Direct Lyman continuum and Lyman-alpha escape observed at redshift 4
E. Vanzella, M. Nonino, G. Cupani, M. Castellano, E. Sani, M. Mignoli, F. Calura, M. Meneghetti, R. Gilli, A. Comastri, A. Mercurio, G. B. Caminha, K. Caputi, P. Rosati, C. Grillo, S. Cristiani, I. Balestra, A. Fontana, M. Giavalisco
MMNRAS , 000–000 (0000) Preprint 2 April 2018 Compiled using MNRAS L A TEX style file v3.0
Direct Lyman continuum and Lyman α escape observed atredshift 4 E. Vanzella (cid:63) , M. Nonino , G. Cupani , M. Castellano , E. Sani , M. Mignoli ,F. Calura , M. Meneghetti , R. Gilli , A. Comastri , A. Mercurio , G. B. Caminha ,K. Caputi , P. Rosati , , C. Grillo , , S. Cristiani , I. Balestra , A. Fontana , andM. Giavalisco INAF – Osservatorio Astronomico di Bologna, via Gobetti 93/3, 40129 Bologna, Italy INAF – Osservatorio Astronomico di Trieste, via G. B. Tiepolo 11, I-34143, Trieste, Italy INAF – Osservatorio Astronomico di Roma, Via Frascati 33, I-00078 Monte Porzio Catone (RM), Italy European Southern Observatory, Alonso de Cordova 3107, Casilla 19, Santiago 19001, Chile INAF – Osservatorio Astronomico di Capodimonte, Via Moiariello 16, I-80131 Napoli, Italy Kapteyn Astronomical Institute, University of Groningen, Postbus 800, 9700 AV Groningen, The Netherlands Dipartimento di Fisica e Scienze della Terra, Universit`a degli Studi di Ferrara, via Saragat 1, I-44122 Ferrara, Italy Dipartimento di Fisica, Universit`a degli Studi di Milano, via Celoria 16, I-20133 Milano, Italy Dark Cosmology Centre, Niels Bohr Institute, University of Copenhagen, Juliane Maries Vej 30, 2100 Copenhagen, Denmark University Observatory Munich, Scheinerstrasse 1, D-81679 Munich, Germany Astronomy Department, University of Massachusetts, Amherst, MA 01003, USA
ABSTRACT
We report on the serendipitous discovery of a z = 4 . M = − .
20 star-forminggalaxy (
Ion3 ) showing copious Lyman continuum (LyC) leakage ( ∼
60% escaping),a remarkable multiple peaked Ly α emission, and significant Ly α radiation directlyemerging at the resonance frequency. This is the highest redshift confirmed LyCemitter in which the ionising and Ly α radiation possibly share a common ionisedchannel (with N HI < . cm − ). Ion3 is spatially resolved, it shows clear stellarwinds signatures like the P-Cygni N v λ β = − . ± . F λ ∼ λ β ) with weak low-ionisation interstellar metal lines. DeepVLT/HAWKI Ks and Spitzer/IRAC 3.6 µm and 4.5 µm imaging show a clear photo-metric signature of the H α line with equivalent width of 1000˚A rest-frame emergingover a flat continuum (Ks − . µm (cid:39) . × M (cid:12) , SFR of 140 M (cid:12) yr − and age of ∼
10 Myr, with a low dust extinction,E(B-V) (cid:46) .
1, placing the source in the starburst region of the SFR − M ∗ plane. Ion3 shows similar properties of another LyC emitter previously discovered ( z = 3 . Ion2 ,Vanzella et al. 2016).
Ion3 (and
Ion2 ) represents ideal high-redshift reference cases toguide the search for reionising sources at z > . Key words: galaxies: formation – galaxies: starburst – gravitational lensing: strong
The definition of a reference sample of Lyman continuum(LyC) emitters at z (cid:46) . z > . f absesc ∼ −
15% have been confirmed in the nearby universe, (cid:63)
E-mail: [email protected] z ∼ − . f esc (cid:39) f esc < . . < z < . f esc ∼ − c (cid:13) a r X i v : . [ a s t r o - ph . GA ] M a r E. Vanzella et al.
Figure 1.
The FORS spectrum at resolution dv (cid:39)
580 km s − (grism 300V) of Ion3 is shown along with the most relevant ul-traviolet lines. In the top-left insets the Ly α spectra at resolution dv (cid:39) − obtained with the grism 600B(300V) areshown. The two-dimensional signal to noise and sky spectra areshown in the middle of the figure, in which the wavelength cov-erage up to 9300˚A is well sampled, despite the fringing patternat λ > S/N >
10 is clearly detected. the available sample of LyC leakers, both at low and highredshift, show very consistent observational features. Suchfeatures include the high Ly α equivalent width EW > iii ] λ ii ] λ α emission like narrow double peaked profiles( < − − ) or Ly α emission close to the systemicvelocity ( < − ). Also intense nebular optical lines[O iii ] λλ , β with EW of 1000˚A − f esc , aswas predicted by radiation transfer models (Verhamme et al.2015) and photoionisation models (e.g., Jaskot & Oey 2013;Nakajima & Ouchi 2014; Zackrisson et al. 2013). In thiswork we report on a serendipitously discovered LyC emitterat redshift 4, dubbed here Ion3 , the highest redshift casecurrently known. We assume a flat cosmology with Ω M =0.3, Ω Λ = 0.7 and H = 70 km s − Mpc − . Ion3 was discovered during a FORS2 spectroscopic programexecuted in visitor mode during the period 15 −
19 Septem-ber 2017 (prog. 098.A-0804(B), P.I. Vanzella).
Ion3 is a rel-atively bright object and was inserted as a filler in the MXUFORS mask (I-band magnitude 23 . ± .
38) and located at 3 (cid:48) (cid:48)(cid:48) from the Frontier Field galaxy cluster AS1063 (Lotzet al. 2017), which lies outside the HST coverage. Giventhe large separation from the galaxy cluster the resultingmagnification is low, a well constrained µ = 1 . ± .
02 (or∆ m = 0 .
15, Caminha et al. 2016). In the following all thereported quantities are corrected for µ .The data reduction was carried on as described in sev-eral previous works (e.g., Vanzella et al. 2014) in which theAB-BA sky subtraction scheme was implemented. The fi-nal spectrum consists of 14 hours integration with an av-erage seeing of 0 . (cid:48)(cid:48) and spectral resolution dv (cid:39)
600 kms − at λ = 7000˚A (R= λ/dλ =500, grism 300V). Figure 1shows the FORS2 spectrum covering the wavelength range3700˚A − dv (cid:39)
300 km s − ,R=1000). We clearly confirmed the double peaked Ly α pro-file of Ion3 not well resolved with the 300V grating (see Fig-ure 1, top-left). Additional four hours X-Shooter integrationon
Ion3 was subsequently obtained during November 2017with an average seeing conditions of 0 . (cid:48)(cid:48) (prog. 098.A-0665,P.I. Vanzella), providing a final spectrum spanning the range3400˚A - 24000˚A with spectral resolution dv (cid:39) − − .We refer the reader to Vanzella et al. (2014, 2017) for detailsabout FORS and X-Shooter data reduction. The most intriguing feature emerging from
Ion3 is the LyCleakage at λ < − − z < . (cid:48)(cid:48) (see Vanzella et al. 2010). Second, there is no traceof any spectral line arising from a foreground object, both inthe deep FORS spectrum (that excludes [O ii ] λ , z < .
45 and Ly α at z > .
9) and wide X-Shooter spectrum,that easily would have captured several UV and/or opticalrest-frame emission lines in the redshift range 0 < z <
Ion3 support a very low col-umn density of neutral gas along the line of sight, makingthe entire picture consistent with the emerging LyC signal.From the FORS spectrum the observed fluxes at900˚A and 1500˚A rest-frame are F λ = (5 . ± . × − erg s − cm − ˚A − and F λ = (3 . ± . × − ergs − cm − ˚A − , respectively, corresponding to a flux densityratio of f ν (1500) /f ν (900) = 19 . ± .
3. Following Vanzellaet al. (2012) this ratio translates to a relative escape frac-tion fesc,rel = 20-100%, assuming an intrinsic ratio of theluminosity densities L ν (1500) /L ν (900) = 1 − MNRAS000
3. Following Vanzellaet al. (2012) this ratio translates to a relative escape frac-tion fesc,rel = 20-100%, assuming an intrinsic ratio of theluminosity densities L ν (1500) /L ν (900) = 1 − MNRAS000 , 000–000 (0000) he highest redshift stellar ioniser Figure 2.
The FORS spectrum is shown with the main spectralfeatures reported at the systemic redshift (blue crosses). In partic-ular the P-Cygni N v λ ii ] ∗ λ .
71, [S iv ] λ , ii λ v λ < T IGM > = 0 .
26 (with a central 68% interval of 0 . − . T IGM is unknown, therefore, any combina-tion of L ν (1500) /L ν (900) > fesc,rel in the range 10% − < T IGM > = 0 .
26 and L ν (1500) /L ν (900) = 3. The inferred ionising photon pro-duction rate from F λ is N phot (900) = 3 . × s − , whichcompared to Starburst99 models for instantaneous bursts(Salpeter IMF, Salpeter (1955), and upper mass limit of100 M (cid:12) , Leitherer et al. 2014) yields a stellar mass involvedin the starburst event of 4 × M (cid:12) with the age not largerthan 20Myrs, and a number of O-type stars dominating theionising radiation of (cid:39) . × (with uncertainties mainlydominated by the aforementioned IGM transmission). The continuum redward the Ly α line up to (cid:39) β = − . F λ ∼ λ β ). Such asteep slope is not significantly affected by the atmosphericdispersion (the atmospheric dispersion compensator, LADC,is part of the system at UT1), and is fully consistent with be-ing powered by massive and hot stars (O and early B stars),which are also responsible for the ionising photons. Thestrongest spectroscopic signature of these stars is providedby the broad P-Cygni profiles in the resonance transitionsof highly ionised species that arise in the stellar winds (e.g.,Leitherer et al. 2014). In our FORS spectrum a N v λ v λ Figure 3.
Top:
The one-dimensional X-Shooter spectrum of theLy α is shown (at dv (cid:39)
35 km s − ) in the velocity space (withreported the relative velocities among the peaks, km s − ). Thesystemic velocity is inferred from the [O ii ] λ , Bottom:
The X-Shooter two-dimensional spectrum is shown with the four mainstructures identified. In the bottom-right inset the same Ly α lineis shown at resolution dv (cid:39)
600 km s − as observed with FORS. al. 2014) spectral population synthesis model to the FORSspectrum with instantaneous burst with age 1, 10 and 20Myr old, 0.4 solar metallicity (Salpeter IMF and 100 M (cid:12) upper mass limit), which reproduces the observations well(especially the 1-10 Myr old templates, Figure 2). Low ioni-sation interstellar absorption lines like [Si ii ] λ i ] λ ii ] λ Ion3 also shows the non-resonant fluorescent emission line [C ii ] ∗ λ .
71 (see Fig-ure 2). Such a feature has been detected by Jaskot & Oey(2014) on local LyC candidates and interpreted as an evi-dence of the complex geometry of the neutral gas outside theline of sight, like anisotropic ionising emission. The systemicredshift z sys has been derived from the [C ii ] ∗ λ .
71 lineand more accurately from the X-Shooter detection of the[O ii ] λ , z sys = 3 . ± . V ] λλ , ii λ σ . Table 1summarises the most relevant spectroscopic properties. α emission In the conventional scenario for Ly α emission, Ly α scat-ters many times before escaping, which significantly altersand broadens the original line profile. The kinematics, thecolumn density and geometry of the H i gas are the mainingredients that shape the Ly α emission (not to mentionthe dust attenuation). A prominent Ly α emission line withrest-frame EW= 40 ± MNRAS , 000–000 (0000)
E. Vanzella et al.
Table 1.
Spectral and physical properties of
Ion3 . Line fluxesare reported in units of 10 − erg s − cm − (no slit losses areconsidered). Quantities are corrected for the lensing magnification µ = 1 . σ z is the redshift error on the last digit.Line/ λ vacuum Flux( SN ) z[ σ z ], Resolution( kms )Ly α (0) 1215.7 0.37(5) 3.9902[4], 35 (XSHO)Ly α (1) 1215.7 1.16(12) 3.9950[3], 35 (XSHO)Ly α (2) 1215.7 1.37(15) 3.9984[3], 35 (XSHO)Ly α (3) 1215.7 6.23(32) 4.0033[3], 35 (XSHO)Ly α (total) 9.13(73) − , 580 (FORS)[C ii ] ∗ λ .
71 0.14(4) 4.000[4], 580 (FORS)[He ii ] λ .
42 0.12(2.7) 4.000[4], 580 (FORS)[O ii ] λ .
09 0.38(4.5) 3.999[1], 55 (XSHO)[O ii ] λ .
88 0.25(3) 3.999[1], 55 (XSHO)SED-fitting output value uncertaintyM(stellar) [ × M (cid:12) ] 1.5 [1 . − . − M (cid:12) yr − ] 140 [110-150]E(B-V) (cid:39) . − . M UV (1500) − . ± . α EW(rest) [˚A] (cid:39) − s − ] 3 . × [10 − ]log ( ξ ion [Hz erg − ]) 25.6 25 . − . f esc,rel . − . X-Shooter spectral resolution ( dv = 35 km s − ), in whichwe identify four emitting structures marked as 0, 1, 2 and3. While the presence of blue peaks typically suggest a lowcolumn of H i gas (e.g., Henry et al. 2015; Yang et al. 2016),the emission at peak (2) is remarkable and emerges at theresonance frequency ( z = 3 . (cid:46) − fromthe systemic) where the opacity to Ly α photons would bethe highest (Figure 3). This is fully consistent with a sce-nario in which the LyC and (part of) the Ly α photons areescaping along the same optically thin direction (to LyC, N HI < . cm − ) and likely from the same cavity (e.g.,Behrens et al. 2014; Verhamme et al. 2015; Zackrisson etal. 2013). The LyC − Ly α escape through a ionised chan-nel discussed in Behrens et al. (2014) consider the possi-bility that the gas is outflowing perpendicular to a galacticdisk (and is reminiscent of a wind breaking out of a galacticdisk). In this scenario the quadruply peaked Ly α emissionobserved in Ion3 might be associated to a face-on disk.
Ion3 represents the highest redshift empirical evidence of sucha LyC − Ly α escaping mode. The Ly α escape degenerateswith the H i column and the outflow velocity such that fastwinds can mimic low columns (Verhamme et al. 2015). Inthis case both the LyC emission and a relatively fast windare detected. The evidence of an outflowing gas is imprintedin the blueshifted interstellar metal lines [S iv ] λ .
76 and[S iv ] λ .
77, with an average dv (cid:39) − ± − . Thisis also consistent with the well developed red tail of peak (3)possibly suggesting backscattering from the receding gas.Interestingly, a very similar direct Ly α escape has re-cently been identified in a lensed z = 2 .
37 galaxy by Rivera-Thorsen et al. (2017), in which the central peak is also atthe systemic redshift and indicative of a possible perforatedchannel of very low H i optical depth. SED fitting has been performed using BC03 templates(Bruzual & Charlot 2003) including the nebular prescriptionand assuming exponentially declining star formation histo-ries with e -folding time 0 . < τ <
15 Gyr, (see Castellano etal. 2016 for details). It has been applied to the ground − basedESO/WFI imaging (B [842] , V [843] , R [844] and Ic [879] , anddeep ESO VLT/HAWKI K s (obtained with 0 . (cid:48)(cid:48) seeing,Brammer et al. 2016) and space − based Sptizer/IRAC 3.6 µm and 4.5 µm bands (see Figure 4). While Ion3 is detected atS/N (cid:46) (cid:38)
10 in the Ks and IRAC bands.The most interesting features are the flat continuum at rest-frame wavelengths 4400˚A and 9000˚A, and the clear excess inthe 3 . µm band consistent with an H α emission with rest-frame EW of 1000˚A. The best-fit solution implies Ion3 isa relatively low stellar mass (1 . × M (cid:12) ) system under-going a starburst phase (SFR (cid:39) M (cid:12) yr − ) consistentlywith the presence of prominent Ly α , N v λ α and measured LyC. Ion3 appears as a stillrapidly growing system with a specific star formation rateof (cid:39) − (see Table 1 for a summary of the proper-ties of Ion3 ). The photometric estimate of the H α line lu-minosity ( (cid:39) × erg s − ) implies a high LyC photonproduction efficiency, ξ ion = 25 . ± . − ]. It resem-bles the values derived by Bouwens et al. (2016) for thebluest galaxies ( β < − .
3, see also Shivaei et al. 2017),and consistent with the values reported from local candi-date and confirmed LyC emitters, that also show large rest-frame EW(H α ) ∼ z > Ion3 belongs to the starburst region of the z ∼ . − M ∗ bimodal distribution recently identified by Caputiet al. (2017). Based on current data, there is no evidence of,nor any need for, any contribution to the UV emission byan AGN. The relatively large ratios of Ly α / N v ( (cid:39) ± α / C iv (cid:38)
20 tend to exclude the presence of an ob-scured AGN (e.g., Alexandroff et al. 2013). A relatively shal-low Chandra exposure of 130 ksec in the 0 . − yields a limit of 1 . × − erg s − cm − (3 σ ), corresponding to an upper limit to the X-rayluminosity of 3 × erg s − , which rules out a luminousAGN. Assuming a low-luminosity AGN is present, the veryyoung burst detected would imply the ionising photons area mixture of stellar and non-stellar radiation escaping alonga transparent medium ( N HI < . cm − ), that, however,should not be able to attenuate the expected high-ionisationemission lines, making Ion3 either a very special case amongthe AGN category or a pure star-forming dominated object.Another possibility is that
Ion3 has been captured just afterthe AGN has turned off, such that the optically thin channelproduced by the previous nuclear activity enables the Ly α and LyC stellar radiation to escape toward the observer. MNRAS000
Ion3 has been captured just afterthe AGN has turned off, such that the optically thin channelproduced by the previous nuclear activity enables the Ly α and LyC stellar radiation to escape toward the observer. MNRAS000 , 000–000 (0000) he highest redshift stellar ioniser Figure 4.
The cutouts of
Ion3 (top, 6 . (cid:48)(cid:48) across) and the bestSED fit with only stellar (red) and stellar and nebular (black)templates are shown. The photometric jump at 3.6 µm is evident. Ion3 is a bright star-forming galaxy showing copious LyCleakage identified just 1.5 Gyr after the Big-Bang ( z = 4).This makes Ion3 the highest redshift confirmed LyC emitterknown so far. In particular, • The spectral features and the SED-fitting suggest that
Ion3 is a young, low mass system undergoing a starburstphase containing hot and massive stars, with a specific starformation rate of 90 Gyr − . • The FORS and X-Shooter spectra reveal for the firsttime at such redshift a transparent ionised channel throughwhich the Ly α photons escape at the resonance frequency,plausibly along the same path of the LyC photons.The absence of HST imaging prevents us from deriv-ing any conclusion on the morphology of the galaxy. Theimage with the best seeing is the VLT/HAWKI Ks-band(0.39 (cid:48)(cid:48) ) from which Ion3 appears marginally resolved with aFWHM ∼ − α emis-sion. However, it is worth noting that Ion3 shows similarproperties of another LyC emitter we identified at z = 3 . Ion2 , Vanzella et al. 2016), e.g., a structured Ly α shape, the blue UV slope and weak low-ionisation interstel-lar absorption lines. While in the case of Ion2 we confirmedalso very strong [O iii ] λ > iii ] λ ii ] λ >
10) makingit among of the highest redshift Green Pea galaxy and sug-gesting a density-bounded condition,
Ion3 might also have aperforated medium and will need JWST to probe the rest-frame optical wavelengths (imaging and spectroscopy) andHST to image directly the LyC. Irrespective of the nature of the ionising radiation,
Ion3 represents a unique high-redshiftlaboratory where ionised channels carved in the interstellarmedium by one or more feedback sources can be studied.
Ion3 and
Ion2 represent ideal reference cases to guide thesearch for reionising sources at z > . ACKNOWLEDGMENTS
We thank the referee for constructive comments. EV grate-fully acknowledge the excellent support by ESO staff atParanal during the observations. We thank G. Zamorani,A. Jaskot, S. Oey, D. Schaerer and A. Grazian for usefuldiscussions. F.C., A.M acknowledge funding from the INAFPRIN-SKA 2017 program 1.05.01.88.04. MM acknowledgessupport from the Italian Ministry of Foreign Affairs andInternational Cooperation, Directorate General for Coun-try Promotion. Based on observations collected at the Eu-ropean Southern Observatory for Astronomical research inthe Southern Hemisphere under ESO programmes P098.A-0804(B), P098.A-0665(B).
REFERENCES
Alexandroff, R., et al. 2013, MNRAS, 435, 3306Behrens, C., Dijkstra, M., & Niemeyer, J. C. 2014, A&A, 563,77Bian, F., Fan, X., McGreer, I., Cai, Z., & Jiang, L. 2017, ApJ,837, L12Bouwens, R., et al. 2016, ApJ, 831, 176Brammer, G. B., et al. 2016, ApJS, 226, 6Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000Caminha, G. B., et al. 2016, A&A, 587, 80Caputi, K. I., et al. 2017, ApJ, 849, 45Castellano, M., et al. 2016b, A&A, 590, A31Chisholm, J., Orlitov´a, I., Schaerer, D., et al. 2017, A&A, 605,67de Barros, S., et al. 2016, A&A, 585, A51Grazian, A., et al. 2017, A&A, 602, A18.Heckman, T. M., et al. 2011, ApJ, 730, 5.Henry, A., Scarlata, C., Martin, C. L., & Erb,D. 2015, ApJ, 809,19Inoue, A. K., Shimizu, I., Iwata, I., & Tanaka, M. 2014, MNRAS,442, 1805Izotov, Y. I., Schaerer, D., Thuan, T. X., et al. 2016a, MNRAS,461, 3683Izotov, Y. I., Orlitov´a, I., Schaerer, D., et al. 2016b, Nature, 529,178Izotov, Y. I., et al. 2017, arXiv:1711.11449Jaskot, A. E., & Oey, M. S. 2013, ApJ, 766, 91Jaskot, A. E., & Oey, M. S. 2014, ApJ, 791, L19Jaskot, A. E., Oey, M. S., Scarlata, C., & Dowd, T. 2017,arXiv:1711.09516Jones, T. A., Ellis, R. S., Schenker, M. A., & Stark, D. P. 2013,ApJ, 779, 52Leitherer, C., Ekstr¨om, S., Meynet, G., et al. 2014, ApJS, 212, 14Lotz, J. M., et al. 2017, ApJ, 837, 97Nakajima, K., & Ouchi, M. 2014, MNRAS, 442, 900Rivera-Thorsen, T. E., et al. 2017, arXiv:1710.09482Salpeter, E. E. 1955, ApJ, 121, 161Schaerer, D., Izotov, Y. I., Verhamme, A., et al. 2016,A&A, 591,8Shapley, A., Steidel, C. C., Strom, A. L., et al. 2016, ApJ, 826,24Shivaei, I., et al. 2017, arXiv:1711.00013Siana, B., et al. 2015, ApJ, 804, 17.Stark, D. P., et al. 2017, MNRAS, 464, 469Vanzella, E., Siana, B., Cristiani, S., & Nonino, M. 2010, MNRAS,404, 1672Vanzella, E., et al. 2012, ApJ, 751, 70Vanzella, E., et al. 2014, ApJ, 783, L12MNRAS , 000–000 (0000)
E. Vanzella et al.
Vanzella, E., et al. 2015, A&A, 576, A116Vanzella, E., et al. 2016, ApJ, 825, 41Vanzella, E., et al. 2017, ApJ, 842, 47Verhamme, A., Orlitov´a, I., Schaerer, D., & Hayes, M. 2015, A&A,578, A7Verhamme, A., Orlitov´a, I., Schaerer, D., et al. 2017, A&A, 597,13Worseck, G., et al. 2014, MNRAS, 445, 1745.Yang, H., Malhotra, S., Gronke, M., et al. 2016, ApJ, 820, 130Zackrisson, E., Inoue, A. K., & Jensen, H. 2013, ApJ, 777, 39 MNRAS000