A Census of Intrinsic Narrow Absorption Lines at z 3

1
A Census of Intrinsic Narrow Absorption Lines at
z 3 T. Misawa, J. C. Charlton, M. Eracleous
(Penn State), R. Ganguly (Wyoming), D. Tytler,
D. Kirkman, N. Suzuki, and D. Lubin (UCSD)
QAL group
SUMMARY For the purpose of studying narrow
absorption lines that are intrinsic to the
quasars (intrinsic NALs), we have searched 40
Keck/HIRES spectra of quasars at z2-4. In these
spectra, we have identified 150 systems that
contain 124 C IV, 12 N V, and 50 Si IV doublets
as a complete sample, of which 18 are associated
systems (within 5,000 km/s of the quasar
redshift). Since the spectral resolution is very
high ( 8 km/s), we were able to use partial
coverage analysis to separate intrinsic NALs from
intervening NALs (produced in the quasar host
galaxies, intervening galaxies, or intergalactic
gas clouds). We found 39 intrinsic systems (28
reliable and 11 possible cases). From the
statistical properties of these systems we
conclude the following (a) at least half of the
quasars have one or more intrinsic NAL, (b) a
minimum of 17 of high-ejection velocity NALs
(normally thought to be intervening) are actually
intrinsic, (c) there are at least two distinct
families of NALs characterized by either strong N
V line or strong C IV lines. This work was
supported by NASA grant NAG5-10817.
2. Fraction of quasars that have one or more
intrinsic NALs One effective method of
investigating the geometry of absorbing gas
around quasars is counting the number of
intrinsic NALs per quasar, and to evaluate the
fraction of the quasars that have intrinsic NALs.
These values probably depend on the global
covering factor around central continuum source
and broad emission line regions (BLRs) and the
distribution of absorbers around the emitting
regions.
DATA ANALYSIS 1. Covering Factor Analysis
The optical depth ratio of C IV or other doublet
lines sometimes deviates from the values expected
from atomic physics, 21. This discrepancy can be
explained if the absorber covers the continuum
source only partially, and an unabsorbed
continuum changes the relative depths of the
lines (e.g., Wampler et al. 1995). Based on this
idea, we can evaluate the covering factor by,
where R1 and R2 are the continuum normalized
intensities of the weaker and stronger components
of the doublet. We adopt two methods, (i)
evaluating Cf for each pixel, and (ii) fitting
absorption profiles with Cf , the column density,
and Doppler parameter as free parameters. 2. NAL
Classification Method Based on the results of
the covering factor analysis, we attempted to
separate all NALs into three categories class-A
and B as reliable and possible partial
coverage NALs (i.e., intrinsic NAL candidates)
and class-C as full coverage NALs (i.e.,
intervening or unclassified NALs), respectively.
We classified 4 mini-BALs (sub-class between BAL
and NAL their line profiles are smooth, and
blue/red members of doublets are self-blending)
into class-A. We also classified three NALs into
class-B regardless of the covering factors,
because these are line-locked, which is often
seen in intrinsic NALs at zabs zem. See
examples below
At least half of quasars have one or more
intrinsic NALs, which implies that (i) either
at least 50 of the solid angle around a
typical quasar is covered by intrinsic
absorber (ii) or at least 50 of all quasars
are fully covered. This is higher than
in zlt2 quasars (25), suggesting that intrinsic
NALs evolve away at low redshift.
Number of quasars that have 0-4 intrinsic NAL
systems.
3. Covering Factor Distributions In order to
explain significant differences between the
ejection velocity distributions of C IV NALs in
flat-spectrum quasars and steep-spectrum quasars,
Richards (2001) posited that 36 of
non-associated C IV systems, normally thought to
be intervening, are actually intrinsic. We find a
smaller fraction of intrinsic, non-associated
NALs in our sample, 10-17. This fraction is a
lower limit, since not all intrinsic NALs need to
exhibit partial coverage, thus our findings are
consistent with the conclusions of Richards.
We also found that N V NALs have smaller
covering factors (Cf 0.5) than C IV and Si IV
NALs. All the N V NALs in our sample are within
5,000 km/s of the quasars, and are detected at
(or near) the top of Lya emission lines of the
quasars (see figure on the right) If the N V
NALs cover the continuum source but not the BELR
(cf, Arav et al. 1999 Misawa et al. 2005),
dilution by the Lya emission line flux would lead
to smaller apparent covering factors.
RESULTS 1. Fraction of Intrinsic NALs The
numbers of C IV, N V, and Si IV NALs, after this
classification, are summarized in the pie charts
below. There are no major differences between
radio-loud and quiet quasars, although our sample
includes only very few radio-loud quasars.
Numbers in the graphs denote the number of NALs.
4. Ionization conditions in the intrinsic
absorbers We consider the ionization structure
of the 39 class-A and class-B intrinsic NAL
systems in our sample, and find two major
categories (see examples below) Strong C IV
systems (27 of 39 intrinsic systems)
characterized by strong C IV, weaker or absent N
V, and strong Lya absorption. Sometimes
low-ionization lines are detected, perfectly
aligned in velocity with C IV, which implies the
low ionization lines arise not from quasar host
galaxies, but from the intrinsic
absorbers. Strong N V systems (12 of 39 intrinsic
systems) characterized by strong N V
absorption, and relatively weak, non-black Lya
absorption, less than twice the equivalent width
of the N V. Low-ionization lines are rarely found
in these systems. Among 19 quasars with
intrinsic NALs, 10 have at least one
low-ionization system (i.e., a system with
low-ionization lines). These quasars are
analogous to LoBAL quasars, but the fraction of
low-ionization NAL quasars (53) is much higher
than that of LoBAL quasars (13 17) among BAL
quasars.
19 of all C IV NALs are intrinsic. Among
associated NALs (AALs within 5,000 km/s of the
quasar emission redshift), 33 are intrinsic.
18 of all Si IV NALs are intrinsic. None of them
are associated i.e., all have vej gt 5,000
km/s. Because of the Lya forest, we were
unable to study N V doublets with vej gt 5,000
km/s. Of the associated N V NALs, 75 are
intrinsic.