Large-scale environmental dependence
of gas-phase metallicity in low-luminosity galaxies
Kelly Douglass, Michael Vogeley
Department of Physics, Drexel University
We study how the cosmic environment affects galaxy evolution in the Universe by comparing the metallicities of galaxies in voids with similar-sized galaxies in more dense regions. Ratios of the fluxes of emission
lines, particularly those of the [OIII] and [SII] transitions, provide estimates of a region’s electron temperature and number density. From these two quantities and the emission line intensities, we estimate the
abundance of oxygen with the Direct Te method. We estimate the metallicity of 2865 void low-luminosity galaxies and 6448 wall low-luminosity galaxies using data from SDSS via the MPA-JHU Garching catalogue.
We find very little difference between the two sets of galaxies, indicating little influence from the large-scale environment on the history of star formation. Of particular interest are a number of extremely metalpoor low-luminosity galaxies.
Metallicity
Metallicity dependence on large-scale environment
Metallicity is often described in terms of the ratio of oxygen to hydrogen (O/H) present in the
interstellar gas of a galaxy. The region of a galaxy which recently experienced star formation is known
as an HII region; here, UV photons from the stars keeps the interstellar gas partially ionized. Optical
photons are either absorbed and re-emitted throughout the region at resonant frequencies (resulting
in electron dipole transitions), or the electrons are collisionally excited (resulting in nominally forbidden
electron transitions).
It is suspected that the metallicity of low-luminosity void galaxies is less than that of low-luminosity
galaxies in higher density regions, for a variety of reasons:
•
More pristine gas surrounding void galaxies (reduces metallicity)
•
Fewer galactic interactions (induces star formation, which eventually increases metallicity)
Since the metallicity is an indicator of star formation history, this would indicate a very different star
formation rate between the two galactic populations. Indeed, previous studies of select void dwarf
galaxies have found them to have low metallicities (Pustilnik et al. (2006, 2011) for example). However,
as the histograms below indicate, there is no significant difference between the metallicity of galaxies in
the different regions. There is a collection of extreme low metallicity galaxies, but these exist in both
the voids and walls. These results are a strong test for galaxy formation models of the ΛCDM theory,
as void galaxies are currently found to have lower mass and be retarded in their star formation when
compared to wall galaxies.
SDSS.org
0.5
0.45
0.4
From the flux of these forbidden emission lines, the electron temperature and number density of the HII
region can be found; it is then possible to calculate the ratios for all heavy elements with respect to H.
The total abundance of an element is equal to the sum of the abundances of each of its ionized states:
A metallicity Z = 12+log(O/H] < 7.6 is considered metal-poor; High metallicities are Z > 8.2. The solar
metallicity Z! = 8.66.
•
•
Mr > -20 – low-luminosity galaxies
0.02 < z < 0.10 – The apparent magnitudes of low-luminosity galaxies at z > 0.10
are outside the limits of the survey, as mr < 17.77 for SDSS, and [OII] λ3727 is
outside the SDSS spectrum for z < 0.02.
0.25
0.2
0.15
[OII] λ3727 line is used to calculate the abundance of singly-ionized oxygen
•
Method published by Izotov et al. (2006), which is based on the astrophysics published by
Osterbrock (1989)
0.3
0.25
0.2
0.15
0.1
0.05
0.05
7
7.5
8
12+log(O/H)
8.5
9
0
6.5
9.5
7
0 > Mr > -17 (dwarf galaxies)
660,000 low-luminosity galaxies are available to analyze for z < 0.10. Of those,
22,104 galaxies have valid emission lines (particularly [OIII] λ4363 > 2σ); 9319
have a detected [OII] λ3727 line.
0.45
Large-scale environment – Void or Wall?
All galaxies are classified as either void or wall based on the Void catalog compiled by
Pan et al. (2012). This catalog is based on the VoidFinder algorithm developed by
Hoyle & Vogeley (2002), which locates statistically significant voids in the SDSS.
Since [OIII] λ4363 is inversely proportional to metallicity and is much weaker than the strong
emission lines, the method is best for low-redshift, low-metallicity galaxies.
8
12+log(O/H)
8.5
9
9.5
0.5
Void
Wall
nvoid = 839
nwall = 1630
0.45
0.3
0.25
0.2
0.15
0.35
0.3
0.25
0.2
0.15
0.1
0.1
0.05
0.05
7
7.5
8
12+log(O/H)
8.5
9
Void
Wall
nvoid = 1671
nwall = 4055
0.4
0.35
0
6.5
7.5
-17 > Mr > -18
0.5
Fraction of galaxies
[OIII] λ4363, λλ4959, 5007 lines are used to calculate the electron temperature and the
abundance of doubly-ionized oxygen
0.35
0.1
0.4
•
•
0.3
0
6.5
Star-forming – ultraviolet photons are needed to excite the interstellar gas, so all
required lines must be emission lines, with a minimum 3σ detection of Hβ and a
minimum 2σ detection of [OIII] λ4363.
Direct Te method – used for metallicity calculations
•
0.35
Void
Wall
nvoid = 319
nwall = 695
0.4
Fraction of galaxies
[OIII] emission lines are sensitive to electron temperature
[SII] emission lines are sensitive to electron number density
•
nvoid = 36
nwall = 68
0.45
Fraction of galaxies
De Robertis et al. (1987)
Fraction of galaxies
Galaxy selection criteria
0.5
Void
Wall
0
6.5
9.5
-18 > Mr > -19
7
7.5
8
12+log(O/H)
8.5
9
9.5
-19 > Mr > -20
Histograms comparing the gas-phase metallicity of void low-luminosity (red dashed line) and wall low-luminosity
(black solid line) galaxies. Our survey covers a much broader range of metallicity values than others, including
Tremonti et al. (2004), which only contains galaxies with values between 8 and 10. As can be seen, there is
not a special population of low-metallicity galaxies in voids.
How does the metallicity of a galaxy change?
H
H ! He
Future work
He ! C
C ! Ne
•
Investigate the N/O ratio in these galaxies – this will help us discern the extreme low-metallicity
values
•
Use the metallicity and other physical characteristics to trace the evolution of dark matter
Ne ! O
Sample void dwarf
galaxy
and spectrum.
satrec.kaist.ac.kr/fims/fims.htm
The “galactic fountain” model. Here, the elements are made within the stars via nucleosynthesis.
When the stars die, their supernovae expel the gas out of the local interstellar medium, where
some of the gas eventually falls back onto the galaxy, increasing the galaxy’s metallicity. In
addition, clouds of neutral hydrogen can also fall onto the galaxy, decreasing the metallicity.
12+log(O/H) =
8.60 ± 0.197
SDSS.org
References
•
•
•
•
De Robertis, M.M., Dupour, R.J., and R.W. Hunt. 1987,
JRASC, 81, 195
Hoyle, F. and Vogeley, M.S. 2002, ApJ, 566, 641
Izotov, Y. et al. 2006, A&A, 448, 955
Osterbrock, D.E.. Astrophysics of Gaseous Nebulae and
Active Galactic Nuclei. University Science Books, Mill Valley,
•
•
•
•
CA (1989)
Pan, D.C. et al. 2012, MNRAS, 421, 926
Pustilnik et al. 2006, AstL, 32, 228
Pustilnik et al. 2011, MNRAS, 417, 1335
Tremonti, C.A. et al. 2004, ApJ, 613, 898