搜档网
当前位置:搜档网 › An Overdensity of Lyman-alpha Emitters at Redshift z=5.7 near the Hubble Ultra Deep Field

An Overdensity of Lyman-alpha Emitters at Redshift z=5.7 near the Hubble Ultra Deep Field

An Overdensity of Lyman-alpha Emitters at Redshift z=5.7 near the Hubble Ultra Deep Field
An Overdensity of Lyman-alpha Emitters at Redshift z=5.7 near the Hubble Ultra Deep Field

a r

X

i

v

:a

s t

r o

-

p

h

/

5

1

4

7

9v

1

2

1

J

a

n

2

5

An Overdensity of Lyman-αEmitters at Redshift z ≈5.7near the Hubble Ultra Deep Field J.X.Wang 1,3,S.Malhotra 2,J.E.Rhoads 2ABSTRACT We have identi?ed an obvious and strong large scale structure at redshift z ≈5.75in a wide (31′×33′)?eld,narrowband survey of the Chandra Deep Field South region.This structure is traced by 17candidate Ly αemitters,among which 12are found in an 823nm ?lter (corresponding to Ly αat z =5.77±0.03)and 5in an 815nm image (z =5.70±0.03).The Ly αemitters in both redshift bins are concentrated in one quadrant of the ?eld.The Hubble Ultra Deep Field,Chandra Deep Field South,and GOODS-South ?elds all lie near the edge of this overdensity region.Our results are consistent with reports of an overdensity in the UDF region at z ≈5.9.This structure is the highest redshift overdensity found so far.Subject headings:cosmology:observations —galaxies:evolution —galaxies:high-redshift —large-scale structure of universe 1.Introduction The study of the large-scale clustering of high redshift galaxies is essential to understand the formation and evolution of galaxies.To date,many possible detection of large-scale clus-

tering of high redshift galaxies have been reported (e.g.,Steidel et al.1998,2000;Venemans et al.2002,2004;Miley et al.2004;Palunas et al.2004;Ouchi et al.2001,2003,2004;Foucaud et al.2003;and references therein).

The Lyαemission line searches have been very e?ective at?nding high redshift galaxies, especially at z>5where most of the spectroscopically con?rmed galaxies show Lyαemission. These include our Large Area Lyman Alpha(LALA)survey(e.g.,Rhoads et al.2000)and other recent searches(Cowie&Hu1998;Hu et al1998,2002,2004;Kudritzki et al2000; Fynbo,Moller,&Thomsen2001;Pentericci et al.2000;Stiavelli et al.2001;Ouchi et al. 2003;Fujita et al.2003;Shimasaku et al.2003;Kodaira et al.2003,Ajiki et al.2004;Kurk et al.2004;Tran et al.2004,and references therein).Large-scale clustering of high redshift Lyαemitters has also been reported in some of these surveys(Venemans et al.2002,2004; Ouchi et al.2001,2003,2004).

In this letter we report the detection of a large-scale structure of Lyαemitters at z~5.7 in Chandra Deep Field South(CDF-S).The candidate Lyαemitters at z~5.7were selected using deep narrowband images,following the selection criteria developed in the Large Area Lyman Alpha survey(Rhoads et al.2000,2001,2003,2004).Throughout this paper,we assume a cosmology with H0=71km s?1Mpc?1,?M=0.27and?Λ=0.73(cf.Spergel et al.2003).

2.Deep Narrow Band Imaging

We imaged the area around Chandra Deep Field South using two narrowband?lters with central wavelengthλc of815and823nm.The full width at half maximum(FWHM) transmissions of the two?lters is7.5nm,with peak throughputs of~90%.These were new ?lters of high quality.In particular,their throughput does not vary substantially across the ?eld of view(unlike the similar but older?lters used by Rhoads&Malhotra2001).The deep narrowband images were obtained using the Mosaic II CCD imager at the CTIO4m V.M.Blanco telescope,on2003Dec.22–30(UT).Chip3of the8SITe2Kx4K CCDs was turned o?during the observations due to high ampli?er noise.Imaging data reduction followed the methods described in Rhoads et al.(2000,2004).To summarize,in addition to the standard CCD data reduction steps(overscan subtraction,bias frame subtraction, and?at-?elding),we also removed electronic crosstalk between Mosaic chip pairs sharing readout electronics,and the residual,large-scale imperfections in the sky?atness using a smoothed supersky?at derived from the science data.Cosmic rays were rejected in each exposure using the algorithm of Rhoads(2000).Astrometry of USNO-A2catalog stars was used to adjust the world coordinate systems of individual frames.Satellite trails were?agged manually for exclusion from the?nal stacks.Weights for image stacking were determined using the“ATTWEIGHT”algorithm(Fischer&Kochanski1994),and the task“mscstack”in the MSCRED package(Valdes1998)was used to stack the individual exposures.

The total integrated exposure time of the?nal stacked images are34.7ks for the815 nm?lter and36.0ks for the823nm?lter.The?nal stacked images have seeing of0.9′′, while the seeing of823nm band image is slightly better(0.02′′)than that of815nm band image.We also used deep B,V,I broad band images of ESO Imaging Survey(EIS)in the CDF-S,obtained using the Wide Field Imager(WFI)of the ESO-MPG2.2meter telescope at La Silla.The data were download from ESO archive,and registered to the narrow band images for candidate selection.The overlap between the narrowband stacks and the EIS images was31×33arcminutes in size with a gap of10.5×13arcminutes due to the missing chip.The total solid angle searched for Lyαgalaxies was thus887square arcminutes.The corresponding volume over the full redshift range,5.67≤z≤5.80,is3.06×105Mpc3.A summary of the images is given in Table1.

3.Candidate z=5.7LyαEmitters

Source detections were performed using SExtractor version2.2.2(Bertin&Arnouts 1996)on the narrowband images,and measured their colors using SExtractor’s two-image mode to measure the EIS B,V,I band photometry for the narrowband detected sources.All ?uxes are measured in2.43′′(9pixel)diameter aperture.

Candidate z~5.7Lyαemitters were selected following the same criteria that have proven highly successful in the LALA?elds(Rhoads&Malhotra2001;Rhoads et al.2003, 2004).These are:1)Narrowband detection>5σ;2)A narrowband excess of>0.75mag-nitude,so that>50%of the narrowband?ux comes from an emission line;3)Signi?cance of the narrowband excess>4σ;and4)<2σdetection in?lters blueward of the expected Lyman break location(B,V band in this letter).The success rate of the criteria has been spectroscopically con?rmed to be≥70%(Rhoads et al.2003,Dawson et al.2004,Dawson et al.2005in prep).

In order to make full use of our narrow band images,which are deeper than the I band image,we ran our selection using the weighted mean?ux from I and815nm band as the underlying continuum for823nm selection,and the weighted mean?ux from I and823nm band as the continuum for815nm selection.A total of17candidates are selected,5in the 815nm band,and12in the823nm band.In Fig.1we plot the underlying continuum and narrow band?uxes for all detected sources and candidate z~5.7Lyαemitters selected. Our selection criteria are also plotted in the?gure:the vertical line stands for5σdetection in narrow band;the solid inclined line represents a narrow excess of0.75magnitude and the dashed line corresponds to the requirement of4σsigni?cance of the narrowband excess.

https://www.sodocs.net/doc/7a16528679.html,rge scale structure

Among the17candidate z~5.7Lyαemitters we detected in CDF-S,12of them are selected in the823nm band image,5of them in the815nm band image.In Fig.2we plot the sky distribution of the17candidate z~5.7Lyαemitters in CDF-S.An obvious structure of the distribution can be seen in the?gure:almost all the17sources are located in the lower left half of the?eld.

Before studying the clustering properties of the17Lyαemitters,we consider whether the structure could be an observational artifact.Such a spurious“structure”in the sky distribution of the selected Lyαemitters could be due to two reasons:It might be due to a di?erence in narrow-band sensitivity across the?eld.It might also due to the sky distribution of very bright sources in the?eld,which could a?ect our source detection and candidate selection.

To check these possibilities,we?rst divide the823nm band image into four quarters and plot the source counts for the four parts in Fig.3.We can see that the source counts in four regions are consistent within the Poisson counting uncertainties.In region a where we identi?ed the most Lyαcandidates,the overall source density is among the lowest.This indicates that the detection e?ciency in region a is not higher than in other3regions.We found a similar result in the815nm band image.

Second,we added arti?cial point sources with narrowband magnitude of24.3(the mean value of the12Lyαemitters in the823nm band image)at random locations in the narrow-band image.We ran exactly the same procedures of source detection and candidate selection on the derived narrowband images.The recovered arti?cial sources passing our criteria show no evidence of strong clustering.Inserting arti?cial sources with di?erent magnitudes(the highest and the lowest magnitudes of the12sources)yields the same results.Likewise,no sign of clustering is found among arti?cial sources inserted and recovered from the815nm band image.A two-dimensional Kolmogorov-Smirno?test shows that the sky distribution of the12Lyαemitters at823nm is di?erent from the corresponding simulation at the99.4% con?dence level.Including the5Lyαemitters from the815nm band would slightly increase the signi?cance to99.6%.The sky distribution of the5Lyαemitters in815nm band alone is di?erent from the simulated sources at the level of98%.We checked the limiting magnitudes of the823nm,I,B and V band images over the whole?eld,and found no inhomogeneity. The photometric zero points of the narrow and broad band images are also found to be accurate within0.1mag over the whole?eld.We point out that there is no evidence of signi?cant spatial variations in bandpass of the narrow band?lters either.All the above checks indicate the clustering of the Lyαcandidates is not arti?cial.

We see clear evidence of large scale clustering of the12Lyαemitters in the823nm band. The same trend is also seen in the815nm band,but with lower con?dence level because the number of Lyαemitters in815nm band is only one third of that in823nm band.In Fig.4 we compare the source counts from the815nm and823nm band images.We can see that the detection e?ciencies are consistent with each other above the5σdetection limit,which is24.45mag for the815nm band image,and~0.2mag deeper for the823nm band.Note there are2Lyαemitters in823nm band with narrow band magnitudes fainter than24.45. Removing the2sources,the space density of Lyαemitters in823nm band is still twice of that in815nm band.The corresponding limiting Lyα?uxes are2.00&1.66×10?17ergs/cm2/s for815nm and823nm bands respectively.Based the luminosity function of z~5.7Lyαemitters given by Malhotra&Rhoads(2004),we predict the numbers of Lyαemitters in the two narrow band images to be4.5&7.5.This corresponds to an overdensity factor of1.6 for823nm band candidates.In the densest quadrant of the survey,the overdensity factor reaches3-4.

5.Conclusions

The structure we have found in the Lyαgalaxy distribution at z≈5.8is the most distant large scale feature yet reported in the spatial distribution of galaxies.The space density of these galaxies varies by a factor of two between our two narrowband?lters,and the projected surface densities span a substantially larger range(up to4×the expected mean).

The Hubble Ultra-Deep Field(HUDF;Beckwith et al2005in prep)lies near the edge of this large scale galaxy sheet,and the e?ects of this overdensity are also apparent there.An overdensity at z≈5.9near the HUDF was?rst suggested by Stanway et al(2004)on the basis of3redshifts in that region.This has now been con?rmed in the redshift distribution of23i-dropout Lyman break galaxies in this?eld(Malhotra et al2005),as measured using slitless spectroscopy from the GRAPES project(Pirzkal et al2004).The redshift extent of the feature within the UDF is larger than the coverage of our narrowband?lters,suggesting that the structure is more complex than a single sheet.

We would like to thank Bahram Mobasher for providing the calibrated photometric zero points for the EIS images.The work of JW was supported by the CAS”Bai Ren”project in USTC and the CXC grant GO3-4148X.

REFERENCES

Ajiki,M.et al.2004,PASJ,56,597

Bertin,E.,Arnouts,S.1996,A&AS,117,393

Cowie,L.L.,&Hu,E.M.1998,AJ,115,1319

Dawson,S.et al.2004,ApJ,617,707

Foucaud,S.et al.2003,A&A,409,835

Fynbo,J.U.,Moller,P.,&Thomsen,B.2001,A&A,374,443

Hu,E.M.,Cowie,Lennox L.,&McMahon,Richard G.1998,ApJ,502,L99

Hu,E.M.,et al.2002,ApJ,568,L75

Hu,E.M.,Cowie,L.L.,Capak,P.,McMahon,R.G.,Hayashino,T.,Komiyama,Y.2004, AJ,127,563

Kodaira,K.et al.2003,PASJ,55,L17

Kudritzki,R.-P.et al.2000,ApJ,536,19

Kurk,J.D.,Cimatti,A.,di Serego Alighieri,S.,Vernet,J.,Daddi,E.,Ferrara,A.,Ciardi,

B.2004,A&A,422,L13

Malhotra,S.&Rhoads,J.E.2004,ApJL,617,5

Malhotra,S.,Rhoads,J.E.,Pirzkal,N.,Haiman,Z.,Xu,C.et al.2005,ApJ submitted Miley,G.K.et al.2004,Nature,427,47

Ouchi,M.et al.2001,ApJ,558,83

Ouchi,M.et al.2003,ApJ,582,60

Ouchi,M.et al.2004,ApJ,611,685

Palunas,P.,Teplitz,H.I.,Francis,P.J.,Williger,G.M.,Woodgate,B.E.2004,ApJ,602, 545

Pentericci,L.et al.2000,A&A,361,25

Pirzkal,N.,et al2004,ApJS,154,501

Rhoads,J.E.2000,PASP,112,703

Rhoads,J.E.,Malhotra,S.,Dey,A.,Stern,D.,Spinrad,H.,&Jannuzi,B.T.2000,ApJ, 545,L85

Rhoads,J.E.&Malhotra,S.2001,ApJ,563,L5

Rhoads,J.E.et al.2003,AJ,125,1006

Rhoads,J.E.et al.2004,ApJ,611,59

Shimasaku,K.et al.2003,ApJ,586,L111

Spergel,D.N.et al.2003,ApJS,148,175

Stanway,E.,et al2004,ApJ,604,L13

Steidel,C.C.,Adelberger,K.L.,Dickinson,M.,Giavalisco,M.,Pettini,M.,Kellogg,M.

1998,ApJ,492,428

Steidel,C.C.,Adelberger,K.L.,Shapley,A.E.,Pettini,M.,Dickinson,M.,Giavalisco,M.

2000,ApJ,532,170

Tran,K.H.,Lilly,S.J.,Crampton,D.,Brodwin,M.2004,ApJ,612,L89

Valdes,F.1998,ADASS VII,ASP Conf.Series145,ed.R.Albrect

Venemans,B.P.et al.2002,ApJ,569,L11

Venemans,B.P.et al.2004,A&A,424,L17

Table1.Optical images

Band Telescope Exposure Time(ks)m AB(lim)a

a Limiting magnitudes(AB)for a3σdetection on a

2.0′′diameter aperture.

Fig.1.—Underlying continuum?uxes vs narrow band?uxes for SExtractor detected sources in the two narrow band images.Our candidate z~5.7Lyαemitters(5in815nm band,and 12in823nm band)are marked as big dots.Lines show our selection criteria:the vertical line stands for5σdetection in narrow band;the solid inclined line represents a narrow excess of0.75magnitude(F lux narrow/F lux continuum=2)and the dashed line corresponds to the requirement of4σsigni?cance of the narrowband excess which is F lux narrow-F lux continuum =4×

Fig.2.—Sky distribution of the17photometrically selected z~5.7Lyαemitters in CDF-S. The shaded region in the plot is due to the Chip3of the Mosaic II camera which was turned o?during the observations.North is up and east is to the left.Solid and open circles are candidates selected from823nm and815nm band respectively.See text for details.

Fig. 3.—The source number counts in the823nm band image,which was divided into 4quarters.The numbers of detected sources in each quarter agrees well with each other, showing that the clustering of our z=5.7candidates is not due to the di?erent detection e?ciency over the?eld.(The counts in quadrant c are adjusted for the reduction in solid angle due to the missing chip.)

Fig.4.—Comparing the source number counts in the815nm and the823nm band images. The vertical lines stand for the5σdetection in the narrow bands as one of our selection criteria.

相关主题