Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.eso.org/~bleibund/papers/Paris.ps Äàòà èçìåíåíèÿ: Sun Jan 7 21:59:00 2001 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 01:55:42 2012 Êîäèðîâêà:

Emission in the Damped Lyman ff Trough of Q2059­360
Gordon Robertson 1;2 , Bruno Leibundgut 2
1 School of Physics, University of Sydney, NSW 2006, Australia
2 European Southern Observatory, Karl­Schwarzschild­Strasse 2, D­85478 Garching,
Germany
Abstract. We confirm the detection of emission within the Damped Lyman
Alpha (DLA) absorption line of Q2059­360. The emission line has a velocity offset
of 490 km s \Gamma1 with respect to the DLA absorption system, making it unlikely that
both absorption and emission arise in a single galaxy. The emission feature itself
is resolved in velocity, probably consisting of two components separated by ¸ 300
km s \Gamma1 . It is also spatially extended and we trace it to at least 2 00 (corresponding
to 15 h \Gamma1
50
kpc) from the QSO line of sight. It may arise from a galaxy grouped with
the DLA absorber, or could be due to Narrow Line Region emission filaments of the
QSO itself.
1 Introduction
The systems causing Damped Lyman Alpha (DLA) absorption troughs in
quasar spectra are of interest because they may represent the precursors of
today's disc galaxies [8], and they have been selected by passive absorption
rather than by the presence of an active nucleus.
At low to medium redshifts, where Lyman ff is in the UV, a variety of galaxy
types have been shown to be responsible for the observed DLAs [5]. For higher
redshifts, with Lyman ff visible from the ground, many attempts have been
made to detect Lyman ff emission from DLA systems. Such observations are
aided by the broad damped absorption which blanks out the (brighter) quasar
over the relevant wavelength range. But only a small number of detections have
been reported: Q0836+113 [4]; PKS 0528­250 [11]; Q2233+131 [2]; Q2059­
360 [7] and Q0151+048A [3]. The low Lyman ff fluxes may be the result of
absorption by dust, in combination with resonant scattering of Lyman ff line
photons (eg [1]).
2 The emission feature in the DLA of Q2059­360
We have observed the quasar Q2059­360, using the 3.5 m ESO New Technology
Telescope. Figure 1 shows the result of 5 hours exposure with the slit on or
close to the quasar. The DLA trough redshift (z abs = 3:0825) is very close to
that of the QSO's broad Lyman ff emission line (z em = 3:0954). Hence the
DLA removes much of the emission line flux, and the QSO line+continuum

4700 4800 4900 5000 5100
0
5â10 ­17
10 ­16
1.5â10 ­16
2â10 ­16
Si III
(1206)
Si II
(1193)
Fe II
(1144) Si II
(1190)
Fe II
(2586) Fe II
(2600)
l hel (Angstrom)
F
l
(erg
s
­1
cm
­2
A
­1
)
Figure 1: Spectrum of Q2059­360 recorded with the EMMI spectrograph on 29­31
August 1995. The spectral resolution is 1.5 š A, and the slit was oriented East­West for
all exposures. The metal lines indicated are in the same redshift system as the DLA,
except for the probable Fe ii 2586,2600 lines in a low redshift intervening system.
spectrum that we show in Figure 1 (as the dashed line) is a reconstruction
from a simultaneous fit to the DLA profile and a smooth incident spectrum
[6]. Figure 1 also shows (as the heavy line) our fit of a Voigt profile to the
DLA trough, with log N(H I) = 20:85 \Sigma 0:03. The observed spectrum shows
clear excess emission at the red edge of the DLA trough, confirming the earlier
detection [7].
Reasoning that the object causing the DLA system is likely to have an an­
gular size of one to a few arcseconds if it is indeed of galactic dimensions, and
that any putative emission need not peak at the exact position of the back­
ground quasar, we also observed with slit positions displaced from the quasar
(a further 7 \Theta 1 hr exposures). With the long­slit spectrograph, this provided
a rudimentary two­dimensional spatial coverage with spectral resolution ade­
quate to clearly resolve any faint emission feature from the quasar's spectrum.
The results showed that the emission feature near 4970 š A is spatially extended,
with detection continuing to our greatest slit displacement of 2:3 00 south (where
none of the quasar's seeing disc remains on the slit). The observed extension
establishes that the feature seen at the red edge of the DLA trough is indeed
due to emission independent of the quasar nucleus, and cannot be caused by
an unusual distribution of absorption. The line flux values that we find are
0:6 \Gamma 0:9 \Theta 10 \Gamma16 erg s \Gamma1 cm \Gamma2 , and the angular extent of at least 2 00 indicates
a physical size of ? 15 kpc (H 0 = 50 km s \Gamma1 Mpc \Gamma1
,\Omega = 1).
With the good fit of a Voigt profile model DLA to the data as shown in

Figure 1, we have been able to subtract the quasar's light (suitably scaled)
from all runs, including those on or near the quasar itself, to leave only the
faint DLA emission feature. We find it to be resolved spectrally as well, with
a suggestion of two components separated by ¸ 300 km s \Gamma1 .
3 Relationships between the DLA absorber, emitter, and
the QSO
Owing to the similarity of the redshifts of the DLA and the quasar itself, the
interpretation of the physical system is more complex than for intervening
absorbers well away from their background quasar. In Figure 2(a) we indicate
schematically where the various emissions and absorptions occur on a radial
velocity scale. The velocity origin has been taken as the QSO's Lyman ff
emission line (reconstructed as shown in Figure 1), with – rest
eff = 1214:97 š A [9].
There are several points to note:
(i) A probable velocity difference of 40 km s \Gamma1 between the DLA hydrogen
absorption and the metal lines in the same system;
(ii) The emission features cannot be located behind the DLA absorber,
so peculiar velocities may be invoked to account for the relative velocities as
shown (another alternative is discussed below);
(iii) The broad Lyman ff and C iv emission lines of quasars are inaccurate
indicators of the true systemic redshift of the quasar host galaxy, being subject
to both a mean bias and large scatter [9]. If the QSO redshift is instead
estimated from the broad C iv line (in the spectrum from [10], and using
– rest
eff = 1547:46 š A [9]) the QSO's `zero' velocity actually falls within emission
component 2. Thus the observed redshifts allow interpretation of the emission
feature(s) as emission line filaments of the quasar itself.
(iv) The mean velocity offset of the emission features with respect to the
DLA absorption is +490 km s \Gamma1 .
In Figure 2(b) we show two possible spatial configurations of the absorber,
emitter, and QSO. In both cases the minimum distance between the QSO and
the DLA absorber is substantial. This limit arises from the requirement that
the radiation flux from the QSO incident on the DLA cloud be low enough to
avoid complete ionisation of the cloud in a time short compared with the age
of the universe at z = 3 [6].
In the upper model of Figure 2(b), the emission features are shown as sep­
arate objects from the DLA absorber, owing to the large velocity offset (as
suggested in [7]). The absorber and emitter may be members of the same
group or cluster. The lower model accounts for the near equality of redshifts
of the quasar and the emission feature. In this case, the emission is interpreted
as coming from Narrow Line Region emission filaments surrounding the QSO
itself or close companions to the QSO. This implies that the DLA cloud ab­
sorbs only over a small angular size, allowing part of the spatially extended or
displaced emission region to be seen within the seeing disc of the quasar. The

0
QSO AGN
Emission
Component 1
feature
Emission
feature 2
DLA
metal abs.
Ly alpha
DLA
­1200 ­1000 ­500
Telescope
absorption
Ly alpha
emission
(a)
QSO AGN
4 Mpc
>
DLA
Emitters
Emitters
DLA
QSO AGN
(b)
v km/s
Figure 2: (a) Placement of the emission and absorption features on a radial velocity
scale. (b) Two possible spatial configurations for the Q2059­360 system (not to
scale).
DLA is thus acting as an occulting disc, allowing us to see any surrounding
faint objects which emit in the appropriate band of wavelengths.
Further observations will include: (i) Low­dispersion spectra of the emis­
sion features, to search for other lines (especially C iv) and the continuum; (ii)
an accurate systemic redshift of the quasar, eg from Hff at 2:7¯; (iii) a more
complete study of the metal lines in the DLA absorption system.
Acknowledgements. We thank the ESO Observing Programs Committee for al­
location of telescope time. The Science Foundation for Physics within the University
of Sydney supported travel for JGR.
References
[1] Charlot, S., & Fall, S.M., 1991, ApJ 378, 471
[2] Djorgovski, S.G., Pahre, M.A., Bechtold, J. & Elston, R., 1996, Nature 382, 234
[3] Fynbo, J., MÜller, P. & Warren, S.J., 1998, these Proceedings
[4] Hunstead, R.W., Pettini, M. & Fletcher, A.B., 1990., ApJ 356, 23
[5] Le Brun,V., Bergeron, J., Boiss'e, P. & Deharveng, J.M., 1997, A&A, 321, 733
[6] Leibundgut B. & Robertson G., 1998, MNRAS (submitted)
[7] Pettini, M., Hunstead, R.W., King, D.L. & Smith, L.J., 1995, QSO Absorption
Lines, ed. G. Meylan, (Berlin: Springer), 55
[8] Prochaska, J.X. & Wolfe, A.M., 1996, ApJ 470, 403
[9] Tytler, D. & Fan, X­M., 1992, ApJS 79, 1
[10] Warren, S.J., Hewett, P.C. & Osmer, P.S., 1991, ApJS 76, 23
[11] Warren, S.J., & MÜller, P., 1996, A&A, 311, 25