A Coarse 3D Model of E0102-72 Derived from HETG Data

1
A Coarse 3D Model of E0102-72 Derived from HETG
Data

• by Dan Dewey
• MIT Center for Space Research
• Thanks to "3D SNR" collegues at MIT
• Johns Davis Houck, Mikes Noble, Nowak Wise,
• Claude Canizares, Kathy Flanagan, Amy Fredericks,
• Glenn Allen, Norbert Schulz, Mike Stage
• Contact dd_at_space.mit.edu

2
HETG Analysis of E0102 Ne X

• Analysis of the Ne X dispersed images suggests
  regions of red and blue shift appearing on the
  sky as displaced rings.
• Red 900 and 1800 km/s
• Green -900 km/s
• Blue -1800 km/s
• See Flanagan et al. 2004, ApJ
• Interpret this as cylinder viewed almost end-on

"True-color" image (above) Ne X line
"color-velocity" map
3
O III Long-slit spectrum
"Thanks to You-Hua Chufor thedata! "
FP image of E0102
Long slit spectrumin O III, 5007 Å,shows
similar spatial-velocity structure tothe Ne X
X-ray line.
- 1800km/s
- 900km/s
900km/s - 1800km/s
? But where's the red-shifted emission in the
optical flux. Perhaps see next slide.
4
An aside
Transmission of gas dust
O VI
Perhaps there's equatorial dust absorbing the
back-half of the ejecta in the optical?
X-ray
O III
Dust attenuatesthe 5007 Å line butnot X-ray
orIR emission.
Dust?
Spitzer IR bands
X-ray and IR black bodies
Observer
O, Ne, etc. ejecta
Blastwave sphere
5
Constructing a 3D Model
Analogous to the Ne X resultconstruct cylinders
of emissionfrom each ion. From Flanagan et al.
2004the radii of the cylinders varyfrom ion to
ion, see Table below. Also include a
Blastwavesphere of emission (Hughes 1994)with a
vnei spectrum andabundances 0.3 solar.
ions
6
Cylindrical Components
Ion Lambda (Å) T_e(keV) Cylinderaverageradius(")
O VII 21.8 0.34 12.50
O VIII 18.97 0.34 13.95
Ne IX 13.57 0.58 13.40
Ne X 12.13 0.58 15.00
Mg XI 9.17 0.50 14.50
Mg XII 8.42 0.50 16.20
Si XIII 6.65 0.60 15.30
7
Encoding the Model
Spatial
O VIIemissionmeasurearray
Each component is a3-D array of values,(e.g.,
873 elements)giving the "norm" ofthe spectral
model ineach local cell.
Spectral
"Single ion" (or "vnei") spectra are createdat
nominal values of T_e and for v0.
local EM
8
"Evaluating" the Model

• Assume the source is optically thin
• concatenate component photon/count outputs.
• Create "output" photons from each component as
  a
• random spatial location determined by "norm"
  array values.
• and project to 2-D on sky location.
• random spectral wavelength based on
• spectrum assigned.
• number created determined by spectrum
• and optionally an "arf" as well.

v r
los
r

• Modify wavelengths using
• line-of-sight angle for Doppler
• v r spherical velocity field.

9
Comparing Model and Data

• Fold simulated photons through instrument, e.g.,
  with a ray-trace.
• Or modify simulated counts to approximate the
  instrument effects.
• Variety of HETGS observation outputs that can be
  compared, e.g.,
• MEG minus order
• MEG plus order
• Zeroth-order image
• Zeroth-order spectrum
• Adjust "norms" of the components for overall
  coarse agreementof model to data - do it by hand
  with human "comparison".
• Use automated fitting to do fine adjustment of
  parameters.

10
Z-o Data and Model
Data
Comparing the Zeroth-order image and
theZeroth-order Spectrum.
The real data includesbackground eventsand
morespatial complexitythan the simple model.
Data
Model
Model
These images are in detector coord.s. Roughly,
North is downand East is to left.
11
MEG-Dispersed Data and Model
Data
O VII
Ne X
MEG -1
MEG 1
Fe needed?
Ne X
Model
O VII
MEG -1
MEG 1
12
Comments on Dispersed Data-Model
The Model reproduces the coarse structure of
the Data. The velocity Doppler effect (due to
the cylindricalion emission and spherical
velocity field) shows upin both Data and Model
as narrower/crisper minus-order ring images.
Is there a need to add some Fe lines in a ring
component?E.g., to add emission between O VIII
and Ne IX lines?
13
Model Calculations
Given the coarse model,the product (n_e n_i)
foreach ion species isdetermined in 3D
space. At right the valuesalong a radial
sliceare plotted. In the plot on the nextpage
the values n_e(r)and n_i(r) have
beenself-consistantly calculated.
Blastwave,n_e n_H
Ne IX
Ne X
O VIII
O VII
O IX
O VI
Mg XI
Ne VIII
Mg XII
Mg X
14
Radial Profile of Densities
n_e
n_H, Blastwave
Blastwavedensities ofelements(0.3 solar)
Mg XI
O IX
O VII
O VIII
O VI
Ne X
Ne IX
Ne VIII
Mg XII
Mg X
n_O
n_Ne
n_Mg
15
Calculating Element Masses
Ion Mass, M_solar
O IX 0.91
O VIII 2.43
O VII 1.92
O VI 1.03
O TOTAL 5.38

Ne X 1.04
Ne IX 0.66
Ne VIII 0.43
Ne TOTAL 2.13

Mg XII 0.08
Mg XI 0.35
Mg X 0.28
Mg TOTAL 0.71

Si XIII 0.12
Blastwave (H,He,etc.) 103.4
Using the values of n_ion obtained above,the
mass of each element in the model canbe
calculated as the sum of masses of theindividual
ions, see Table at right. Note that non-X-ray
visible ions such asO VI and Ne VIII have been
included atlevels with comparable densities to
theX-ray visible ions --- compatible with
auniform homogeneous ejecta elementdistribution.
16
Further work
Another aside Is E0102's morphology/flux
changing over the 3 yearsbetween HETGS
obervations ? Due to ionization changes rather
than motion ?

• Add spatial complexity to the model, e.g.,
  azimuthal variation

Note that the 3D array implementation allows
adjusting each of the 3Darray values to create
essentially arbitrary emission distributions.

• Develop better comparison methods and tools

There is a lot of information in the data - on
small and large scales.

• Compare multiple data sets

A single model can be used to output simulated
HETGS, XMM/RGS, andAstro-E2/XRS data
simulations. Compare each/all of these to their
realdata sets and adjust the model appropriately.
17
Hydra The Bigger Picture
The work to create a 3D model of E0102 described
above is part of a larger effort we've named
"Hydra". The block diagram at right shows the
many heads of Hydra - the many components to the
full 3D modeling process. The E0102 activies
above are used to describe Hydra concepts and
goals in thepages at right.
(Hydra graphic from http//www.pantheon.org/areas
/gallery/folklore/greek_heroic/hydra.html)
18
Hydra Block Diagram
19
Hydra 3D Source Modeling andX-Ray "Rendering"
The method of using 3D arrays to specify
propertiesof the 3D model described above is
just one of manyways to encode 3D information.
Other possibilites includeanalytic descriptions
with a component algebra. The generation of rays
from the model can be viewedas a form of
"rendering" and in the extreme it takeson all
the complexity of radiation transport.
Efficienciesin rendering may be improved by
matching the source modelproperties with the
method of encoding the problem.
20
Hydra Comparing Data and Folded Model
One issue to explore more is the method to
comparemulti-dimensional data sets, e.g., the
dispersed imagesfrom "real data" with a forward
folded (ray-traced)simulation output. It may be
important and useful for human insight toguide
the selection and definition of a fit metric
fordifferent fitting situations. For example,
the ratioof flux (projected) interior to the
E0102 ejecta ringcompared with the flux in the
limb of the blast wavemay be very sensitive to,
say, the blastwave shellthickness.
21
Hydra Instrument(s) Modeling
One goal of the Hydraeffort is to allow
andsupport multiple datasets from
multipleobservatories. As a simple
example,the spectral image andhistogram at left
werecreated from the coarse E0102 model and
foldedthrough a simple approx-imation of the
XMM RGS Grating response.
O VIII
Ne X
O VII
The width of the lines is due toboth the spatial
and velocitystructure in the 3D model.