![Page 1: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/1.jpg)
1
STIS Calibration
Enhancement: Implementation of Model
Based Techniques for Wavelength Calibration and
Charge Transfer InefficiencyHST Calibration Workshop 26th October 2005
Paul Bristow, Florian Kerber & Michael Rosa
www.stecf.org/poa
![Page 2: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/2.jpg)
2
Empirical Wavelength Calibration - Quick Review Polynomial dispersion solution:
Deep “wavecal” observations of on-board lamp
Line list for on-board lampFine tuning with contemporaneous short
wavecal Finds zeropoint offsets due to monthly offsets,
MSM uncertainty & central wavelength
![Page 3: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/3.jpg)
3
Short and Long Wavecal
![Page 4: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/4.jpg)
4
Short and long wavecal, detail
![Page 5: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/5.jpg)
5
Echelle model Matrix based model of
STIS optical components ~35 parameters Inversion is not
possible
![Page 6: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/6.jpg)
6
STIS Spectrograph Model Optimization with deep wavecals establishes
exact configuration of fixed optical components Change in central wavelength for different MSM
positions turns out to be a linear function of just one of the physical model parameters.35 parameters 2
Compare to the ~7parameters of the empirical model
![Page 7: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/7.jpg)
7
Simulated Wavecals
Config determined with optimisation can be plugged back into the model to produce simulated wavecals.
Given a model SED, astronomical sources can also be simulated in this way - advanced ETC
![Page 8: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/8.jpg)
8
Other visualisation products
![Page 9: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/9.jpg)
9
Optimisation of Contemporaneous Wavecal
Find strongest features Optimise just X-disperser
angle NO NEED FOR ANY
FURTHER SHIFTS cf A1CENTER &
A2CENTER (nominal location) , MOFFSET1 & 2 (MAMA commanded) and SHIFTA1 & 2 (wavecals) Shifts used by calstis
![Page 10: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/10.jpg)
10
Science case
Thermal Pressures in ISM from CI Fine-Structure Excitation
Jenkins, E.B. & Tripp, T.M., 2001, ApJS 137, pp297
E140H, 1180Å-1280Å O5LH01010-1040, O5LH02020-2040, O5LH08010-8030
Authors noted that features in order overlaps had discrepant wavelengths
![Page 11: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/11.jpg)
11
Order Overlap
![Page 12: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/12.jpg)
12
Order Overlap
![Page 13: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/13.jpg)
13
Model BasedCTI Correction
![Page 14: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/14.jpg)
14
Large charge packets fill available traps Subsequent smaller packets receive the charge emitted by
the traps Relative positions of sources can be important Remember: Profile in figure could be vertical x-section of
stellar image or stellar spectra
•All other “CTI effects” are manifestations of the effects upon x-section:
•Flux loss•CTI tails•Centroid shift•PSF degradation
How CTI effects x-section
Readout directionRed is observed
![Page 15: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/15.jpg)
15
Empirical Solution Estimate flux loss from point sources as a function of:
Position on chip Size of charge packet (flux) Background Time on orbit
Only applies to photometry, not astrometry, structure etc. Only applies to the type of objects it was calibrated for
(usually point sources or spectra of point sources) Requires extensive calibration observations Cannot account for image complexity
![Page 16: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/16.jpg)
16
Readout model: Motivation
A physical solution, CTI effects understood in terms of the CCD operation and environment
A correction that applies to the entire 2D array: Equally valid for imaging and spectroscopy Also applicable to extended objects
Portable to other space based detectors (WFPC2, ACS, WFC3?, Kepler instruments etc.)
Extendable to model further aspects of CCD readout
![Page 17: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/17.jpg)
17
Readout Model: Concept Forward Simulation:
Start with 2D Charge distribution on chip Distribution of bulk traps
Shift the charge under each electrode as during readout
Calculate capture and emission of charge in each shift
Take into Account: Status of bulk traps Dark current Trap capture and emission timescales Chip clocking frequency, architecture, gain etc.
![Page 18: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/18.jpg)
18
Restoring data with simulation results CTI effect is grossly exaggerated in figure! --- Original profile --- Raw (observed) profile, after CTI readout --- Raw profile + sim CTI --- Raw profile - sim CTI (red-[blue-red]) Further iterations will converge on original profile
Readout direction
![Page 19: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/19.jpg)
19
Correction:
Further iterations (usually not necessary):
Correctedimage
Differenceimage
Simulationoutput
Rawdata
=-
=
Rawdata
-
Differenceimage
Rawdata
>> SIMULATION
Correctedimage
>> SIMULATION
Rea
do
ut
dir
ect
ion
![Page 20: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/20.jpg)
20
Rea
do
ut
dir
ect
ion
![Page 21: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/21.jpg)
21
![Page 22: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/22.jpg)
22
Quantitative Validation Matches photometric and spectroscopic empirical
corrections (Goudfrooij & Kimble 2002, Bohlin & Goudfrooij 2003)
well with physically realistic parameters Charge restored reliably to central isophotes of point
sources => whole image array is corrected reliably. 2 CE-STIS ISRs describing these results Large parameter space explored in order to find the
optimum model configuration => physical properties of STIS CCD
![Page 23: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/23.jpg)
23
Comparison to Empirical Corrections for Imaging Data Empirical calibration used
as a test of physical model On average we should
expect agreement This comparison can also
help to calibrate the physical model
Good general agreement (especially for objects to which the empirical calibration applies)
![Page 24: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/24.jpg)
24
Special Cases: Images
Rea
do
ut
dir
ect
ion
![Page 25: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/25.jpg)
25
Science Case: Proper motions of local dwarf galaxies Slawomir Piatek, Carlton Pryor, Edward
Olzewski et al. AJ126,2346 Nov 2003 (Carina) &
AJ124,3198 Dec 2002 (Fornax) Highly accurate astrometry of stars in local
Dwarfs (Carina, Sculptor, Fornax) and background QSO (as reference)
Repeat observations with STIS and WFPC2 taken several years apart.
![Page 26: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/26.jpg)
26
Proper motions - CTI effects
CTI during readout will shift the centroids Row dependent shift Luminosity dependent
shift Epoch dependent shift
- later observations suffer significantly more CTI
![Page 27: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/27.jpg)
27
Proper motion data with and without CTI correction
Rea
do
ut
dir
ect
ion
March2000
March2003
![Page 28: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/28.jpg)
28
Proper motions - Full Analysis See poster from Piatek et al Piatek et al. AJ, 130, pp95 (Ursa Minor) Piatek et al. AJ, in press (Sculptor) Data processed for further Carina study
![Page 29: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/29.jpg)
29
Summary -STIS Pixel based CTI Correction Status
Pipeline version delivered to STScI Feb’ 04 Testing delayed (STIS safe mode in mid 2004) Imaging results not bad:
Point sources, sparse field - As good as empirical Point sources, crowded field or in presence of artefacts -
Better than empirical extended sources - Only approach available Proper motions science case
Application to spectra not as good Slow:
Could be more efficiently coded… Any pixel based solution will necessarily be computationally
expensive
![Page 30: HST Calibration Workshop 26th October 2005 Paul Bristow, Florian Kerber & Michael Rosa](https://reader035.vdocuments.net/reader035/viewer/2022062723/56813b7f550346895da4a23c/html5/thumbnails/30.jpg)
30