Spitzer Observations of Star Formation in Several

Download Report

Transcript Spitzer Observations of Star Formation in Several

Mid-Infrared Observations of
Nearby Interacting/Starburst Galaxies
with the Spitzer Space Telescope
8/15/05
Zhong Wang
For Giovanni Fazio and the IRAC Science Team
1
Science Goals:
•
Trace locations of active star formation, often with heavy extinction
from dust. Explore possible connections to more distant ULIRGs.
•
Measure spectral energy distribution (SED) of the star forming regions,
identify energy source.
•
Compare with kinematics, spectroscopic probes, and other types of star
formation tracers to determine the role of interaction in starburst and
evolutionary history of these systems
8/15/05
2
Why interacting
systems?
• Most of the infrared luminous galaxies are
found to be interacting systems
• This is even more true for the ultraluminous infrared galaxies (ULIRGs)
Characteristics of Our Sample:
• Using IRAC/MIPS to observe ~30
infrared luminous, interacting systems
• All galaxies have significant amount of
gas/dust, and active star formation
• Systems are mostly recent interactions,
at various stages of final merging
• Many have high-quality imaging data in
UV/optical, NIR, radio and X-ray
8/15/05
Ishida et al (2005)
3
GTO Program on
Interacting/Starburst Galaxies
Observations of ULIRGs in the Sample: (see presentation by Jason Surace)
8/15/05
4
Nearby interacting/starburst systems:
NGC 4038/9 (the Antennae)
NGC 6240
NGC 6090
NGC 1068
NGC 3690
NGC 520
NGC 7252
NGC 2623
Arp 220
8/15/05
5
A Primary Example:
the “Antennae”
A Proto-typical interacting
pair/merger
8/15/05
6
The ISM component
Optical and Radio 21-cm
8/15/05
Hibbard et al (2001) have mapped HI gas in
the two extended tidal tails as well as in
the central region.
CO and other molecular line observations with
IR, mm/submm telescopes detect denser
gas in the center e.g., Standford et al
(1990), Wilson et al (2000), Gao et al
(2001), Zhu et al (2003), Hass etal
(2005), Iono etal (2005).
CO contours
7
The dust lanes
Overlap region
8/15/05
B, V
Dust emission is also measured
with SCUBA (Hass et al 2000)
H
8
Hubble and ISO findings
Whitmore et al (1995, 1999) and Zhang et al (2001) found based on HST B, V and H images that
the Antennae contains numerous (1,000+) young stellar clusters or super-clusters, perhaps as a
result of the interaction.
The ages of these clusters range from one to a few hundred million years.
Mid- and far-infrared images from KAO and ISO (7 to 160m) show that the center of NGC 4038
and the overlap region are the major contributors of the infrared flux (Vigroux et al 1996;
Kunze et al 1996; Bushouse et al 1998; Mirabel et al 1998; Xu et al 2000).
ISO
WFPCII data
Of individual
Young stellar
Clusters
(Whitmore et al
1999)
15 m peak
8/15/05
9
Galaxy SEDs in the mid-IR
Potential Contributors:
•Stellar continuum (B-B)
•Small dust grains
•Large dust grains
•PAHs
•Possible AGN component
(power-law)
8/15/05
10
Galaxy SEDs in
the mid-IR
Empirical modeling:
Dale et al (2001, 2002)
have shown that the
warm dust emission
from galaxies can be
parameterized as a
function of heating
intensity of UV
photons.
The spectral region that is
most sensitive to the
level of star forming
activities is between 20
and 42 m
8/15/05
11
Dale et al. (2001)
Antennae in IRAC bands
At longer IRAC wavelengths, warm dust
emission becomes increasingly more
important
3.6 µm
R
8.0 µm
8/15/05
IRAC
12
The stellar mass component
8/15/05
Ks (Martini 2005)
13
IRAC 3.6m
The warm dust emission
8/15/05
IRAC 4.5m
IRAC 8.0m
Wang et al. (2004)
14
The photon-heated dust grains --recent star formation sites
8/15/05
MIPS 24m
MIPS 70m
15
Star formation indicator
The MIPS [24-70] map of the Antennae
provides a quantitative indicator of
star formation sites as well as star
formation intensities in various
parts of the system, especially in
the overlap region, where the
starburst activities approach the
level of those in the ULIRGs.
MIPS [24 -70] map
Wang et al (2005)
8/15/05
Dale et al (2000)
16
How to measure the mid-IR fluxes
This would be a typical approach
(e.g., SINGS).
But it would not be appropriate
for interacting/irregulars such as
the Antennae
8/15/05
17
How to measure the mid-IR fluxes
Using analysis tool HIIPhot to measure
fluxes, based on high-resolution H image
(with B. Penprase)
In effect, we divide emission areas into
large “HII” regions, based on their
morphology
2
3
4
8/15/05
11
18
IRAC colors of different regions
in the Antennae
8/15/05
19
Unusually strong 4.5 m flux
Why is the [4.5-3.6]
color very
different in active
star forming
regions?
Maybe we’re seeing a
large amount of
shocked molecular
gas as those found
in certain outflow
regions in the
Milky Way.
?
3.6 4.5 8.0
Reflection nebular: NGC
7129 (Megeath et al 2004)
R 4.5 H
8/15/05
20
Nine-panel comparison
V
3.6
70
8/15/05
H
8.0
450
Ks
24
850
21
Details of nine-panel image
H
3.6
70
8/15/05
8.0
450
Ks
24
850
22
SEDs in the mid-IR
Instead of measuring
fluxes over the
entire galaxies, we
can now trace the
SEDs of individual
regions In the
interacting pair,
tracing the
progression of the
merging effects.
From SINGS
8/15/05
23
WF2
Hα
Lípari et al ‘04
“Off-nuclear”
Star formation sites
NGC 3256
B, I
“Secondary
Nucleus”
WFPC2
IRAC
3.6,4.5,8.0 µm
Zepf et al ‘99
8/15/05
24
NGC 3690
8/15/05
1.6
3.6
8.0
24
70
25
Summary
With the higher angular resolution and sensitivity of spitzer imaging observations, individual
regions of nearby interacting/starburst galaxies can be studied in greater detail in terms of
gas/dust contents, SEDs, star forming rate/efficiency, and the effect of shocks.
IRAC/MIPS data can be better analyzed with the help of short wavelength data (e.g. HST).
In the Antennae and NGC 3690, the most intensive starburst activities, appear to have been
triggered well before the final merging of the two main gas disks and well outside the
gravitational center of the system, but with efficiencies comparable to those of ULIRGs.
The spatial distributions of atomic gas, molecular gas, warm dust and ionizing photons are
indicative of a progressive change from denser gas concentrations to forming stars in
clusters. Thre observed active supernova explosions are apparently related to this.
In the most active regions we also find particularly strong emission at 4.5m, suggesting an
possible origin from shocked molecular gas. This, along with the variation of MIPS [24-70]
ratio suggest that the most active star formation can be a very localized phenomena in
early-stage mergers.
Combined with observations from other wavelengths, a more complete picture of galaxy-galaxy
merger and associated star formation process is emerging and ready to be compared with a
new generation of theoretical models.
8/15/05
26