The evolution of Weak Mg II Absorption Lines

1
The evolution of Weak Mg II Absorption Lines
Toru Misawa1, Ryan Lynch1, Anand Narayanan1,
Nikola Milutinovic1, Tae-Sun Kim2, and Jane
C. Charlton1 (1 Penn State University, 2
University of Cambridge)
Summary We present two surveys of weak Mg II
(rest frame equivalent width Wr(2796)lt0.3Å)
absorbers at ltzgt0.15 and at ltzgt1.75. For the
low redshift survey, we use 25 HST/STIS echelle
quasar spectra (R45,000) which covered Si II
1260 and C II 1335 over this redshift range.
These transitions were used as a tracer of Mg II
2796, allowing a survey of a redshift pathlength
g(z)5.3 from 0 lt z lt 0.3. Our high redshift
survey covers a redshift pathlength of 6.7 in 15
high resolution QSO spectra obtained with UVES on
VLT. We find that dN/dz0.960.19 for 0 lt z lt
0.3 and 0.02 lt Wr(2796) lt 0.3Å, and that
dN/dz1.050.16 for 1.4 lt z lt 2.2 over the same
equivalent width range. These numbers are to be
compared to the results of a previous survey of
weak Mg II absorbers dN/dz1.740.10 for 0.4 lt z
lt 1.4 (Churchill et al. 1999). There appears to
be a peak epoch for weak Mg II absorbers at
z0.9. At higher redshifts, the observed value
is significantly less than expected for
cosmological evolution, and somewhat smaller than
the expectation considering the larger
extragalactic background radiation at ltzgt1.75.
We conclude from this that the processes that
produce weak Mg II absorbing structures are
equally, or somewhat less, active at ltzgt1.75
than at ltzgt0.9. At lower redshifts, ltzgt0.15,
we would expect a significantly larger number of
weak Mg II absorbers than observed. A
significant fraction of these expected objects
would result from the evolution of low density
structures that produced only observable C IV
absorption at higher redshifts. The fact that we
see fewer than expected suggests that the
mechanism for generation is less active at
present than at ltzgt0.9. The overall evolution of
weak Mg II absorbers is consistent with them
being related to some type of global star
formation activity, but detailed examination of
their properties will be needed to be more
specific.

• Expected Evolution of Weak Mg II Absorbers Due to
  Changing Extragalactic Background
• The extragalactic background radiation intensity
  decreases by about half a dex from ltzgt1.75 to
• ltzgt0.9, and almost another dex from ltzgt0.9
  to ltzgt0.15 (Haardt and Madau 2001).
• For the same physical entity we would expect
  weaker Mg II absorption at higher redshift and
  stronger
• Mg II absorption at lower redshift.
• To illustrate this point we simulate the expected
  evolution of three real ltzgt0.9 weak Mg II
  systems back
• to z1.75 and forward to z0.

PRESENT
z 0.9
z 1.75
The Nature of Weak Mg II Absorbers Weak Mg II
absorbers are metal-rich, with metallicities
approaching or exceeding the solar vale (Rigby et
al. 2002, Charlton et al. 2003). Despite these
high metallicities, weak Mg II absorbers are not
found to be directly associated with giant
luminous galaxies. They may be related to
supernova activity and/or star formation activity
in dwarf galaxies and/or the cosmic web. Weak Mg
II absorbers constitute a significant class of
objects, as their observed redshift path density
at redshift one is twice that of the strong Mg II
absorbers which are associated with luminous
galaxies. Weak Mg II absorption appears to be
produced by regions on the order of 1-100 pc
(Rigby et al. 2002, Charlton et al. 2003). The
same absorbers also show C IV absorption, which
is thought to be related to a larger, high
ionization region at the same velocity. Flattened
geometries are favored by the number statistics
and kinematic structures of Mg II and C IV
absorbers (Milutinovic et al. 2005). The exact
location and process that gives rise to these
absorbers remains a mystery.
low ionization
high ionization

• Goals of Present Study
• We know that dN/dz 1.740.10 for systems with
  0.02 lt Wr(2796) lt 0.3Å in range 0.4 lt z lt 1.4
  (Churchill et al. 1999).
• To determine dN/dz at lower redshift (0 lt z lt
  0.3)
• To determine dN/dz at higher redshift (1.4 lt z lt
  2.2)
• To compare the observed dN/dz values to
  expectation from evolution of extragalactic
  background and
• cosmological evolution.
• To consider possible implications about the
  nature of weak Mg II absorbers based on
  relationship between
• their evolution and the evolution of the
  global star formation rate in the universe.

Mg II from low density phase
Data and Survey Method 1.4 lt z lt 2.2 Survey
Spectra were obtained using UVES on VLT. The
resolution of these spectra is 45,000, and they
typically provide a search region for Mg II from
the Lyman-alpha emission line up to 10,000 Å.
The spectra had sufficient S/N that our survey is
100 complete for Wr(2796) gt 0.02 Å lines in
regions of the spectrum relatively free from
atmospheric absorption. Regions contaminated by
atmospheric absorption were eliminated from our
redshift path. 0 lt z lt 0.3 Survey Spectra were
obtained using the E140M grating of the STIS
spectrograph on the HST. The resolution of these
spectra is 45,000, and they cover the Si II 1260
and C II 1335 transitions in the redshift range
of interest. We calibrate the use of these
similar transitions as tracers of the Mg II
doublet using data for which coverage of all is
available. The completeness of the survey is 30
at the minimum equivalent width limit of Mg
II, corresponding to Wr(2796) gt 0.02A. The
results are not significantly sensitive to the
ratio of Si II and C II to Mg II that we adopt.
Figure 2
Figure 2 - Middle panels represent the observed
profiles of the three real weak Mg II absorbers
at z0.8, z0.9, and z0.6. We show both Mg II
2796 and CIV 1548. The red curve represents the
model contribution of a relatively high density
(0.01-0.1 g/cc) phase to the absorption, while
the blue curve represents the contribution of a
lower density (0.0001 -0.001 g/cc) phase which
gives rise only to observable C IV absorption in
the real system, and not to Mg II. The left set
of panels show the expected evolution of these
systems to the present, subject only to the
changing extragalactic background evolution
(allowing for no other evolution of the
structure). The right set of panels show the
expected evolution back to z1.75.
Estimate of Expected dN/dz at ltzgt 1.75 Based on
Evolving EBR We estimate the expected evolution
of the weak Mg II absorber population subject to
the evolving extragalactic background radiation
and cosmological evolution (in the favored Lambda
cosmology cosmological evolution is shown as the
solid curve in Figure 1). We begin by computing
the rest frame equivalent widths of Mg II 2796 at
ltzgt0.9 that correspond to 0.02 lt Wr(2796) lt 0.3Å
at ltzgt1.75 and at ltzgt0.15. We then integrate
the observed ltzgt0.9 equivalent width
distribution to predict the expected number of
weak Mg II absorbers absorbers expected at
ltzgt1.75 and at ltzgt0.15 (solid squares in Figure
1). The expected number at ltzgt0.15 also includes
a dominant contribution from low density,
kiloparsec-scale absorbers that produced only C
IV absorption at ltzgt0.9. These predictions
assume static populations of objects or ones that
regenerate at constant rates.
Predicted dN/dz3.4 at ltzgt0.15
Predicted dN/dz1.61 at ltzgt1.75

• Conclusions
• ltzgt0.15
• Processes that generate weak Mg II absorbers are
  less active at ltzgt0.15 than at ltzgt0.9.
• There are not as many low density, kiloparsec
  scale structures at low redshift as at ltzgt0.9.
  Those
• structures are within 100kpc of galaxies and
  have been hypothesized to relate to satellites or
  
• galactic fountains (Chen, Lanzetta, and Webb
  2001).
• The number of weak Mg II absorbers from higher
  density, parsec-scale structures is also likely
  to be
• evolving/decreasing from ltzgt0.9 to ltzgt0.15.
• The observed evolution is consistent with a
  relationship between the overall star formation
  rate in the
• universe and the generation of weak Mg II
  absorbers.
• ltzgt1.75
• Processes that generate weak Mg II absorbers are
  similar or more probably somewhat less active at
• ltzgt1.75 than at ltzgt0.9.
• The global star formation rate in the universe
  is likely to be a factor of 2 higher at ltzgt1.75
  than at
• ltzgt0.9.
• Weak Mg II absorber evolution roughly traces
  global star formation rate, however, small
  differences
• are not unexpected because they are not
  well-correlated all of the galaxies in which star
  formation is
• occurring.

Figure 1

• Results
• Figure 1 summarizes the redshift evolution of
  weak Mg II absorbers, combining the results of
  our two
• surveys with the earlier survey of Churchill
  et al. (1999).
• 1.4 lt z lt 2.2
• Detected 25 weak Mg II absorbers in 15 VLT QSO
  spectra.
• 7 weak Mg II absorbers with 1.4 lt z lt 2.2 in a
  redshift path of 6.7.
• dN/dz 1.05 ? 0.16 at ltzgt 1.75 (Solid blue
  points and errorbars in Fig. 1).
• 0 lt z lt 0.3
• Detected 5 weak Mg II absorbers in 20 HST/STIS
  spectra in a redshift path of 5.3 from 0 lt z lt
  0.3.
• dN/dz0.960.19 at ltzgt0.15 (Solid green point
  and errorbars in Fig. 1).
• As a check
• 18 weak Mg II absorbers with 0.4 lt z lt 1.4 in a
  redshift path of 11.5.
• dN/dz 1.570.14 (Our binned data over this
  range shown as dotted blue points in Fig. 1).
• Agrees with dN/dz 1.74 0.10 measured by
  Churchill et al. (1999) in a redshift path of
  17.0.

• REFERENCES
• Charlton et al. 2003, ApJ, 589, 111 Chen,
  Lanzetta, Webb, 2001, ApJ, 556, 158
• Churchill et al. 1999, ApJS, 120,
  51 Haardt Madau, 2001, astro-ph/0106018
• Milutinovic et al. 2005, in
  preparation Rigby, Charlton, Churchill, 2002,
  ApJ, 565,743