Transcript Document

Coronagraphy with HST/NICMOS*
*The Near Infrared Camera & Multi-Object Spectrometer
Extending HST’s UV/Optical Panchromatic Vision
into the Near IR (0.8mm —2.4 mm)
Glenn Schneider
Steward Observatory, University of Arizona (NICMOS/IDT)
Hubble Space Telescope
Third Calibration Workshop
18 October 2002
Baltimore, Maryland
http://nicmosis.as.arizona.edu:8000
[email protected]
NICMOS Coronagraphy Takes Advantage of HST’s
Unique Venue for High Contrast Imaging
• Diffraction Limited Imaging in Optical/Near-IR
•> 98% Strehl Ratios @ all ls
Background Rejection*
• Highly STABLE PSF
1.6mm: ~10-6 pix-1 @ 1”
1.1mm: ~10-5 in 2”-3” annulus
• Highly Accurate Pointing
Repeatability
& Control
*w.r.t. central pixel
• Intra-Orbit
Fcentral(H) = 11% Fstar
Field
Rotation
NIR High Dynamic Range Sampling
NICMOS/MA: Dmag=19.4 (6 x 4m)
Scientific Areas of Investigation Enabled
With Today’s Capabilities on HST
via PSF-Subtracted Coronagraphic Imaging
Young Extra-Solar Planet* &
Brown Dwarf
Companions
HR 4796A
1"
Circumstellar Disks
fdisk/f* > few x 10-4 at 1”
q = 0.1”
TWA 6
* < few x 106 yr at 1”
Damped La Absorbers LBQS 1210+1713
Cooling Curves for Substellar Objects
0
Evolution of M Dwarf Stars, Brown Dwarfs
and Giant Planets (from Adam Burrows)
Log L/L(sun)
-2
200M jup
80M jup
-4
-6
14M jup
-8
STARS (Hydrogen burning)
BROWN DWARFS (Deuterium burning)
JUPITER
PLANETS
-10
6
7
SATURN
8
Log10 Age (years)
9
10
Planet-Building Timeline
Taurus,
Ophiuchus
star forming
regions
TW Hydrae Tucanae
Hyades
Assoc Pleiades
Assoc
a Persei
106
yr s
Collapsing
protostar
forms protoplanetary disk
107
yr s
108
yr s
Giant planets
accrete
gaseous
atmospheres
Rocky cores
of giant
planets form
Era of heavy
bombarment
by comets
Terrestrial
planets
form
Disk Evolution/Dissipation(?)
Primary Dust (≤ mm) Secondary Dust (≥mm)
Locked to Gas
Collisional erosion
Sun
109
yr s
Current
age of
the Sun:
5x109 yrs .
Clearing of
inner solar
system,
formation of a
Kuiper
cometary
belt?
Clearing Timescales: P-R drag few 10 6
Rad. Pressure: ~ 104
From: R. Webb
Coronagraphic Companion Detection
PSF “Roll Subtraction”
Separate into Positive
& Negative Conjugates
Coronagraphic Images
D Orientation = 30°
HD 102982
H = 6.9
G3V
DH = 5.3
r = 0.9"
Difference Image
Rotate about Hole
Center and Co-Add
(Resampled)
 (Multiaccum) Imaging at two S/C orientations in a single HST visability period.
 Background objects rotate about occulted Target. PSF and optical artifacts do not.
 DRoll = 30°, DTime ≈ 20 min., Total time per Orientation ≈ 11 min.
Combined detection floor in absence of background light: H ≈ 23
Combined detection floor in absence of background light: H ≈ 23
H = 21.9
r = 9.36”
DH = 12.6
TA Persistence
Ghost Images
H = 22.3
r = 13.34”
DH = 12.9
LHS 3003
H = 9.3
Coronagraphic Performance (G2V)
.
Radius (Pixels) from Hole Center
5
7
9
11
13
B
A
C
K
G
R
O
U
N
D
16
R
E
D
U
C
T
I
O
N
9
15
14
13
12
11
10
8
INTENSITY (AZIMUTHAL AVERAGE)
10 0
5
17
19
21
23
25
27
29
31
33
35
37
REDUCTION IN BACKGROUND FLUX FROM F160W PSF
w.r.t. central pixel
Fcentral(H) = 11% Fstar
10-1
Unocculted PSF
Coronagraph
Coronagraph & PSF Subtraction
10-2
1
pixel
10-3
10-4
10-5
7
6
15
10-6
0
Coronagraphic
Hole
Radius = 0.3"
0.075 0.15 0.225
0.3 0.375 0.45 0.525
0.6 0.675 0.75 0.825
0.9 0.975 1.05
ARCSECONDS
4
3
2
0.3 0.45 0.6 0.75 0.9 1.05 1.2 1.35 1.5 1.65 1.8 1.95 2.1 2.25 2.4 2.55 2.7 2.85
Radius (Arcsec) from Hole Center
General Description
Coronagraphic Field of View
• NICMOS Coronagraph is in Camera 2*:
256 x 256 pixels @ ~ [76.2, 75.5] mas / pixel
FOV ~ 19.49” x 19.33” (377 ”)
0.9% X:Y Linear Geometrical Distortion
• Radius of Occulted Region = 0.3”
Size “Optimized” for H-band Imaging (1st Airy Ring fully contained)
~ [+73, -45] pixels (or [+5.6”, -3.4”] from [-X,+Y] corner of FOV
• Field Asymetric w.r.t Occulted Star
For maximum S/C Roll (at one epoch) of 29.9°:
475 ” Survey Area with
280 ” Overlap Area
* http://www.stsci.edu/hst/nicmos/performance/platescale/rel_platescale.html
Coronagraphic Companion Detection
PSF “Roll Subtraction”
EXAMPLE: TWA 6, H = 6.9
Two Integrations from Median of 3 Multiaccums Each
Total Integration Time = 640 seconds at Each Orientation
DRoll = 30°, D Time = 20 minutes
Linear Display 0—20 ADU/sec/pixel; 2.19E-6 Janskys/ADU/sec/pixel
Coronagraphic Companion Detection
Unresolved (Point-Like) Object: H =20.1, DH = 13.2, r=2.5”
Linear Display 0—2 ADU/sec/pixel; 2.19E-6 Janskys/ADU/sec/pixel
Coronagraphic Companion Detection
PSF “Roll Subtraction”
Difference Image: H =20.1, DH = 13.2 (La/Lb = 200,000:1), r=2.5”
At r=2.5” background brightness is reduced by an ADDITIONAL
factor of ~50 over raw coronagraphic gain (of appx 4).
Linear Display -0.4 — +0.4 ADU/sec/pixel; 2.19E-6 Janskys/ADU/sec/pixel
Coronagraphic Companion Detection
PSF “Roll Subtraction”
Each independent point- Geometrical Rectification
And De-Spiking*
source image is S/N ~ 20
*NICMOS/IDT
Post-Processing & Analysis S/W: DSKP & IDP3
ftp://nicmos.as.arizona.edu/
Linear Display -0.4 — +0.4 ADU/sec/pixel; 2.19E-6 Janskys/ADU/sec/pixel
Coronagraphic Companion Detection
PSF “Roll Subtraction”
“Final” Image After Additional Post-Processing
2.0
1.0
0.0
1.0
Arc Seconds
2.0
TWA6
PSF FWHM = 0.16"
Image Combination
& Spatial Filtering
A spatial filter is applied to
the combined image to further
reject image artifacts with
characteristic frequencies
not commensurate with the
size of a stellar PSF.
1
S/N ~ 35
I(r)/I(peak)
0.8
0.6
0.4
0.2
NICMOS
F160W
25 OCT 1998
Camera 2 (0.076"/pixel)
Coronagraph (0.3" radius)
Integration Time =1280s
0
0
1
2
3
4
Radius (Pixels)
5
6
Sensitivity (S/N=25) vs. Detectability (50% Probability*)
H-Band Two-Roll Coronagraphic PSF Subtraction 22m Total Integration
DH(50%) = 9.7±0.3 + 2.1 x r {M–G Stars}
Delta-H
TWA6 and Median of 50 G-K Stars in NICMOS Survey
* Determined by Noise Statistics AND Model Star Implantation
PRELIMINARY Post SM-3B Coronagraphic Performance
Characterization for HST Cycle 11/12
Data from SMOV3B Test Programs:
• Coronagraphic Target Acquisition Test*
• Coronagraphic Focus Verification*
• Initial (Part 1) Performance Check - Characterization†
*Executed
Prior to “Final” Plate Scale / Aperture Rotation Updates
†Executed Prior to Low Scatter Point Determination / Adjustment
To Be Executed (Next Week) Under Cycle 11 Cal Program
• Coronagraphic Light-Scatter Minimization
• “Final” (Part 2) Performance Check - Calibration
Coronagraphic First Light Post-SM3B
“Out of the Box”
CYCLE 7
GTO/7227
CYCLE 11
SMOV3B/8983
0 — +2.0 ADU/sec/pixel
2.19E-6 Jy/ADU/sec/pixel
0 — +2.75 ADU/sec/pixel
1.59E-6 Jy/ADU/sec/pixel
Coronagraphic First Light Post-SM3B
“Out of the Box”
CYCLE 7
GTO/7227
CYCLE 11
SMOV3B/8983
-0.4 — +0.4 ADU/sec/pixel
2.19E-6 Jy/ADU/sec/pixel
-0.55 — +0.55 ADU/sec/pixel
1.59E-6 Jy/ADU/sec/pixel
I(Direct)/I(Coronagraphic)
25
F
1
6
0
W
Coronagraphic Performance (M9.5V+)
Direct/Coronagraphic
20
15
10
5
100-3
Coronagraphic
Direct (Total = 23700 ADU/sec)
I(pixel)/I(star)
10-4
10-5
10-6
10-7
10-8
Direct
Coronagraphic
10
20
30
Radius (Pixels)
40
50
Coronagraphic PSF-Subtraction Induced Image Artifacts
The Dominant Source of Systematic Error (“Noise”)*
Imperfections in PSF-subtractions
result in residuals larger than
expected from pure photon noise.
Systematics:
OTA “Breathing”
Target Re-centration
Coronagraphic Hole Edge Effects
Cold-Mask “Wiggles”
Opto-Mechanical Stability
*For
properly reduced/calibrated images
“Breathing” - The Coronagraphic Nemesis
De-spaceing of the HST secondary mirror along the telescope optical
axis from (orbit driven) thermal instabilities in the OTA causes
variations in the PSF structures which are typically THE dominant
source of systmatic errors in coronagraphic PSF subtraction.
The thermal time constant of the OTA is longer than sub-orbit
timescales.
“Two roll” coronagraphic observations should be completed in a
single target visibility period to minimize PSF variations.
Reference PSFs should be obtained as close in time (very preferabley
in the same visibility period) as target images WITHOUT any
intervening changes in Sun angle.
Coronagraphic Optics General Description
• Coronagraphic “Hole” On Camera 2 Field Divider Mirror
@ OTA f/24 Focus
Physical Radius: 170mm
Projected Radus: 0.3”
• Lyot Stop (85% Unobscured Area)
At Cold Pupil in VCS (near Filters)
Obscurations for (warm):
Primary Mirror Outer Edge
Secondary Mirror Housing
Primary Mirror Hold-Down Pads
F160W PSF “Mapped” Onto Coronagraphic Hole
A small change in energy distribution in the first Airy ring (due to
breating induced focus shifts) cause scattering sites on the hole-edge
to “light up” and change, significantly, the downstream scattering.
HH30 Obscured A Few
GMWords
AUR on
Unembedded
Circumstellar
(AvDisks…
< 0.5)
Red Polar Lobes
10 mJy arcsec-2
Direct Image
Lower Scattering Surface
0.2 mJy arcsec-2
Faint Blue Ribbon
J* = 0.33 Jy
H* = 0.40 Jy
Observing young circumstellar
disks
Coronagraph
+
With obscured central stars isPSF
notSubtraction
difficult.
Disk systems with unembedded, or only
marginally obscured central stars are
much more observationally challenging
and require PSF-subtracted coronagraphy.
For Disk Imaging, to Minimize Image Artifacts Resulting from
Reference Subtraction, Reference PSFs Should Be:
• Obtained in the same visability period as the target whenever possible
• Of Similar Spectral Type (Within One Spectral Class)
• At Least as Bright as The Target
HR 4796A
For Disk Imaging, to Minimize Image Artifacts Resulting from
Reference Subtraction, Target Images Should Be
• Obtained at Two or More Spacecraft Roll Orientations
TW HYDRAE
HD 141569A
Calibrating Coronagraphic Data
Be Critical of “Pipeline” Results
Performance Levels Discussed ASSUME Properly Calibrated Data
Local and Global Deviations from True Photometric Backgrounds
MUST Be Corrected (Zeroed) Before PSF-Subtraction, Otherwise:
 Loss of Sensitivity (Against Residual Background)
 Degraded Detectability in PSF-Subtracted Images
 Photometric Zero-Point Errors
 Spatial Non-Uniformity in Detection Limits
Calibrating Coronagraphic Data
Be Critical of “Pipeline” Results
 REFERENCE FLATS:
- Hole “imprint” in CDBS flats is static, in reality it moves.
Augment Reference Flats with Contemporaneous TA Lamp Flats.
- Construct Reference Flats So As Not To Rely on Assumed High
Fidelity of Knowledge of Linearity Transfer When Approaching
Saturation. I.e., “throw away” reads > 50—70% full well when
making reference flats.
REFERENCE DARKS: See Silverstone Poster (This Workshop)
- “Synthetic” (Decomposed
Models)
Generated by OTFR vs.
Three Day
“Snapshot”
- Median ObservedOf
vs.Coronagraphic Hole
- Combined (Temperature &Motion
SAA Decay) Selected
Calibrating Coronagraphic Data
Be Critical of “Pipeline” Results
OTFR/CALNIC
10 October 2002
Using “Best” Ref Data
Flat-Field Imprint
Non-Zero Background
Quadrant Offsets
“Photometrically Challenged” Column
Dead, Grotty, Excessively Hot Pixels
CALNICA ANALOG
(+ Bad Pixel Replacement)
ObsDARKS/LinFLATS
Or…
Post-Processing Tools Exist To Mitigate
Calibration Errors, But Often Do Not
Work Well In Regions of High Flux
Densities and Large Signal Gradients
Calibrating Coronagraphic Data
Be Critical of “Pipeline” Results
OTFR/CALNIC
10 October 2002
Using “Best” Ref Data
Progressively “Better” Flat-Field / Zero-Point Calibration Sequentially
Through the Orbit Is a Tell-Tale Sign of A DC Offset Matching Problem.
Calibrating Coronagraphic Data
Be Critical of “Pipeline” Results
OTFR/CALNIC
10 October 2002
Using “Best” Ref Data
CALNICA ANALOG
(+ Bad Pixel Replacement)
ObsDARKS/LinFLATS
Post-Processing to Remove
Electronic Image Artifacts:
Saturation Bands & Echos
“Mode 2” Target Acquisition (TA)
 Target Blind-Pointed into 128x128 Pixel Acquisition Sub-Array
 Allowing for GSC errors co-ordinates must be known to ± 3.8”
 Central region of TA field-of-regard
nearly free of detector defects.
 CYCLE 11 BRIGHT OBJECT LIMIT:
H = 4.0*
*Using F187N (1%) filter
TA Performance Verified: SMOV 8979
CYCLE 11 FAINT OBJECT LIMIT:
H ~ 18*
*Acquisition in one orbit, imaging in subsequent orbit(s)
“Mode 2” Target Acquisition (TA)
Cycle 11 (77K) Exposure Time Requirements (F160W)
“Mode 2” Target Acquisition (TA)
ASTROMETRIC ANCHOR (Where is My Target?)
 TA Images with S/C Pointing & Acquisition (“Engineering”)
Data Provide Necessary Information to Accurately Determine
Occulted Target Position AFTER Offset Slew Maneuver
- Early Cycle 7: SPT file in raw “engineering” units
- Later Cycle 7: SPT file in detector pixels in FSW coordinates
- Cycle 11: _RAW, _CAL files in detector pixels (FSW)
SIAF[X, Y] = 256 - FSW[Y, X]
 May Need to Correct “Requested” vs. Actual Post-Slew
Target Position Due to Secular Change(s) in Image Scale and/or
Aperture Rotation Angle. (FSW uses Fixed constatnts).
“Mode 2” Target Acquisition (TA)
PHOTOMETRIC ANCHOR (How Bright is My Target?)
 Stellar PSF Cores Will Saturate at Shortest (0.2s) Exposure Times for:
F160W: H < 7.2
F165M: H < 6.5
F171M: H < 5.5
F187N: H < 4.0
If H < 4 Need Mode-1 Target Acquisition
 TA Images Can Be Used To Establish In-Band Magnitudes of
Target for Acquisition Filter Used.
 “Hole Locate” Lamp-Flat Background (2x7s ACCUM) Images
May Be Used to Obtain H-Band Magnitude of Target.
 For subsequent Coronagraphic Imaging in Other Filters Take
Unsaturated UNOCCULTED Images (when possible) to Establish
PSF Core Photometry
“Mode 2” Target Acquisition (TA)
ASTROMETRY / PHOTOMETRY
 BUT… TA Images are *NOT* Calibrated in OPUS Pipeline.
 Shading Biases Target Centroids With Horizontal
Field Gradients AND Photometry Through Flat-Field Errors
 TA Process PROVIDES:
- Two F160W Lamp Flat Images & Backgrounds
(used by on-board hole-location algorithm)
And… Necessary To Augment Reference Flats Used
In Calibrating Follow-On Coronagraphic Imaging
(But Not Used in OPUS Pipeline)
- Two Acquisition “ACCUM” Mode Images (for CR Minimization)
“Mode 2” Target Acquisition (TA)
ASTROMETRY / PHOTOMETRY
 Dark Current is (Generally) Not An Issue, Shading Is
TA Images may also be corrupted by “the bands”,
which could be a problem if they go through the target
 OPUS Does Not Currently Use Observed or “Synthetic”
ACCUM Darks for TA Image Processing
 Options:
- Take “ACCUM” Mode Darks (+0.025s)
- Build Source-Clipped Column-Medians from TA Images to
Remove Shading Signature & DC Offsets Before Flat-Fielding
EXAMPLE…..
Note: 7s F160W Target Images in Background Frames
A contemporaneous
reference flat for
the region around
the coronagraphic
hole can (should)
be made from the
TA lamp flats &
backgrounds (S/N
~120, combined),
And used to flatfield the subsequent
coronagraphic
Images.
Coronagraphy with HST/NICMOS
SUMMARY
SMOV3B Program Has Demonstrated Full Return of Capabilities
Coronagraphic Diffracted & Scattered Light Rejection Comparable
to Cycle 7. Should be Fully Restored After Low Scatter-Point
Mapping and Compensation.
Coronagraphic Detectability (Direct and with PSF-Subtraction)
Comparable to Cycle 7, with Increased Sensitivity Due to
QE Improvement @ 77K (DQE~37% in H-band relative to cycle 7).
Final Performance Metrics and Calibration Pending Completion
of Cycle 11 Calibration Test Plan.
Ready to Resume NICMOS Coronagraphic Science
(if any proposals are accepted for HST Cycle 12).
Coronagraphy with HST/NICMOS*
*The Near Infrared Camera & Multi-Object Spectrometer
Extending HST’s UV/Optical Panchromatic Vision
into the Near IR (0.8mm —2.4 mm)
Glenn Schneider
Steward Observatory, University of Arizona (NICMOS/IDT)
http://nicmosis.as.arizona.edu:8000
[email protected]
General Description
“Coronagraphic Focus”
• The f/24 (FDM) and f/45 (detector) image planes are suppose
to be confocal. Because of the “dewar anomoly” they are not.
1200
DIRECT DET
DIRECT FDA
C
O 1000
U
N
T
800
CORON DET
CORON FDA
R
A
T 600
E
/
P 400
I
X
E
200
L
0.3
0.45
0.6
0.75
RADIUS
0.9
1.05
( ARCSECONDS)
1.2
1.35
To achieve “Best
Focus” at the
detector (for “direct”
imaging), a star
image on the FDA
mirror is de-focused,
so light from the 1st
Airy ring scatters off
the edge of the
coronagraphic hole
with much greater
intensity (3x at
1.51.6mm).
General Description
“Coronagraphic Focus”
• The f/24 (FDM) and f/45 (detector) image planes are suppose
to be confocal. Because of the “dewar anomoly” they are not.
1200
DIRECT DET
DIRECT FDA
C
O 1000
U
N
T
800
CORON DET
CORON FDA
R
A
T 600
E
/
P 400
I
X
E
200
L
0.3
To reduce edge
scattering, and
recover image
contrast, the PAM
mirror is moved
(by ~ 2mm) for
coronagraphic
imaging.
As a result the
unocculted PSF is
slightly de-focused.
0.45
0.6
0.75
RADIUS
0.9
1.05
( ARCSECONDS)
1.2
1.35
1.5
Question: Can you comment on the Coronagraphic Focus?
Answer: Follows…
SMOV/3B Coronagraphic Focus Check
Detector
F187N
F160W
F110W
Intermediate
Field Divider Mirror
Coronagraphic Focus Check - F187N
Azimuthal Average Per Pixel Intensity
Focus @ Detector
Focus @ FDA Mirror
Coronagraphic Focus Check - F160W
Azimuthal Average Per Pixel Intensity
Focus @ Detector
Focus @ FDA Mirror
Coronagraphic Focus Check - F110W
Azimuthal Average Per Pixel Intensity
Focus @ Detector
Focus @ FDA Mirror
“Coronagraphic Focus”
The peak of an unocculted
stellar PSF at the
coronagraphic focus is
reduced in intensity by ~ 17%.
This is more than an
acceptable trade given the
reduction by a factor of 3 in
the scattered background near
the coronagrahic hole.
DATA from SMOV/7157
(Cycle 7) and SMOV/8984
(Cycle 11).
PSF CORE FWHM
DIRECT
CORON
INTENSITY
General Description
.
1.934 pix
0.1466”
1.954 pix
0.1481”
I
N
T
E
N
S
I
T
Y
RADIUS (PIXELS)