Chandra Calibration Status

1
Chandra Calibration Status
There have been 3 CALDB releases since the last
CUC meeting CALDB 4.1.0 Dec. 15, 2008
CALDB 4.1.1 Jan. 21, 2009 CALDB 4.1.2
Mar. 27, 2009.
HRMA

• HRMA encircled energy fraction released in CALDB
  4.1.0 for use with mkpsfmap.
• Updated HRMA effective area released in CALDB
  4.1.1

ACIS

• Gain correction files for epochs 35 (Aug. Oct.
  2008) and 36 (Nov. 2008 -Jan.2009) were released
  in CALDB versions 4.1.0 and 4.1.2.

HETG

• Updated HETG transmission efficiencies were
  released in CALDB 4.1.1

2
Two corrections were applied to the predictions
of the raytrace code before CALDB 4.1.1
Updated HRMA Effective Area
Empirical XRCF correction
HRMA overlayer of 22A
3
Re-Analysis of the Ground-Based (XRCF) Data
SSD Continuum Measurement
HETG Continuum Measurement
4
Empirical Correction to Updated HRMA Effective
Area
5
XMM-Newton / Chandra Cross-Calibration
6
Current HRMA Calibration

• Do the mirror shutters block a greater fraction
  of the beam than presently predicted by the
  raytrace code Analysis of HSI ring focus images
  show that the shutters reduce the throughput of
  the mirrors by at most 1.
• Is a greater fraction of the beam scattered
  beyond the 2mm pinholes used for effective area
  than presently predicted by the raytrace code
  The measured 2mm/35mm flux ratio is consistent
  with the predictions of the raytrace code below 2
  keV and is approximately 3 less than the value
  predicted by the raytrace code at 6 keV.
• Improve the pile-up correction for the SSD
  continuum measurements.
• Investigate the discrepancy between FPC and SSD
  measurements.

7
HIS Ring Focus Images
8
Measured FPC 2mm/35mm flux ratio vs. energy

• Improve the pile-up correction for the SSD
  continuum measurements.
• Investigate the discrepancy between FPC and SSD
  measurements.

9
Update on the ACIS Contamination Model
ACIS Calibration Status
10
Spectrum of ECS L-Complex
F-K edge
11
ECS data fit with the old and new spectral model
for the L-complex
Old Spectral Model
New Spectral Model
Old Spectral Model
12
Decline in ECS Flux with Time
13
Optical depth of the contaminant at 700 eV on
ACIS-S
14
Optical depth of the contaminant at 700 eV on
ACIS-I
15
Revised HETG Efficiencies
CALDB 4.1.1
16
HRC-I/ACIS Cross-Calibration with Updated HRMA
Model
17
HRC-S QE Re-Calibration

• HRC-S 0.5-10keV QE calibrated and fine-tuned
  in-flight using LETG blazar continua
• HRMA EA revision requires QE re-calibration
• Account for
• HRMA EA changes
• Accumulated in-flight data suggesting overly-hard
  power-law indices than HETGACIS by of order 5
• Will require relative QE correction between
  0.5-5keV of 15

Power law Fits to contemporaneous HETG and LETG
observations of PK2155-304
HEG
MEG
LEG

• Complicated by
• Spectral shape/intensity variability
• Parameter non-orthogonality
• Possible PL curvature

18
CUC Calibration Wish List
1) Astrometric CalibrationDo observers have
ready access to any study of the
astrometriccalibration of the mission PSF tools,
e.g. if a field of sourceswith well-known
positions is both observed and simulated by
thePSF tools, how well do the observed and
simulated data sets line up?
Astrometry - The overall 90 uncertainty in
absolute positions is 0.6. This is given on the
CXC web pages and in the POG. The mean off-set
between CSC and SDSS positions is 0.2 On-axis
PSF - There are several on-line Chandra
calibration workshop presentations as well as a
discussion in the POG on this topic. The CSC
Catalog Statistical Characterization memo shows
that the uncertainties in relative astrometry for
bright point sources is 0.2 within off-axis
angles of 10 and approximately 1 at larger
off-axis angles. A study has begun to simulate a
Chandra observation of the open star cluster
NGC2516 with SAOTrace and then compare the x-ray
source positions as determined by wavdetect with
the optical positions.
19
2) ACIS Bad Pixel TableData discarded via
entries in the ACIS Bad Pixel Table
(obviously)reduces the effective area of the
mission.  Currently, the standardBad Pixel Table
aggressively removes ACIS columns and is not
necessarilyappropriate for all kinds of ACIS
science.  Modifications to the ACISBad Pixel
Table should pass through an approval process
which includesconsultation with the Users
Committee.  Several different ACIS Bad
PixelTables, each optimized
Data systems is revising the CIAO tool
acis_build_pixel so the user has much more
freedom in specifying the bad pixels.
20
3) RMFs and ARFsUsers need the capability of
creating RMFs and ARFs for user-selectedgrades. 
There are several benefits  - for bright
piled-up sources filtering on only single events
will    in general reduce pileup  - for faint
non-piled-up sources a non-standard grade
selection    can increase effective area  - for
bright non-piled-up sources a non-standard grade
selection    may improve energy resolution.
Since the Chandra beam is comparable to the size
of an ACIS pixel, grade 0 events are not
necessarily un piled-up. The standard grade set
is 02346. In general, grades 1 and 5 add about
1-2 to the standard grade set. For soft
sources, grade 7 is heavily dominated by charged
particles. For photon energies greater than 6
keV, grade 7 events can add an additional 10 to
the standard grade set, however the background
rate is 75 times higher when including grade 7
events.
21
ACIS Flight Grades
Grade 0 single pixel Grade 1 Diagonal
split Grade 2 vertical split Grades 3 and 4 -
horizontal splits Grade 5 L-shaped
corner Grade 6 L shaped Grade 7 everything
else
22
Grade 0 fraction vs. count rate
23
E0102-72
.
ACIS-I3
ACIS-S3
Grade 02346 1
Grade 02346 1
5
7
5
7
24
ECS Data
25
G21.5-09
.
ACIS-S3
26
E0102-72
.
ACIS-S3
ACIS-I3
27
Cas A
.
ACIS-S3
ACIS-I3
28
4) ACIS Contamination Model 5) Soft Proton
ContaminationProper statistical treatment of
backgrounds requires a model of thebackground
spectrum, but what is this for the soft proton
contamination?In addition, what is the spatial
distribution of the soft protoncontamination? Is
it uniform, or is it vignetted at all by the
HRMA?
The most common type of BI flare is very
well-behaved, has a reproducible spectrum (a
power law with an exponential cutoff (see
Markevitch et al. 2003, ApJ 583, 70), and a
spatial distribution that is non-uniform but
energy-independent. These results were presented
at the 2003 Chandra Calibration
meeting (http//cxc.harvard.edu/ccr/proceedings/03
_proc/presentations/markevitch2/s005.html) This
information can be added to the CIAO thread for
generating background images for extended
sources. We can also provide a tool which will
add in the background flare component in a blank
sky image.