Detection of an H-alpha Emission Line on a Quasar, RX J1759.4+6638, at z=4.3 with AKARI

a r X i v :0706.0253v 1 [a s t r o -p h ] 2 J u n 2007Detection of an H αEmission Line on a Quasar,RX

J1759.4+6638,at z =4.3with AKARI

Shinki Oyabu 1,*,Takehiko W ada 1,Youichi Ohyama 1,Hideo Matsuhara 1,Toshinobu Takagi 1,Takao Nakagawa 1,Takashi Onaka 2,Naofumi Fujishiro 1,Daisuke Ishihara 2,Yoshifusa Ita 1,Hirokazu Kataza 1,Woojung Kim 1,

Toshio Matsumoto 1,Hiroshi Murakami 1,Itsuki Sakon 2,Toshihiko Tanab ′e 3,Kazunori Uemizu 1,Munetaka Ueno 4,Fumihiko Usui 1,Hidenori W atarai 5,

and Kanae Haze 1

*e-mail:

oyabu@ir.isas.jaxa.jp 1Institute

of Space and Astronautical Science,Japan Aerospace Exploration Agency,3-1-1Yoshinodai,Sagamihara,Kanagawa 229-8510,Japan 2Department of Astronomy,School of Science,University of Tokyo,Bunkyo-ku,Tokyo 113-0033,Japan 3Institute of Astronomy,University of Tokyo,2-21-1Osawa,Mitaka,Tokyo,181-0015,Japan 4Department of Earth Science and Astronomy,University of Tokyo,3-8-1,Komaba,Megro-ku,Tokyo 153-8902,Japan 5O?ce of Space Applications,Japan Aerospace Exploration Agency,Tsukuba,Ibaraki,305-8505,Japan (Received —;accepted —)Abstract We report the detection of an H αemission line in the low resolution spectrum of a quasar,RX J1759.4+6638,at a redshift of 4.3with the InfraRed Camera (IRC)onboard the AKARI.This is the ?rst spectroscopic detection of an H αemission

line in a quasar beyond z=4.The overall spectral energy distribution (SED)of RX J1759.4+6638in the near-and mid-infrared wavelengths agrees with a median SED of the nearby quasars and the ?ux ratio of F (Ly α)/F (H α)is consistent with those of previous reports for lower-redshift quasars.

Key words:galaxies:quasars:emission lines —galaxies:quasars:individual(RX J1759.4+6638)—infrared:spectroscopy

1.Introduction

The broad emission lines of high-redshift quasars are luminous enough to study physical properties of quasars such as central black hole mass,accretion rate and metallicity of broad emission line regions in the early universe.Recent optical observations of emission lines such as C IV and N V reveal supersolar abundances in quasar broad emission line regions even at z>4 (Dietrich et al.2003).The lack of evolution of the Fe II/Mg II UV emission line ratio of quasars apparently continues out to z=6.4(Barth et al.2003;Iwamuro et al.2004).The rest-frame optical Fe II emission as well as hydrogen Balmer emission lines in high-redshift quasars cannot be measured from the ground facilities.

These measurements can be pursued with the AKARI(Murakami et al.2007)which has a unique capability to take near-infrared spectra in2?5μm from the space as well as mid-infrared(5-14,18-26μm)with the InfraRed Camera(IRC;Onaka et al.2007;Ohyama et al. 2007).Therefore one can trace the redshifted emission lines toward the high-redshift universe with the AKARI.

A quasar,RX J1759.4+6638,at a redshift of4.320was discovered as the most distant ROSAT X-ray selected object known at the time by Henry et al.(1994)near the North Ecliptic Pole(NEP).In this paper,we report the near-infrared spectroscopy of this quasar and the detection of an Hαemission line at z=4.3.Previous studies(Espey et al.1989;Nishihara et al.1997)performed the measurements of the Hαemission line in high-redshift quasars in the near-infrared spectroscopy and the redshifts of their sample only reached z～2.4because the wavelength of redshifted Hαemission lines is a?ected by strong thermal emission of a telescope. Therefore this is the?rst spectroscopic detection of an Hαemission line in a quasar beyond z=4,while Hαdetections in photometric observations are reported by Jiang et al.(2006)for quasars at z～6by using the Spitzer IRAC.Our spectroscopic detection is caused by a bene?t of high sensitivity spectroscopy with a cooled telescope from the space.

Through this paper,we adopt a?at cosmology with H0=71km s?1Mpc?1,?=0.27 andλ=0.73(Spergel et al.2003).

2.Observation and Data Reduction

RX J1759.4+6638was observed by the AKARI IRC on2006April,May,October and 2007February.As summarized in Table1,we made six pointed observations using the IRC spectroscopic mode AOT04.There are three observing modes:NP,NG,and NPNG modes. The NP mode uses the near-infrared prism with spectral resolving power R～19at3.5μm.In this mode,a target is put on the imaging area of near-infrared detector.In the NP mode,a MIR-S spectroscopic observation of a target is performed simultaneously.The NG mode uses the near-infrared grism(R～120at3.6μm),in which a target is put on a small aperture for a point source grism spectroscopy.The other,NPNG,is a special observing mode for a calibration

Table1.Observing Log

Obs.ID Obs.Date Obs.mode?

Note.

?NP is a spectroscopy on the imaging area of

detector with the near-infrared prism for reso-

lution,R～20,and NG is a spectroscopy on the

slit of“Np”position with R～80of the near-

infrared grism.NPNG is prepared using prism

and grism in the?rst half observation and the

second,respectively.

using the prism and the grism in the?rst half observation and the second,respectively.The ?rst four observations with the NPNG mode were done for the wavelength calibration using the planetary nebula NGC6543in the performance veri?cation phase.During these observations, the spectra of RX J1759.4+6638were obtained serendipitously.The remaining observations were targeted observations for this quasar using the AKARI Director Time in order to con?rm the serendipitous detections in the performance veri?cation phase.

The data were processed through the IRC Spectroscopic toolkit(Ohyama et al.2007)to produce calibrated data frames.The data were converted into dark-subtracted,linearity cor-rected and?at-?eld corrected frames after data that had strong cosmic rays on the target object were removed with visual investigation.Multiple frames were combined,and one-dimensional background-subtracted spectra of the target quasar were extracted from the combined data with the aperture width of3pixels(4.5′′in the NIR).The resultant spectra were scaled by a factor of1.6for the aperture correction.In addition to the error estimation by the IRC Spectroscopic toolkit using the sky variation near a target,we also calculated the root mean square of signal between each frames.The original error estimation was recognized to be overestimated due to the contamination of other sources and the variations of signals between each frame were used as the background error of spectra here.

The photometric calibration of the spectrum was based on AKARI IRC observations of standard stars,while the wavelength calibration was based on observations of emission line stars and planetary nebulae(Ohyama et al.2007).At this time we estimate that the overall uncertainty in the wavelength calibration is～0.05μm at3.5μm.

3.Result

The series of PV phase observations(ID5020047.1,5020048.1,5020049.1and5020050.1) detected the Hαemission line only tentatively because the contamination of a faint near-infrared source a?ected positions of the Hαemission line.The targeted NP observation(ID5124044.1) was arranged after we checked that the dispersion direction was free from any other object using the deep N3-band image of the AKARI North Ecliptic Pole(NEP)Survey(Figure1; Oyabu et al.2007).The NG spectroscopy(ID5124035.1)and all MIR-S data,which were taken simultaneously during NP observations,provided spectra with low signal-to-noise ratio only.

Figure2presents the resultant observed spectrum of the NP observation on the top panel with a black thick line.On the spectrum,there are two features;a bump at the wavelength of 3.47μm and a dip on the shorter wavelength side of the bump.The bump is located at the Hαemission line corresponding to z=4.29.For the check of their reliabilities,we divided their data in half and reduced them separately.On the top panel of Figure2,two spectra of?rst and second half data are also shown in blue and red lines,respectively.Both the dip and the bump are shown in both data sets.In addition,we also checked that the response function of the NP prism did not produce these features as shown on the bottom panel of Figure2.Thus both features,the bump and the dip,are real.The other possible explanation of the bump is the contamination of other sources.However,as mentioned above,we arranged the NP observation when the dispersion direction was free from any other objects.There is no contamination in the dispersion direction as shown in Figure1,and therefore we conclude that the detection of the Hαemission line is reliable.

To determine the wavelength center and the full width half maximum(FWHM),we?t a Gaussian function to this spectrum after smoothing the spectrum with3pixels(Figure2). The detail measurements of the Hαemission line are summarized in Table1.The redshift from the Hαemission line is found to be z=4.29±0.06,in agreement with the previous measurements of redshifts,z=4.320(Henry et al.1994)and4.32(Constantin et al.2002)by using restframe-ultraviolet emission lines from the ground-base optical spectroscopy.The Hαline?ux is4.9±1.1×10?22W cm?2,corresponding to the luminosity of Hαemission line L(Hα)=7.5±1.7×1036W at z=4.3.While the Hαemission line?ux is a?ected by two[N II] lines located at blue and red side of the Hαline,their contribution of[N II]emission lines to the Hαline?ux is only3percents in the Sloan Digital Sky Survey(SDSS)composite(Vanden Berk et al.2001)and is ignored in this paper.The FWHM of the Hαemission line is found to be unresolved and18000±4000km s?1which is comparable to the instrumental resolution ～20000km s?1when we made the3-pixel smoothing data.The restframe equivalent width (EW)of the Hαemission line is0.071±0.015μm which is converted from the observed EW of 0.38±0.08μm at z=4.3.However there is a dip of the continuum around3.2μm in Figure2

.4+6638

E

10"

N

Fig.1.The image from the NEP-Deep in the N3-band(Oyabu et al.2007).The image is rotated as the dispersion direction is upper at2007Feb.9th as the spectrum position is shown with a green box.

and the dip makes it di?cult to measure the Hαline?ux and the EW accurately.This dip might be a broad absorption line feature,while there are no features of broad absorption lines in the restframe UV spectra(Henry et al.1994;Constantin et al.2002).Thus the reason of this dip is still uncertain.

We also measured the NIR and MIR-S photometric?uxes of this quasar with the IRC during the course of the AKARI NEP-Deep Survey(Oyabu et al.2007)1.The results of the NIR and MIR-S bands are summarized in Table3.Figure3presents the comparison of the IRC spectrum of RX J17759.4+6638with IRC photometric result as well as the median spectral energy distribution(SED)of low-redshift quasars(Elvis et al.1994).Assuming that the N3-band?ux includes the Hαemission line?ux and the N2-and N4-bands represent continuum level,the?ux of the Hαemission line,F(Hα)=4.0±2.6×10?22W cm?2,from the photometric data agrees with the spectroscopic measurement of the Hαemission line.The uncertainty of the line?ux from the photometric data is dominated by the systematic errors of photometric calibration.The observed?uxes of emission lines and continuum level are consistent with the photometric study within their https://www.wendangku.net/doc/528785739.html,paring our measurements with the median SED of low-redshift quasars(Elvis et al.1994),RX J1759.4+6638has quite typical spectral energy distribution,which suggests no signi?cant evolution in the rest-frame optical-and near-infrared wavelengths.

-0.10.0

0.1

0.2

0.3

F l u x (m J y )H α

2.5

3.0 3.5

4.0 4.5

Observed Wavelength(μm)0.00.2

0.4

0.60.81.0R e s p o n s e Fig.2.The NP spectrum of RX J1759.4+6638(black solid line;top).Blue and red lines indi-cate the result from ?rst and second half of data,respectively.The vertical lines are error bars consisting of background variation,wavelength calibration,?atten calibration and ?ux calibration er-rors.The dashed and dotted lines show the ?tted Gaussian and the assuming continuum,respec-tively.For reference,a normalized response curve of the NP as a function of wavelength (bottom).

1

10

Observed Wavelength(μm)-0.1

0.0

0.1

0.2

0.3F l u x (m J y )0.5 1.0 2.0 3.0

Rest Wavelength(μm)https://www.wendangku.net/doc/528785739.html,parison of the spectrum of RX J1759.4+6638with the photometric study.The solid line is the IRC spectrum as shown in Figure 1with 1σerror bars,while red diamonds presents the pho-tometric result from Oyabu et al.(2007).The red horizontal and vertical bars on each point presents the band width and the 1σphotometric error,respectively.For comparison,the median spectral en-ergy distribution of low-redshift quasars (Elvis et al.1994)is plotted in a dotted line after scaled at the N4-band.In Oyabu et al.(2007),this quasar is called as CXOSEXSI J175928.1+663851.

Table2.Observed Hαline measurement of RX J1759.4+6638

Line Observed Wavelength Redshift Observed Flux?FWHM?Observed EW?

(μm)(10?22W cm?2)(km s?1)(μm)

Note.

?Flux is measured by direct integration of the line?ux.

?The observed full width half maximum(FWHM)of emission line.

?The observed equivalent width(EW)of emission lines.

?The observed FWHM is unresolved with the instrumental resolving power～20000km s?1after3pixel smoothing at3.5μm.

Table3.The AKARI/IRC photometric measurements of RX J1759.4+6638from Oyabu et al.(2007)

Band(λref)Flux(μJy)Flux error(μJy)?

Note.

?Only statistical errors of the?uxes are presented.About20

percents of the systematic errors exist during the photometric

calibrations.

4.Discussion

The?ux ratio of the observed Hαemission line to Lyαwould be useful to investigate the physical conditions in the broad-line region and the reddening to this region.Henry et al.(1994)observed the rest-frame ultraviolet spectra from ground-based telescopes,and they measured the Lyαemission line?ux of F(Lyα)=1.4×10?21W cm?2on June1993.Constantin et al.(2002)made a new measurement of F(Lyα)=6.4×10?22W cm?2on June1999with the Keck Telescope2.The Lyαline?ux from Constantin et al.(2002)changed into less than half of the measurements in Henry et al.(1994)on a timescale of about1yr(rest frame)suggesting that this quasar is variable.We note that the X-ray observations of this quasar(Grupe et al.2006)also reported variability.In addition,the line pro?le of Lyαis strongly a?ected by the intervening column of intergalactic neutral hydrogen(Henry et al.1994;Constantin et al. 2002).These problems make straightforward comparisons between Hαand Lyαline?uxes di?cult.However assuming that the Lyαemission line?ux in Constantin et al.(2002)be minimum,the lower limit,F(Lyα)/F(Hα)>1.3,is calculated.This ratio is consistent with other observational results;F(Lyα)/F(Hα)=3.2of the SDSS composite quasar(Vanden Berk et al.2001)and F(Lyα)/F(Hα)=1.2?7.4of the low-redshift quasars at z=0.061-0.555

(Tsuzuki et al.2006),although the theoretical values of F(Lyα)/F(Hα)=10?12from pure recombination calculation(Osterbrock&Ferland2006)are greater than observations.

AKARI will extend the spectroscopic sample of high-redshift quasars’spectroscopy for more detailed studies of Hαemission lines at z>4.Speci?cally,one of the AKARI Mission Programs,“The unbiased Slit-less spectroscoPIC surveY of galaxies(SPICY)”(Matsuhara et al.2006),is designed and conducted to make～0.5square degree scale survey with slit-less spectroscopy with AKARI IRC wavelength and will provide a new sample of Hαemission lines in the high-redshift universe.

5.Summary

We present the detection of an Hαemission line on the low resolution spectrum of a quasar RX J1759.4+6638at a redshift of4.3with the IRC onboard the AKARI after careful consideration of possible artifact and contamination.This is the?rst spectroscopic detection of an Hαemission line in a quasar beyond z=4.Our spectroscopic measurement shows a good agreement with the photometric data from the AKARI NEP-Deep Survey within their uncertainties.The overall SED of RX J1759.4+6638in the near-and mid-infrared wavelengths also agrees with a median SED of the nearby quasars.The?ux ratio of F(Lyα)/F(Hα)of this quasar is consistent with those of previous report for lower-redshift quasars.These results suggest no signi?cant evolution of the quasar’s features at z=4.3in the optical-and near-infrared wavelength(rest frame).

AKARI is a JAXA project with the participation of ESA.We thank all the members of the AKARI project for their continuous help and support.

References

Barth,A.J.,Martini,P.,Nelson,C.H.,&Ho,L.C.2003,ApJL,594,L95

Constantin,A.,Shields,J.C.,Hamann,F.,Foltz,C.B.,&Cha?ee,F.H.2002,ApJ,565,50 Dietrich,M.,Hamann,F.,Shields,J.C.,Constantin,A.,Heidt,J.,J¨a ger,K.,Vestergaard,M.,& Wagner,S.J.2003,ApJ,589,722

Elvis,M.,et al.1994,ApJS,95,1

Espey,B.R.,Carswell,R.F.,Bailey,J.A.,Smith,M.G.,&Ward,M.J.1989,ApJ,342,666 Grupe,D.,Mathur,S.,Wilkes,B.,&Osmer,P.2006,AJ,131,55

Henry,J.P.,et al.1994,AJ,107,1270

Iwamuro,F.,Kimura,M.,Eto,S.,Maihara,T.,Motohara,K.,Yoshii,Y.,&Doi,M.2004,ApJ,614, 69

Jiang,L.,et al.2006,AJ,132,2127

Matsuhara et al.2006,PASJ,58,673,

Murakami et al.2007,submitted to PASJ

Nishihara,E.,Yamashita,T.,Yoshida,M.,Watanabe,E.,Okumura,S.-I.,Mori,A.,&Iye,M.1997, ApJL,488,L27

Ohyama et al.2007,submitted to PASJ

Onaka et al.2007,PASJ,in press

Osterbrock,D.E.,&Ferland,G.J.,2006,Astrophysics of Gaseous Nebulae and Active Galactic Nuclei Second Edition(University Science Books)

Oyabu et al.2007,submitted to PASJ

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

Tsuzuki,Y.,Kawara,K.,Yoshii,Y.,Oyabu,S.,Tanab′e,T.,&Matsuoka,Y.2006,ApJ,650,57 Vanden Berk,D.E.,et al.2001,AJ,122,549