2. Target: SNR W28 (G6.4-0.1)
The Northeastern Shell (XMM-Newton data)
The Central Region (Chandra data)
Aims:
Detect powerlaw component (non-thermal X-
ray radiation) from W28’s NE shell for
broadband model
Study plasma properties of SNR W28
3. Ejected stellar material (ejecta) from a supernova, mixed with interstellar
medium (ISM) and compressed by forward shock, forms a SNR.
What is supernova remnant (SNR)?
4. SNR Evolution
Free Expansion
Adiabatic or
Sedov phase
Radiative phase
Mejecta >> Mswept-up
Ejecta expands without deceleration
Outer ISM shell by forward shock
Mejecta ~ Mswept-up
Significant deceleration
Ejecta merges into forward shock
Energy loss by adiabatic expansion
Efficient particle acceleration by
strong forward shock
Tejecta < 106K
Efficient cooling by radiation
Well-mixing between ejecta and ISM
No efficient particle acceleration due to slow forward shock
Disappearance when velocity drops to the typical value of ISM
ejecta
outer shell of
ISM by forward
shock
5. Types of SNR
Shell-type (e.g. Cassiopeia A)
• non-thermal X-ray emitting
shells
Crab-like (e.g. Crab Nebula)
• non-thermal X-ray or radio
radiation from a pulsar at
the centre
• faint shells or no shell at all
Composite-type
• if appear as both shell-like
and Crab-like
Mixed-morphology (e.g. W44)
• radio shell emission
• thermal X-ray center-filled
emission
• uniform temperature profile
not common in shell-type
SNRs
Ejecta Cassiopeia A
Crab Nebula
Radio
shell
Radio image
Thermal X-ray
center-filled
emission
X-ray image
W44
Neutron star
(a little spot)Non-thermal X-
ray emitting shell
by forward shock
Pulsar
6. SNR W28
• Mixed-morphology SNR
• 35000-15000 year old
• 50’x45’ angular size
• ~ 2kpc distant away
• Possibly evolved into its
radiative phase
Previous X-ray observation
with ROSAT and ASCA (2002):
Northeastern :
single thermal component
(0.6 keV ),
ISM abundance,
Center:
two thermal components
(1.8 keV and 0.6 keV)
Radio Image
of W28
ROSAT X-ray
Image of W28
7. Why Study SNR
W28?
• γ-ray emission near the
northeastern shell of W28
detected by H.E.S.S in 2008
and Fermi-LAT in 2010.
• γ-ray emission:
Inverse Compton
scattering (electrons)
Non-thermal
bremsstrahlung (electrons)
π0 decay (protons)
Detection of non-thermal X-ray
• helps determine if γ-ray emission
by electron or proton acceleration.
Radio Image
of W28
ROSAT X-ray
Image of W28
8. Download raw data from data
archive HEASARC and check
them, e.g. version, mode, etc.
Create a new event file
Check light curve
Filter the event file
Select region and extract the region’s
spectrum from the event file
Background subtraction
Spectral analysis and
fitting: XSPEC ver.12.5.1
Need reprocessing
for event file?
Flaring exist?
YES
YES
NO
NO
Basic Procedure of
Preparing Spectra
9. The CCD image of W28’s
NE shell from XMM (10
regions)
The CCD image of W28
from Chandra (8 regions)
Parameters concerned in
spectral fitting:
1. Temperature
2. Abundances of elementsROSAT X-ray
Image of W28
11. O
Mg
Si
S
The CCD image of W28’s NE
shell from XMM
pn-CCD
component
MOS-CCD
components
Northeastern Shell
12. Results for Northeastern Shell
All regions’ spectra are best
fitted by two thermal
components, different to single
thermal component observed by
ROSAT and ASCA.
Low temperature components
correspond to interstellar
medium (ISM) and high
temperature components
correspond to SNR ejecta.
Power-law model was also used
for fitting spectra, but photon
indices Γ > 4.0 for all the
regions, except REG G that Γ =
3.8.
13. Enhanced abundances of O
were detected in all regions,
and most of them show
enhanced abundances of Mg
as well, confirming the
presence of ejecta in W28’s NE
shell.
Emission from ejecta is still
strong enough to be resolved by
XMM, not expected from old
SNRs like W28.
Results for Northeastern Shell
14. Results from Chandra Data
Image of the Chandra
central chip with point
sources moved
15. Results from Chandra Data
Three-thermal model best fits the central
spectra. The component of ~1.8keV was also
observed by ROSAT and ASCA (2002), but two
thermal only.
The third thermal component may correspond
to reflected shock by (molecular) cloud.
Further analysis is needed to determine
whether the reflected shock model fitted the
data of the central region.
Image of the Chandra
central chip with point
sources moved
Three
thermal
components
Z
also observed by
ROSAT & ASCA
16. Results from Chandra DataChandra image with
point sources moved
Power-law component was
detected in R12.
Γ
17. Constraint on Non-thermal X-ray Radiation in
Northeastern Shell
The multi-band spectra of the Fermi-LAT source at the
NE boundary of SNR W28 (taken from Abdo et al. 2010).
Fermi
H.E.S.S
EGRET
18. We also attempt to estimate the mass of the
progenitor star associated with W28 by:
1. calculating the ejecta masses of
overabundant elements
2. comparing our values to the well-known
results (Thieleman et al. 1996).
The progenitor star’s mass:
8 M⊙ < M < 13 M⊙
There should be a neutron star
near the remnant’s center.
Mass of Progenitor Star
associated with W28
Our values are 10
times smaller than
that for 13M⊙
19. Identifying Background Point Sources and
Potential candidates for Neutron star
The CCD image of W28
from Chandra
• Five possible candidates
were identified (blue
circles).
• But determination of
their flux is needed to
find out the strong
candidates.
20. Summary
1. New thermal components detected in both the NE shell
and the central region of W28.
2. Confirmation of high temperature plasma of ~1.8 keV at
W28’s center.
3. First detection of ejecta with enhanced abundances of O
and Mg in W28’s NE shell .
4. First detection of powerlaw components in both W28’s
NE and S regions.
5. Determination of lower limit of flux of non-thermal X-ray
for the NE shell’s broadband modeling.
6. First identification of possible candidates of neutron star
associated with W28.
23. Results from Chandra Data
Oxygen line, not detected in other
Chandra’s regions we investigated, was
detected in both R1 and R2.
However I had trouble to determine the
best fits for R1 and R2 due to the
background spectrum dominant at
energies E > 3keV.
Considering consistency, the fits with solar
abundances are the preferred ones.
Chandra image with
point sources moved
Chandra image with
point sources moved
24. Ejecta
Injection of
ions to shock
Efficiency of
particle
acceleration
Amplification
of magnetic
field in shock
by CRs
Particle
acceleration or
re-acceleration
Collision
between
ejecta and
dense cloud
Reverse
shock
Modification
of shock by
CRs
?
Energy
spectrum
of CRs
SNR
Evolution
?
?
25. We use data from X-ray Multi-mirror Mission (XMM-Newton).
(above) XMM-Newton
satellite and (left) its X-ray
telescope’s basic structure.
26. CCDs of MOS1 (left) and MOS2 (right) cameras CCDs of pn camera
European Photon Imaging Camera (EPIC)(EPIC)
Welcome everyone. In the beginning My project as you see from the title is about studying supernova remnant in X-ray band. The remnant I investigated is called SNR W28. Basically, my project’s purpose is to attempt to detect non-thermal radiation associated with particle acceleration in W28, and study the plasma properties of W28.
Because I don’t expect everyone here familiar with supernova remnant, it is better to talk about what SNR actually is. As the name tells us, the SNR was produced by a supernova explosion occurring at the end of the life of a massive star. The mixture of ejected stellar material from the supernova and the surrounding interstellar medium, compressed by the forward shock, forms the supernova remnant.
According to the standard evolution, the supernova remnant will undergo three main phases when ejecta and forward shock are expanding outward to interstellar medium. The first phase is called free expansion phase, in which the ejecta is sweeping the surrounding gas and expanding without deceleration. Meanwhile we can see the shell of ISM formed at the boundary due to compression by the forward shock.
As long as the mass of interstellar medium swept by ejecta is comparable to the ejecta mass, then the SNR evolves into adiabatic phase, in which ejecta slows down significantly and merge into forwad shock, energy loss is due to adiabatic expansion only. This phase only last for thousands of years but it is believed that efficient particle acceleration can occur by the strong forward shock during the adiabatic phase.
Once the temperature becomes lower than a million Kelvin, and cooling by radiation is efficient, then the SNR enters the radiative phase. No efficient particle acceleration is expected because of slow forward shock.
When the velocity of the expanding ejecta drops to the typical value of ISM, then the ejecta totally merge with ISM, and the remnant disappears.
Astronomers used to classify SNRs into four types.
Shell-type is the common one which has non-thermal X-ray emitting shell near its boundary, as shown by the classical example Cas A.
Second type is called Crab-like in which we can see non-thermal X-ray or radio radiation from a fast spinning neutron star, called a pulsar, at the remnant's center.
If the remnants appear as shell-like and Crab-like at the same time, we usually called them composite-type.
The latest type is called mixed-morphology SNR. Mixed-morphology SNRs have radio emitting shell instead of non thermal X-ray shells like shell-type, and their central brightness is associated with thermal X-ray radiation emission coming from SNR material instead of non-thermal emission from pulsars. Also, uniform temperature profile is observed in mixed-morphology SNR.
My target, W28, belongs to the mixed-morphology SNR. It is one of oldest remnants in the Galaxy, and it is likely that W28 has evolved into its radiative phase.
The image at top right-hand corner is the radio image of W28. We can see the bright radio shell near the boundary. The red contour lines represent the X-ray emission observed by ROSAT in 2002.
From the previous X-ray observation with ROSAT and ASCA, there is single thermal component
My project was initially motivated by gamma-ray emission detected by H.E.S.S. near the northeastern region of W28.
Gamma-ray emission, as we know, can be produced through IC scattering or non-thermal bremsstrahlung both associated with electron acceleration, or neutral pion decay associated with proton acceleration.
In order to determine which kind of particle acceleration mainly contribute to the gamma-ray emission, we need detection of non-thermal X-ray emission from the northeastern shell for broadband modeling
The data we used was downloaded from the public data archive. All useful data is contained in event files. We need to inspect the event file and check if solar flare exists in the event files by looking at their light curves. If the event files are all right, we extract the spectra of regions form the event file and do the background subtraction to eliminate any contribution of background noise to the spectra. Then load the spectra into XSPEC, provided by NASA, for spectral analysis and spectral fitting.
For the northeastern shell, I investigated 10 regions. On the other hand, I studied 8 regions from Chandra data.
I used the models available in XSPEC to fit the spectrum of a region to derive its temperature and abundances of different elements.
For the northeastern shell, I was able to best fit all the regions using two thermal X-ray emission models. I am going to just show you the resultant spectra of several regions. You can see they are almost the same. The interesting thing is that we detect oxygen emission line which is unable to be detected by previous generation of X-ray telescopes before XMM and Chandra.
The table on the left side shows the best fit parameters of several regions. The spectra of all regions in the northeastern shell are best fitted by two thermal components, different to the previous observation with ROSAT and ASCA, in which they observed single thermal component only.
Although we could not detect powerlaw component in any region from XMM data for braodband modeling, we could still estimate the constraint on the non-thermal X-ray radiation. To do so, I intentionally add a powerlaw component with a photon index fixed to be the value of 2.2, then fine-tune the normalization of the powerlaw component unit 3 sigma confidence was achieved. Then the resultant photon flux we determined in this way would be the upper limit of photon flux of the non-thermal X-ray at the NE shell. The upper limit we found is about 0.2eVcm-2s-1 at around 1keV.
The XMM observatory contains three EPIC detectors: two MOS cameras and a pn camera. All three cameras can detect photons with energy between 0.15 and 15 keV. Each EPIC detector has the field of view (FOV) of 30 arcmin and energy resolution of 0.15keV at
1keV.
ACIS is composed of 10 planar, 1024 times 1024 pixel CCDs. Each CCD provides 8.4’times8.4’ observational view on the sky. Six of them forms a linear array and the rest is arranged as a square array. Up to 6 ACIS CCDs in any possible combination can
be operated simultaneously. In our case, I0,I2 and S2-S5 CCDs were used during the observation.