAEs, but note that the neighboring objects donot affect the photometry more than 0.1 mag. It is noteworthythat the object NB
711 image, while a bright object isseen close to µ m and4.5 µ m images. Our aperture photometry includes the lightfrom this object. However, we cannot rule out the possibilitythat the object is a foreground red object. When we subtractthe object, the aperture magnitudes in IRAC bands are fainterthan 2 σ limiting magnitudes. Thus it should be kept in mindthat the SED of this object may be contaminated. SELECTION BIAS The adjacent object of NB
711 band, is aseparate object and is likely to be a low-redshift emitter, because it is detectedin V band and does not satisfy the V - I c > .
55 criterion.
Yuma et al.
Table 1
Aperture photometry of isolated LAEsID Field V a,b NB711 a I c a,c z ′ a,c µ m a,d µ m a,d (mag) (mag) (mag) (mag) (mag) (mag)Group I30702 GOODS-N 28 . ± .
69 25 . ± .
18 26 . ± .
36 26 . ± .
36 24 . ± .
14 25 . ± . > .
51 25 . ± .
12 25 . ± .
25 25 . ± .
24 25 . ± .
16 25 . ± . > .
51 25 . ± . > .
50 26 . ± .
42 25 . ± . > . > .
51 25 . ± .
20 26 . ± . > .
38 26 . ± . > . > .
51 25 . ± . > . > .
38 24 . ± .
20 24 . ± . > .
51 24 . ± .
07 26 . ± . > . > . > . > .
51 25 . ± . > . > . > . > . > .
51 25 . ± . > . > . > . > . > .
51 25 . ± . > . > . > . > . > .
51 25 . ± . > . > . > . > . > .
51 25 . ± . > .
50 26 . ± . > . > . > .
51 25 . ± . > . > . > . > . a Errors are 1 σ values. b Upper limits are 1 σ values at 2. ′′ c Upper limits are 2 σ values at 2. ′′ d Upper limits are 2 σ values at 2. ′′ In this section, we check the sample bias in two pointsof views: bias on selecting the isolated LAEs and bias onchoosing the SED fitting sample. Figure 8(a) shows the dis-tributions of rest-frame Ly α equivalent widths of all LAEs,IRAC sample, isolated LAEs (group I–IV), SED fitting sam-ple (group I–III), and the compared LAE sample (see section7). The rest-frame Ly α EWs are calculated by assuming thatall LAEs are at z = 4 .
86 and the Ly α emission line falls intothe center of NB
711 band (Table 2). Note that the values maybe underestimated if the LAEs are not exactly at z = 4 .
86, i.e.,the Ly α emission line does not fall into the center of NB σ upper limit on I c mag-nitudes. For the spectroscopically confirmed LAE ( α emission line is in the NB
711 band, but is out of NB
711 FWHM range. By considering the transmission curveof the NB
711 filter, the corrected EW is estimated to be 40Å (Table 2) and is used hereafter. As a cross-check, we alsoexamine the EW from the spectrum and find that it is consis-tent with that obtained from the imaging. The EWs of LAEsdistribute from 10 Å (the selection limit) to 55 Å. Figures 8(b)and (c) show the distributions of z ′ magnitudes and I c - z ′ col-ors, respectively. Leftmost bins in Figures 8 (b) and (c) rep-resent LAEs with z ′ magnitudes fainter than 26.5 mag andwith magnitudes fainter than 2 σ limiting magnitudes in both I c and z ′ bands, respectively. The large number of LAEs inthe leftmost bins of both figures imply that most of LAEs havefaint UV continuum. As seen in Figures 8 (a)–(c), the isolatedLAEs (red open histogram) occupy the same range of distri-butions as all sample at z = 4 .
86 (orange histogram). In orderto check whether or not selecting isolated sources in IRACimages can cause any bias on properties of the whole LAEs,we applied the Kolmogorov–Smirnov test (KS test) to the dis-tributions of properties of all LAEs at z = 4 .
86 and the isolatedLAEs. We could not reject the null hypothesis that the rest-frame Ly α EW, z ′ magnitude, and I c - z ′ color distributionsof these two samples are drawn from the same populations atmore than 95% confidence levels. However, we could not useall isolated sample in SED fitting because of their faintness either in rest-frame UV or optical wavelengths. To investi-gate if SED fitting in the following section can represent allpopulations of LAEs at this redshift, we have to compare theSED fitting sample (group I–III, green histogram) to all LAEs(orange histogram). It is seen in Figure8 (a) that group I–IIILAEs represent both high– and low–EW LAEs. However, wecan see from Figures 8 (b) and (c) that the SED fitting sam-ple are biased toward the bright UV luminosity and red I c - z ′ colors as we neglect the leftmost bins where magnitudes areunreliable. In addition, Table 1 shows that SED fitting sampleis relatively bright in IRAC bands as compared to all isolatedLAEs implying the brighter luminosity in rest-frame opticalwavelength. Thus we conclude that the SED fitting and the re-sults hereafter are for the LAEs which have the brighter rest-frame UV and optical magnitudes, and the relatively redderUV colors. STELLAR POPULATION SYNTHESIS MODEL AND SED FITTING
In this paper, we intend to compare stellar populations ofthe LAEs to those of LBGs at the same redshift by Yabe et al.(2009). In order to make a fair comparison, we used the samestellar population synthesis model as that used in Yabe et al.(2009). The model SEDs were obtained by Bruzual & Charlotsynthesis code (2003, hereafter BC03). We used Padova 1994evolutionary track as recommended by BC03. Salpeter(1955)initial mass function (IMF) with lower and upper mass cutoffsof 0.1 and 100 M ⊙ is assumed. We made models by fixing themetallicity at 0.2 Z ⊙ and assuming the constant star forma-tion history. BC03 uses quasi-logarithmic 221 time steps from0.1 Myr to 20 Gyr. Time steps were adopted to 51 logarithmicsteps both to reduce the calculation time and to avoid dealingwith an unequally spaced scale of the original 221 models.The age of the Universe at z ∼ ∼ . Fixing the metallicity at the lower abundance (0.005 Z ⊙ ) does not sig-nificantly change the fitting results. The stellar masses derived from the lowermetallicity model differ from those derived from 0.2 Z ⊙ model by ∼
10% atmost. The average differences of age, color excess, and star formation rateare ∼ ± . ∼ tellar populations of LAEs at z = 4 .
86 7
Figure 8.
Distributions of rest-frame Ly α EW, z ′ –band magnitude, and I c - z ′ color of LAE samples. We divide each figure into 2 panels for display purpose. Inthe top panel, an orange histogram shows the distribution of all 24 LAEs at z = 4 .
86, while the blue and red histograms are for the IRAC samples and the isolatedLAEs (group I–IV), respectively. In the bottom panel, red histograms refer to the isolated LAEs (the same as in the top panel). Group I–III LAEs are shownin green histograms. Distributions of group I LAEs and α EWs of LAEs; the number of arrows directly corresponds to the number of LAEs with Ly α EWlower limits. Note that the leftmost bins in panels (b) and (c) indicate, respectively, the LAEs fainter than z ′ = 26 . σ limiting magnitudes in neither I c nor z ′ bands. account by using Calzetti extinction law (Calzetti et al. 2000)by changing the color excess E ( B - V ) from 0.0 mag to 0.8mag with a 0.01 step. Attenuation by the IGM is calculatedwith the prescription by Madau (1995). The model spectrawere then convolved with the appropriate filter transmissioncurves to give model fluxes. This is exactly the same as donein Sawicki & Yee (1998). Except for one spectroscopicallyconfirmed LAE ( z = 4 . z = 4 .
86 under assumption that Ly α emission of our LAEsample is detected at the center of the NB711 bandpass.Some of the observed SEDs including one spectro-scopically confirmed LAE show a significant excess inthe 3.6 µ m band as compared with the magnitude in the4.5 µ mband. This is likely to be due to the H α emission line,which falls into IRAC ch1 (3.6 µ m) at this redshift. The ex-cess is also seen in some fraction of z ∼ α emission shows a bet-ter fit (Chary et al. 2005; Finkelstein et al. 2008; Yabe et al.2009). As adopted by Yabe et al. (2009), the spectrum ofH α emission is included in the synthesis model spectrum bythe following process. The luminosity of H α emission is cal-culated from the star formation rate of the model by using therelation by Kennicutt (1998). The dust extinction to the line(Calzetti et al., 2000) is assumed to be the same as that tostellar component. The H α flux density is finally put intoeach model SED. The existence of H α emission line in themodels makes the fit better without adding a free parameter(Yabe et al. 2009).We use the SEDfit software (Sawicki, in prep.), whichis an evolved version of the SED-fitting software used inSawicki & Yee (1998) and subsequent papers, including the z ∼ χ minimization. Forobjects that are undetected in some of the bands (i.e., group Adopting this relation may be disputable. Dependences on metallicitiesand differences of extinction to stellar continuum and nebular emission arediscussed by Yabe et al. (2009; their Appendix B).
II and III objects), SEDfit employs a modification of the χ formalism: as is standard in the χ approach, for the filters inwhich the object is detected the maximum likelihood calcu-lation considers the likelihood that the detected data deviatefrom the model given the uncertainties; for the undetected fil-ters, the calculation adds in the likelihood that a given modelwould/would not have been detected given the upper limit ofthe non-detection. For more details on this technique see Saw-icki (in prep.).In this paper, we constructed an observed SED of each in-dividual object from the photometry in I c , z ′ , IRAC 3.6 µ m,and IRAC 4.5 µ m bands. As explained in section 2, we donot use V -band photometry in the SED fitting process in or-der to avoid the uncertainty due to IGM absorption. IRAC5.8 µ m and 8.0 µ m photometry is not used because of the lowS/N ratio of the images. Including the upper limits of the pho-tometry in IRAC 5.8 µ m and 8.0 µ m bands does not usefullyimprove the fitting. Free parameters in our fitting process areage, color excess, and scaling normalization. The SFR is ob-tained from the scaling normalization. We then find the stellarmass by using the age and SFR (Sawicki & Yee 1998). RESULTS
The fitting results are summarized in Table 2. Errors inthe table are at 68% confidence level and are determined asfollows. The Monte Carlo realizations for each object areperformed; we vary the input fluxes within their photomet-ric uncertainties, rederive the best-fitting model and repeat it100 times. The error on an individual parameter is then 68%range of the realizations in that parameter. The best-fittingmodel spectra are shown with the observed SEDs in Figure 9.The derived stellar masses are ranging from 10 - M ⊙ with the median value of 2 . × M ⊙ . A typical erroron the stellar mass is ∼ . M ⊙ with the smallertypical error of ∼ . µ m] magnitudes by assuming f λ ∝ λ β with β derived from the z ′ - [4 . µ m] color of an individual ob- Yuma et al. Table 2
The best fit resultsID Field rest Ly α EW a log[Mass] log[Age] E(B-V) log[SFR] χ ν q b (Å) ( M ⊙ ) (yr) (mag) ( M ⊙ yr - )Group I30702 GOODS-N 14 (51) 9 . + . - . . + . - . . + . - . . + . - . .
02 1 . . + . - . . + . - . . + . - . . + . - . .
08 1 . > >
44) 9 . + . - . . + . - . . + . - . . + . - . . < . . + . - . . + . - . . + . - . . + . - . .
63 2 . > >
11) 10 . + . - . . + . - . . + . - . . + . - . . < . a The rest-frame Ly α EWs are determined by assuming that all LAEs are at z = 4 .
86 except for α wavelength to estimate the continuum. b The clumpiness parameter (see text for further details).
Figure 9.
Observed SEDs and the best-fitting models for group I–III LAEs.In each panel, the observed SED is shown with filled circles; the best-fittingmodel SED is indicated by opened squares. Solid line represents the best-fitting model spectrum with H α emission line. Arrows indicate the 2 σ upperlimits. V-band photometry is also shown in the figures but is not used in SEDfitting. ject. For LAEs with upper limit, we used the upper limit toestimate the magnitude. Despite the small size of the sample,the rest-frame optical luminosity is likely to be a tracer of thestellar mass. The relationship between the stellar mass andextinction-corrected optical absolute magnitude also shows acorrelation (Figure 10(b)). A mass-to-light ratio is not per-fectly linear; it becomes larger in the brighter LAEs. Thistrend is similar to the results of LBGs at the same redshiftby Yabe et al. (2009) which will be discussed in more detailsin the next section. From the relationship between the stellarmass and the rest-frame optical luminosity, it is implied thatmost LAEs at z = 4 .
86, which could not be detected above2 σ limiting magnitudes in 4 . µ m band (Table 1), are likely tohave stellar masses lower than 10 M ⊙ .The best-fitting stellar ages are in a wide range from 7.4Myr to 437 Myr with the median age of 25 Myr. Althoughwe did not limit the model age at the age of the universe at z = 4 .
86 ( ∼ . ∼ . E ( B - V ) range from 0.14 mag to thevalue as high as 0.40 mag. The median value is 0.27 mag.The typical error on E ( B - V ) is ∼ . . - .
30 mag; any correlations between the dust extinctionand magnitudes or other properties cannot be seen. The pres-ence of some amount of color excess indicates that LAEs atthis redshift are not free from dust, which is consistent withwhat Finkelstein et al. (2007, 2008, 2009) found for LAEs at z ∼ .
5. Many authors (Neufeld 1991; Hansen & Oh 2006;Finkelstein et al. 2009) proposed that the presence of dust inclumpy clouds can enhance the Ly α EW of a galaxy. To testthis hypothesis, we estimated the clumpiness parameter ( q ): q τ is Ly α line opacity where τ is an optical depth for thecontinuum (Finkelstein et al. 2008). If q <
1, the Ly α EWis enhanced on the hypothesis that Ly α photons suffer lessextinction than the continuum photons and the clumpy cloudmodel would be supported. We calculate q from the ratioof observed and intrinsic Ly α luminosities. The observedLy α luminosity is derived from the observed Ly α EW andthe continuum flux density, while the intrinsic one is calcu-lated from the H α luminosity by assuming case B recombi-nation. Except for q parameter, all LAEs show q > α photons suffer from dust extinction larger thanthe continuum. The uncertainty of q value is large; neverthe-less, the q values are all near 1 - α emission appears to be more attenuated than the con-tinuum, the difference is not so much. In section 3, however,we mentioned that the observed Ly α EWs may be underesti-mated if the Ly α emission line does not fall into the centralpart of NB
711 filter and are further underestimated due to thenon-detection in I c band for some LAEs. Underestimation ofthe Ly α EW results in the overestimation of the q parameter.Accordingly, the objects without spectroscopic confirmation,i.e., q values less than thoseshown in Table 2, if their Ly α line did not fall into the centralpart of the NB
711 filter. The spectroscopic observations aredesirable to investigate the real Ly α EWs and q values of theLAEs.tellar populations of LAEs at z = 4 .
86 9 −23−22−21−20 S t e ll a r M a ss e s ( M s un ) M (a) −25−24−23−22−21 S t e ll a r M a ss e s ( M s un ) M (b) E ( B − V ) ( m ag ) M (c) E ( B − V ) ( m ag ) M (d) Figure 10.
Plots of derived stellar masses and color excesses of group I–III LAEs at z = 4 .
86 against the rest-frame UV or optical absolute magnitudes. Filledcircles, open triangles, and open circles represent group I, group II, and group III LAEs, respectively. Vertical error bars represent the 68% errors as described intext. Arrows indicate the objects whose absolute magnitudes were calculated from the upper limits.
The best-fitting star formation rates (SFRs) of LAEs rangefrom 55 M ⊙ yr - to 209 M ⊙ yr - . The median value of thederived SFRs is 132 M ⊙ yr - . The typical error on SFR foreach object is ∼ . ± .
04 dex).Age, color excess, and SFR for group I LAEs are also ro-bust ( ± .
24 dex, ± .
02 mag, and ± .
10 dex, respectively).For group II and III LAEs, age, color, and SFR do not agreewell with each other, but agree within their large uncertainties.H α emission accounts for the largest expected nebular lineemission to be detected from our LAEs. Including H α emis-sion to the model spectrum when we performed SED fitting in the last section is likely to be the reason why adding the othernebular emissions give the similar results. COMPARISONS TO LBGS AT Z ∼ In this paper, we aim at comparing the derived stellar pop-ulations of LAEs at z ∼ z ∼ z ′ < . V - dropout criteria as those used in this study. In addition, theSED models and the fitting technique are also the same. Inorder to make a fair comparison, we used only LAEs whose z ′ total magnitudes are brighter than 26.5 mag. Because part ofthe LBGs are spectroscopically confirmed with the mean red-shift of z ∼ . z ′ magnitude is equivalent to taking the samelimit on rest-frame UV luminosity. There are six LAEs with z ′ total < . NB
711 image before making photometry (section 2.1.2).0 Yuma et al.
Figure 11.
Relations of output parameters from SED fitting for group I–III LAEs at z = 4 .
86. Filled circles, open triangles, and open circles represent group I, IIand III LAEs, respectively.
Accordingly, magnitudes measured from the images beforeand after smoothing are not necessarily the same.From 617 LBGs at z ∼
5, (Iwata et al. 2007), Yabe et al.(2009) selected 170 LBGs by eye inspection as isolated ob-jects both in z ′ - band and in IRAC images. The 170 LBGsare divided into four categories according to the detection in3.6 µ m and 4.5 µ m filters. The LBGs that are detected in both3.6 µ m and 4.5 µ m images are in category 1, while those onlydetected in either 3.6 µ m or 4.5 µ m are put in category 2 or 3,respectively. The rest that are detected neither in 3.6 µ m nor4.5 µ m images are in category 4. We used categories 1, 2, and3 LBGs in the comparison. Yabe et al. (2009) performed theSED fitting for 64 LBGs in category 1. For categories 2 and3 LBGs, we refitted them by using an upper limit either in4.5 µ m or 3.6 µ m band, respectively. 17 LBGs are best fitted with the stellar ages larger than the cosmic age at z ∼ ∼ . NB
711 filter covers only a small fraction of theredshift range, LAEs that are not in this redshift range are ex-pected among the LBG sample.Before going to the comparison between stellar populationsof LAEs and those of LBGs, we try to figure out what theirphotometry can tell us without SED fitting. Another thingtellar populations of LAEs at z = 4 .
86 11 -0.8-0.4 0 0.4 0.8 23 24 25 26 27 I c - z ’ z’ magnitude (AB) (a) -2-1 0 1 2 3 4 5 6 21 22 23 24 25 26 27 z ’ - [ . m m ] m m magnitude (AB) (b) Figure 12. (a) I c - z ′ versus z ′ magnitude diagram; (b) z ′ - [4 . µ m] versus 4.5 µ m magnitude diagram. Group I LAEs and which should be kept in mind is that three LAEs which areused to compare with the LBGs in this section (i.e., groupI LAEs and z = 4 . z ∼
5. No clear trend is seen be-tween z ′ magnitude and I c - z ′ color (Figure 12(a)). The figureshows that the LAEs occupy the faint part of z ′ -magnitudedistribution and the red part of I c - z ′ color distribution of theLBGs. Distribution of these LAEs at the faint part of UV lu-minosity among LBGs may support the deficiency of strongLy α emission among bright LBGs claimed by Ando et al.(2006), though the sample size is only three. Figure 12(b)shows the color-magnitude diagram of z ′ - [4 . µ m] color and4.5 µ m magnitude. The z ′ - [4 . µ m] colors of LBGs spreadover a wide range from - µ m-band magnitude and the relatively bluer part ofthe color distributions than the LBGs.Output parameters of SED fitting are plotted against therest-frame UV and optical luminosities in Figure 13. Thetop panels show the comparisons between the derived stel-lar masses of LAEs and those of LBGs. The stellar massesof the compared LAE sample range from 10 M ⊙ to 10 M ⊙ .In contrast, LBGs distribute in a larger range from 10 M ⊙ to10 M ⊙ . It is seen in the figure that the LAEs locate in theregion where LBGs distribute and in the almost middle partof LBGs’ distribution. At the same bin of rest-frame UV oroptical luminosity, LAEs seem to have comparable masses toLBGs.Showing no relation between ages and luminosities, Figures13(c) and 13(d) indicate that the LAEs again lie on the samedistribution of LBGs. The median ages are similar; they are25 Myr and 22 Myr for LAEs and LBGs, respectively. How-ever, age uncertainties of the LAEs cover the whole range ofage distribution of the LBGs. It is difficult to draw any con-clusion.Figures 13(e) and 13(f) show the plots of the derived colorexcesses against UV and optical absolute magnitudes, respec-tively. Although the dust extinction of the LAEs is not zero, they seem to lie at the relatively lower region of the color ex-cess distribution of LBGs. At the fixed rest-frame UV or op-tical luminosity, there seems to be no difference in E ( B - V ),even if we take the uncertainties into account. Figure 13(e)shows that there is no relation between the dust extinction andthe rest-frame UV luminosity. In contrast, Figure 13(f) seemsto show a correlation of LBGs between the color excesses andthe optical magnitudes. Though we can not state any correla-tion for only 3 LAEs, the LAEs still lie on the correlation ofLBGs.The SFRs are plotted against rest-frame UV and optical ab-solute magnitudes in Figures 13(g) and 13(h), respectively.Both figures show a correlation of SFRs of the LBGs withUV (but weak) and optical luminosity. It is seen from the fig-ures that the LAEs locate in the lower part of the LBGs’ dis-tribution. A median SFR of the LAEs is 132 M ⊙ yr - , whileit is 187 M ⊙ yr - for LBGs. Note that the LAEs used in thecomparison are biased toward the brightest sample among thewhole LAE sample (section 4), which probably results in thehigher SFRs of the LAEs. In conclusion, no significant dif-ferences in the stellar properties between LAEs and LBGs areseen at the same luminosity.A physical reason that makes an LAE an emitter is stillunclear. In order to investigate whether or not the dif-ference between LBGs and LAEs has any dependence onthe equivalent widths (EWs) of Ly α emission, we plottedthe output parameters of SED fitting against the rest-frameLy α EWs in Figure 14. Group I–III LAEs are plotted with fiveLBGs which have spectroscopic Ly α EWs (Ando et al. 2004;Kajino et al. 2009). Arrows in the figures represent the lowerlimits on EWs of the LAEs; I c magnitudes of them are fainterthan 2 σ limiting magnitude. Note that, except for one spectro-scopically confirmed LAE, the Ly α EWs of LAEs are likelyto be lower limits if their Ly α emission lines do not fall intothe center of NB
711 band. According to the figure, we can-not find any significant correlation between the stellar prop-erties and the rest-frame Ly α EWs. Recently, Kornei et al.(2009) studied these relations for z ∼ α EW, while large Ly α EW is seen in older, lower SFR,and less dusty LBGs. Such trends are difficult to isolate withonly three LAEs. A larger sample at z ∼ −23−22−21−20 S t e ll a r m a ss ( M s un ) M (a) −26−25−24−23−22−21−20−19 S t e ll a r m a ss ( M s un ) M (b) −23−22−21−20 A ge ( y r) M (c) −26−25−24−23−22−21−20−19 A ge ( y r) M (d) E ( B − V ) ( m ag ) M (e) E ( B − V ) ( m ag ) M (f) −23−22−21−20 S F R ( M s un / y r) M (g) −26−25−24−23−22−21−20−19 S F R ( M s un / y r) M (h) Figure 13.
Output parameters of SED fitting against the UV and optical absolute magnitudes. Symbols are the same as those in Figure 12. tellar populations of LAEs at z = 4 .
86 13
0 10 20 30 40 50 60 70 S t e ll a r m a ss ( M s un ) Rest−frame Ly a EW (Å) (a)
0 10 20 30 40 50 60 70 S t e ll a r age ( y r) Rest−frame Ly a EW (Å) (b) E ( B − V ) ( m ag ) Rest−frame Ly a EW (Å) (c)
0 10 20 30 40 50 60 70 S F R ( M s un / y r) Rest−frame Ly a EW (Å) (d)
Figure 14.
Output parameters from SED fitting for group I–III LAEs versus rest-frame Ly α equivalent widths. Filled circles, open triangles, and open circlesrepresent group I, group II, and group III LAEs, respectively. LBGs that have spectroscopic Ly α EWs are also shown in filled squares. SUMMARY
In this paper, we studied the stellar properties of Lymanalpha emitters (LAEs) at z = 4 .
86 using SED fitting. Bynarrowband and broadband observations by Suprime-Camon Subaru telescope, 24 LAEs were selected in the area of ∼ . around GOODS-N field. In addition to theoptical photometry, we obtained the mid-infrared photome-try from data taken by IRAC on Spitzer space telescope inthe GOODS-N field as well as the surrounding area in or-der to cover most part of Subaru area. We selected 12 LAEsthat are isolated from the neighboring objects. We performedSED fitting of five LAEs which are detected above 2 σ mag-nitude limits in more than 2 bands. Selecting those five LAEscould introduce a bias toward the bright red galaxies. ModelSEDs are built by assuming the constant star formation his-tory with fixed metallicity of 0.2 Z ⊙ , the Salpeter IMF rangingfrom 0.1 M ⊙ to 100 M ⊙ , and the extinction law of Calzetti etal. (2000). The derived stellar masses of the LAEs range from10 to 10 M ⊙ with the median value of 2 . × M ⊙ . Thederived ages cover wide range from 7.4 Myr to 437 Myr withthe median value of 25 Myr. The color excess are between0 . - . - M ⊙ yr - . The median color excess and SFRs are 0.27 mag and132 M ⊙ yr - , respectively. The high SFRs are probably dueto the selection effect; we selected the LAEs that are brightenough to be detected in rest-frame UV and optical bands, which results in selecting the LAEs with the higher SFRs. Weinvestigate the correlations between the stellar properties de-rived by SED fitting and the photometric properties of LAEsand found no significant correlation due to both small size ofthe sample and the large uncertainty on fitting results.The main objective of this study is to compare LAEs toLBGs at the same redshift. LBGs were selected by V-dropoutcriteria (Iwata et al.2007). Their stellar populations were de-rived by Yabe et al. (2009). Because those LBGs are se-lected from the same set of data and stellar population wasderived by the same SED fitting model, we can make a faircomparison between their stellar populations. We comparedthree LAEs to 88 LBGs down to the same UV luminositylimit. These three LAEs are the brightest and reddest onesamong the whole LAE sample in this study. The comparisonsof SED-fitting parameters show that LAEs locate in the re-gion where LBGs distribute; the physical properties of LAEsand LBGs occupy similar parameter spaces. At the same rest-frame UV or optical luminosity, there is no difference of stel-lar properties between LAEs and LBGs. In order to figureout properties that control the Ly α EWs, we plotted the out-put parameters against the rest-frame Ly α EWs ranging from0 to 40 Å. We could not find any significant correlation be-tween them. A larger number of sample is needed to see anycorrelation of Ly α EW if exists.We thank the referee for their valuable comments which im-proved the paper. This work is supported by the Grant-in-Aid4 Yuma et al.for Scientific Research on Priority Areas (19047003) and bythe Grant-in-Aid for Global COE program "The Next Genera-tion of Physics, Spun from Universality and Emergence" fromMinistry of Education, Culture, Sports, Science, and Technol-ogy of Japan.
REFERENCESAjiki, M., et al. 2003, ApJ, 126, 2091Ajiki, M., et al. 2004, PASJ, 56, 597Ajiki, M., et al. 2006, ApJ, 638, 596Ando, M., Ohta, K., Iwata I., Watanabe, C., Tamura, N., Akiyama, M., &Aoki, K., 2004, ApJ, 610, 635Ando, M., Ohta, K., Iwata I., Akiyama, M., Aoki, K., & Tamura, N. 2006,ApJ, 645, L9Ando, M., Ohta, K., Iwata I., Akiyama, M., Aoki, K., & Tamura, N. 2007,PASJ, 59, 717Bertin, E. & Arnouts, S., 1996, A&AS, 117, 393Bouwens, R. J., Thompson, R. I., Illingworth, G. D., Franx, M., vanDokkum, P. G., Fan, X., Dickinson, M. E., Eisenstein, D. J., & Rieke, M.J. 2004, ApJ, 616, L79Bouwens, R. J., Illingworth, G. D., Franx, M., & Ford, H. 2008, ApJ, 686,230Bruzual, G., & Charlot, S. 1993, ApJ, 405, 538Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000Calzetti, D. 1997, in AIP Conf. Proc. 408, The Ultraviolet Universe at Lowand High Redshift: Probing the Progress of Galaxy Evolution, eds. W. H.Waller, M. N. Fanelli, J. E. Hollis, & A. C. Danks (Woodbury: AIP), 403Calzetti, D., Armus, L., Bohlin, C., Kinney, A.L., Koornneef, J., &Storchi-Bergmann, T. 2000, ApJ, 533, 682Capak, P., et al. 2004, ApJ, 127, 180Chary, R., Stern, D., & Eisenhardi, P. 2005, ApJ, 635, L5Cowie, L. L. & Hu, E.M. 1998, ApJ, 115, 1319Dickinson, M., Giavalisco, M., & The GOODS Team 2003, in The Mass ofGalaxies at Low and High Redshift, eds. R. Bender & A. Renzini (Berlin:Springer-Verlag), 324Finkelstein, S. L., Rhoads, J. E., Malhotra, S., Pirzkal, N., & Wang, J. 2007,ApJ, 660, 1023Finkelstein, S. L., Rhoads, J. E., Malhotra, S., Grogin, N., & Wang, J. 2008,ApJ, 678, 655Finkelstein, S. L., Rhoads, J. E., Malhotra, S., & Grogin, N. 2009, ApJ, 691,465Fioc, M., & Rocca-Volmerange, B. 1997, A&A, 326, 950Fioc, M., & Rocca-Volmerange, B. 1999, astroph/9912.179Fujita, S. S., et al. 2003, ApJ, 125, 13Fynbo, J. U., Møller, P., & Thomson, B. 2001, A&A, 374, 443Gawiser, E., et al. 2006, ApJ, 642, L13Gawiser, E., et al. 2007, ApJ, 671, 278Giavalisco, M., Steidel, C. C., & Macchetto, F. D. 1996, ApJ, 470, 189Giavalisco, M., Steidel, C. C., Adelberger, K. L., Dickinson, M. E., Pettini,M., & Kellogg, M. 1998, ApJ, 503, 543Gronwall, C., et al. 2007, ApJ, 667, 79Guaita, L., et al. 2009, ApJ, submitted, [arXiv:0910.2244]Hansen, M. & Oh, S. P. 2006, MNRAS, 367, 979Hu, E. M., & McMahon, R.G. 1996, Nature, 382, 231Hu, E. M., Cowie, L. L., & McMahon, R. G. 1998, ApJ, 502, L99Hu, E. M., Cowie, L. L., Capak, P., McMahon, R. G., Hayashino, T., &Komiyama, Y. 2004, ApJ, 127, 563Iwata, I., Ohta, K., Tamura, N., Ando, M., Wada, S., Watanabe, C.,Akiyama, M., & Aoki, K. 2003, PASJ, 55, 415Iwata, I., Ohta, K., Tamura, N., Akiyama, M., Aoki, K., Ando, M., Kiuchi,G., & Sawicki, M. 2007, MNRAS, 376, 1577Iye, M., et al. 2004, PASJ, 56, 381Iye, M., et al. 2006, Nature, 443, 186Kajino, H., et al. 2009, ApJ, 704, 117Kashikawa, N., et al. 2006, ApJ, 648, 7Kennicutt, R. C. 1998, ARA&A, 36, 189 Kodaira, K., et al. 2003, PASJ, 55, L17Kornei, K. A., Shapley, A. E., Erb, D. K., Steidel, C. C., Reddy, N. A.,Pettini, M., & Bogosavljevi´c, M. 2009, arXiv:0911.2000v1Kudrtizki, R.-P., et al. 2000, ApJ, 536, 19Lai, K., Huang, J., Fazio, G., Cowie, L. L., Hu, E. M., & Kakazu, Y. 2007,ApJ, 655, 704Lai, K., et al. 2008, ApJ, 674, 70Lowenthal, J. D., Koo, D. C., Guzman, R., Gallego, J., Phillips, A. C., Faber,S. M., Vogt, N. P., Illingworth, G. D., & Gronwall, C. 1997, ApJ, 481, 673Madau, P. 1995, ApJ, 441, 18Malhotra, S., & Rhoads, J. E. 2002, ApJ, 565, L71Malhotra, S. & Rhoads, J. E. 2004, ApJ, 617, L5Miyazaki, S., et al. 2002, PASJ, 54, 833Murayama, T., et al. 2007, ApJS, 172, 523Neufeld, D., A. 1991, ApJ, 370, L85Nilsson, K. K., et al. 2007, A&A, 471, 71Nilsson, K. K., Tapken, C., Møller, P., Freudling, W., Fynbo, J. P. U.,Meisenheimer, K., Laursen, P., Östlin, G. 2009, A&A, 498, 13Oke, J. B., Gunn, J. E. 1983, ApJ, 266, 713Ono, Y., et al. 2009, MNRAS, in pressOta, K., et al. 2007, ApJ, 677, 12Ouchi, M., et al. 2003, ApJ, 582, 60Ouchi, M., et al. 2004, ApJ, 611, 660Ouchi, M., et al. 2008, ApJS, 176, 301Papovich, C., et al. 2004, ApJ, 600, L111Partridge, R.B. & Peebles, P.J.E. 1967, ApJ, 147, 868Pirzkal, N., Malhotra, S., Rhoads, J. E., & Xu, C. 2007, ApJ, 667, 49Peng, C. Y., et al. 2002 ApJ, 124, 266Pentericci, L., et al. 2007, A&A, 471, 433Pettini, M., Kellogg, M., Steidel, C. C., Dickinson, M., Adelberger, K. L., &Giavalisco, M. 1998, ApJ, 508, 539Reddy, N. A., Erb, D. K., Steidel, C. C., Shapley, A. E., Adelberger, K. L., &Pettini, M. 2005, ApJ, 633, 748Reddy, N. A., Steidel, C. C., Erb, D. K., Shapley, A. E., & Pettini, M. 2006,ApJ, 653, 1004Reddy, N. A., Steidel, C. C., Pettini, M., Adelberger, K. L., Shapley, A. E.,Erb, D. K., & Dickinson, M. 2008, ApJS, 175, 48Rhoads, J. E., Malhotra, S., Dey, A., Stern, D., Spinrad, H., & Jannuzi, B. T.2000, ApJ, 545, L85Rhoads, J. E., & Malhotra, S. 2001, ApJ, 563, L5Rhoads, J. E., et al. 2004, ApJ, 611, 59Salpeter, E. E. 1955, ApJ, 121, 161Sawicki, M., & Thompson, D. 2005, ApJ, 635, 100Sawicki, M., & Thompson, D. 2006, ApJ, 648, 299Sawicki, M., & Yee, H. K. C. 1998, AJ, 115, 1329Sawicki, M. 2010, in prep.Shapley, A. E., Steidel, C. C., Adelberger, K. L., & Pettini, M. 2001, ApJ,562, 95Shapley, A. E., Steidel, C. C., Adelberger, K. L., Dickinson, M., Giavalisco,M. 2003, ApJ, 588, 65Shimasaku, K., et al. 2003, ApJ, 586, L111Shimasaku, K., et al. 2004, ApJ, 605, L93Shimasaku, K., et al. 2006, PASJ, 58, 313Shimizu, I., & Umemura, M. 2010, [arXiv:1004.0781v1]Steidel, C. C., Giavalisco, M., Dickinson, M., & Adelberger, K. L. 1996, AJ,112, 352Steidel, C. C., Adelberger, K. L., Giavalisco, M., Dickinson, M., & Pettini,M. 1999, ApJ, 519, 1Steidel, C. C., Adelberger, K. L., Shapley, A. E., Pettini, M., Dickinson, M.,& Giavalisco, M. 2003, ApJ, 592, 728Taniguchi, Y., et al. 2005, PASJ, 57, 165Veneman, B.P., et al. 2002, ApJ, 569, L11Veneman, B.P., et al. 2004, A&A, 424, L17Williams, R. E., et al. 1996, ApJ, 112, 4Yabe, K., Ohta, K., Iwata, I., Sawicki, M., Tamura, N, Akiyama, M., &Aoki, K. 2009, ApJ, 693, 507Yagi, M., Kashikawa, N., Sekiguchi, M., Doi, M., Yasuda, N., Shimasaku,K., & Okamura, S. 2002, AJ, 123, 66Yoshida, M., et al. 2006, ApJ, 653, 988APPENDIXEFFECTS OF NEBULAR EMISSION
In order to investigate the effects of nebular emission on the synthesis models, we used the stellar population synthesis model byFioc & Rocca-Volmerange (1997, 1999; hereafter PEGASE). The PEGASE can add nebular emission (continuum and emissiontellar populations of LAEs at z = 4 .
86 15
Table 3
The best fit results for PEGASE modelID Field log[Mass] log[Age] E(B-V) log[SFR] χ ν ( M ⊙ ) (yr) (mag) ( M ⊙ yr - )Group I30702 GOODS-N 9 . + . - . . + . - . . + . - . . + . - . . . + . - . . + . - . . + . - . . + . - . . . + . - . . + . - . . + . - . . + . - . . . + . - . . + . - . . + . - . . + . - . . . + . - . . + . - . . + . - . . + . - . . lines) to the model spectrum. The hydrogen emission lines are computed from the number of Lyman continuum photons byassuming case B recombination. Other emission lines are calculated from the observed ratios to H β for typical local starbursts(Fioc & Rocca-Volmerange 1997). Other prescriptions are the same as those used in the case of BC03 model described in themain text. We used Salpeter IMF (1995) with lower and upper mass cutoffs of 0.1 and 100 M ⊙ , respectively, adopted metallicityof 0.2 Z ⊙ , and assumed the constant star formation history. We re-scaled the age step of PEGASE into the logarithmic scale as wedid for BC03 model. The effect of dust attenuation is taken into account by using Calzetti extinction law (Calzetti et al. 2000). E ( B - V ) varies from 0.0 mag to 0.8 mag by 0.01 mag step. The attenuation due to IGM is followed by Madau (1995). Theredshift is fixed at z = 4 .
86 except for
Figure 15.
Observed SEDs and the best-fitting models for LAEs in group I (left panel), group II (middle panel), and group III (right panel). In each panel, theobserved SED is shown with filled circles; the best-fitting model SEDs are indicated by opened squares and asterisks for BC03 and PEGASE models, respectively.Black line represents the best-fitting BC03 model spectrum with H α emission line, while blue one represents the PEGASE model. 2 σ upper limit of V-bandphotometry is also shown with the arrow; however, the data is not used in SED fitting. The results obtained for group I–III LAEs with PEGASE are summarized in table 3. The best-fit spectra are shown as acomparison with BC03 best-fitting spectra in Figure 15. The comparisons of resulting parameters are presented in Figure 16.Figure 16(a) shows that the stellar masses derived by both models are well in agreement with each other. The stellar massesobtained by the PEGASE model are averagely smaller than those by the BC03 model by about 0.04 dex. As seen in Figure 16(b),the stellar ages of most LAEs derived by PEGASE are younger than those derived from BC03 models. Inclusion of nebularemission averagely decreases the ages of all LAEs by 0.46 dex. However, due to the large uncertainties in determining the stellarages, the derived ages seem to be in agreement within the uncertainty range. Figure 16(c) shows that the color excesses derivedby PEGASE are comparable to those by BC03 model. The average difference is 0.04 mag except for the group III LAE. Asseen in Figure 16(d), the SFRs derived by both models seem to be comparable for most of the LAEs. However, the rather largedifferences are seen for the group III LAE and one in group II ( S t e ll a r m a ss ( M s un ) P ega s e Stellar mass (Msun) BC03 (a) S t e ll a r age ( y r) P ega s e Stellar age (yr) BC03 (b) E ( B - V ) P ega s e E(B-V) BC03 (c) S F R ( M s un / y r) P ega s e SFR (Msun/yr) BC03 (d)