The

WFIRST CGI

N. Jeremy Kasdin
Princeton University

HabEx Face-to-Face
May 17, 2016
Coronagraph Design Goal:

Find a coronagraph architecture that:

Results in the most contrast . . .

Over the largest area of the image . . .

At the smallest possible inner working

angle . . .

With the most throughput . . .

With no changes to the telescope . . .

By December 2014.
 The Challenge

0.5 25 0
20 −1
15 −2
10
−3
5

y ( λ /D )
−4
0 0
−5
−5
−6
−10
−15 −7

−20 −8

−0.5 −25 −9
−0.5 0 0.5 −20 −10 0 10 20
x ( λ /D )

The AFTA Pupil The AFTA PSF

 Coronagraph selection based on
maturity, robustness, flexibility
HLC PIAACMC
SPC

Pupil Masking (Kasdin, Princeton Pupil Mapping

Image Plane Amplitude & Phase
University) (Guyon, Univ. Arizona)
Mask (Trauger, JPL)

VVC VNC - DAVINCI VNC-PO

Image Plane Visible Nuller – Phase Occulting

Visible Nulller - DAVINCI
Phase Mask (Serabyn, JPL) (Clampin, NASA GSFC)
(Shao, JPL)
04/30/2014 WFIRST-AFTA SDT Interim Report 4
Approach to Recommendation
• Objective: Recommend a primary and backup coronagraph architecture to
focus design and technology development leading to potential new mission
start in F17
• Recommendation by ExEPO and ASO based on inputs from
• SDT: Sets the science requirements
• ACWG: Delivers technical FOMs and technology plans ACWG = representatives of
> Aim for the positive: a consensus product ExEPO, ASO, SDT,
Community
> SDT delivers science FOMs
• TAC: Analysis of technical FOM, TRL readiness TAC:
plans, and risks Alan Boss (Carnegie Mellon)
Joe Pitman (EXSCI)
Steve Ridgway (NOAO)
• ExEPO and ASO recommendation to APD Director Lisa Poyneer (LLNL)
based on: Ben Oppenheimer (AMNH)

• Technical and Programmatic

• Musts (Requirements), Wants (Goals), and Risks
How do we define a
• Distinguish description from evaluation successful outcome?

• APD Director will make the decision

FOM = Figure of Merit
5
Recommendation Criteria:
Defining a Successful Outcome
MUSTS (Requirements): Go/No_Go
1. Science: Does the proposed architecture meet the baseline science drivers?
2. Interfaces: For the baseline science, does the architecture meet telescope and spacecraft requirements of the
observatory as specified by the AFTA project (DCIL1)
3. Technology Readiness Level (TRL) Gates: For baseline science, is there a credible plan to be at TRL5 at the start
of FY17 and at TRL6 at the start of FY19 within available resources?
4. Is the option ready in time for this selection process?

WANTS (Goals): Relative to each other, for those that pass the Musts:
1. Science: Relative strength of science beyond the baseline
2. Technical: Relative technical criteria
- See details
3. Programmatic: Relative cost of plan to meet TRL Gates

RISKS and OPPORTUNITIES

- See details
1 DCL = Dave Content Interface List
6
 Hybrid Lyot Coronagraph
Transmission = 48%

Baseline design for WFIRST/AFTA

Total PSF core throughput = 4.5% From John Trauger, JPL

7
WFIRST Dark Holes with Pointing Jitter & Finite Star
No jitter 0.4 mas jitter 0.8 mas jitter 1.6 mas jitter
0 mas 1.0 mas star 1.0 mas star 1.0 mas star
star

HLC

r = 9 λ/D

SPC

PIAACMC
WFIRST PIAACMC
uses a single DM for
simplicity, so it has a
single-sided dark hole
Coronagraph simulations use validated
wave-optics code
AFTA Pupil Shaped Pupil Focal plane mask

r = 2.5 – 9 λ c/D
65° opening angle

27% mask transmission

Lyot stop

rsp = 0.3 – 0.9 r pup

Total PSF core throughput = 3.7%

Simulations show e.g. robust performance
against jitter
1.6 mas RMS jitter
0.8 mas RMS jitter
0.4 mas RMS jitter
No jitter

728 – 872 nm
Jitter levels shown here are after coronagraph fast tip/tilt
 Simulations show stable high contrast
with AFTA in thermal scenarios
• Proper EFC correction for telescope nominal wavefront (initial DM setting)
• Gen 1 SPC design , 10% bandwidth, l = 550 nm, 3.9 ~12.3 l /D WA, 56 deg opening angle
• Realistic AFTA surface aberration (amplitude +phase), and
• Piston/tip/tilt/focus correction computed only once initially
• The system configuration is held constant throughout the observations

Raw speckle, S(t) Δ Speckle (rel. to nominal): S(t) – S(0)

Typical Mean Contrast: ~5.3e-9 Typical Mean Δ Contrast: ~4.4e-11

11
PIAA - CMC
DM1/FSM

FPA

DM2
PIAA Inverse
mirrors To LOWFS PIAA

Final design deadline extended to 11/4/2013

DM1, DM2 Pupil mapping Apodizer mask Occulting mask Lyot stop Inverse pupil
mapping
Medium ACAD PIAA mirrors Gray scale, filer Phase Transmission, Inverse PIAA
on both DMs wheels? transmission, on binary, fixed? mirrors
filter wheel

12
 CGI Operational Modes
T2 Telescope
Hybrid Lyot Mode
Imaging in 2 simultaneous polarizations, simulated planets are circled in red

T1

COR F1

M3
COL F1

COR F2
M4

Pupil Focal Plane Lyot Field stop Filter IFS/Img

COL F2 Tertiary Collimator mask
mask mask mask wheel Selector
Assembly (TCA)

FSM Coronagraph IMG

Instrument (CGI)
R1 OAP2
DM1 R2
OAP2
R1 FocM FPA
OAP1
R2 R3 FS FS Radiation
DM2 FM R3
Shield Mirror
OAP1 OAP1 OAP2 OAP1 OAP2 PBS

December 8-9, 2015 to LOWFS 13

 CGI Operational Modes
T2 Telescope
Shaped Pupil Spectroscopy Mode
The IFS uses 3 18% bands to produce an R=70 spectra from 600 to 970 nm

T1

COR F1

M3
COL F1 lenslet
pinhole extracted data
SPC images in 3 18% array dispersed
mask cube
COR F2 bands lenslet images
M4

IFS/Img
Pupil Focal Plane Lyot Field stop Filter Selector
COL F2 Tertiary Collimator mask
mask mask mask wheel
Assembly (TCA)

FSM Coronagraph IFS

Instrument (CGI)
FPA
R1 OAP2
DM1 R2
OAP2 IFS
R1 FocM
OAP1
DM2 R2 FM R3 R3 FS FS
OAP1 OAP1 OAP2 OAP1 OAP2

December 8-9, 2015 to LOWFS 14

CGI Operational Modes
T2 Telescope
Shaped Pupil Disk Imaging Mode
Disk Imaging at wavelengths 465 and 890 nm, in 2 simultaneous polarizations

T1

COR F1

M3
COL F1

Image from 2015 Exo-C STDT Final Report

COR F2
M4

IFS/Img
Pupil Focal Plane Lyot Field stop Filter Selector
COL F2 Tertiary Collimator mask
mask mask mask wheel
Assembly (TCA)

FSM Coronagraph IMG

Instrument (CGI)
R1 OAP2
DM1 R2
OAP2
R1 FocM FPA
OAP1
R2 R3 FS FS Radiation
DM2 FM R3
Shield Mirror
OAP1 OAP1 OAP2 OAP1 OAP2

to LOWFS
 The Coronagraph Instrument

Optical Bench

Triangular
Support Frame
Electronics Platform

Imaging Camera Integral Field Spectrograph

1k x 1k EMCCD detector R70 Spectra in IFS 600 to 970 nm
5 arcsec imaging FOV from 430 to 970 nm 0.5 arcsec FOV
4 filters for color photometry Better than Nyquist sampled PSF
Contrast better than 10-8 (raw)
Dark hole from 0.2 to 1 arcsec
Dual polarizations
CGI Optical Layout
Focal Plane DM2
mask Fast Steering Shaped-pupil
Mirror mask
Field Stop
DM1

LOWFS
Color Filter Camera
Wheel Lyot Stop

Imaging
Camera
IFS Camera
IFS Optics
 WFIRST Telescope LoS Jitter and WFE Drift
• Line-of-sight drift and jitter (Cycle 5 model)
– Drift (<2Hz): ~14 milli-arcsec ACS pointing.
– Jitter (>2Hz): < 10 milli-arcsec. Most around 10 Hz, the telescope optics.
with multiple harmonics at each RW speed.
• RW induced WFE Jitter (Cycle 5 model)
– High frequency WFE. Dominant WFE are: astig
(Z5, Z6), coma (Z7, Z8), trefoil (Z9, Z10).
– Impact to coronagraph contrast is small.
LoS vs RW Speed
• WFE drift (Cycle 5 model)
– Mostly thermal induced rigid body motion of
– Slow varying, typically <10 pm/hour.
– Dominant WFE are: focus (Z4), Astigmatism
(Z5, Z6) and coma (Z7, Z8).
– Severely depredate the coronagraph contrast
if left un-corrected.
WFE Drift

1.6 mas

0.4 mas

18
3
OMC Aberration Sensitivities
HLC and SPC WFE sensitivities modeled by J. Krist
• Compared to 2013 ACWG down select, HLC sensitivities are lower,
SPC sensitivities higher (performance trade-off with the addition of Lyot
stop)
• Sensitivity highest to spherical and coma

19
Low Order Wavefront Sensing and Control (LOWFS/C)
SPC mask Focal plane Lyot Field stop Filter
From Telescope mask mask mask wheel

HOWFS/C

HLC
FSM
R1 SPC
R1
OAP1 OAP2 DM1
R3 R3
OAP1 OAP2
Foc M
FPA
DM2 R2 R2 FS FS PBS
OAP1 OAP2 OAP1 OAP2

LOWFS/C
5 mHz

5 mHz
1 kHz

Coronagraph IFS
Instrument (CGI)

• WFIRST LOWFS/C subsystem measures and controls line-of-sight (LoS) jitter and
drift as well as the thermally induced low order wavefront drift
• Differential sensor referenced to coronagraph wavefront control: maintains
wavefront established for high contrast (HOWFS/C)
• Using rejected starlight from occulter which reduces non-common path error
• LOWFS/C telemetry can be used for coronagraph data post-processing
 • Disturbance: WFIRST Cycle 5 CBE
• ACS LoS drift
• LoS jitter @ RW of 10 rev/sec (worst case)
• Small plot shows the zoomed in region

Lab Noise
ACS & Jitter On

FB Loop On
On
FF Loop
Converging

Both FB & FF Loops

Typical Observing Sequence
• The typical scenario involves two stars:
1. A nearby bright star for getting a dark hole (~ 3 Hrs)*
2. The planet host target star (~20 Hrs)* * numbers notional

1
RV planets
Bright star for ~20 deg
acquisition of DH 𝑉 ∼ 6 𝑚𝑎𝑔
Target star and planet
𝑉 ≤ 3 𝑚𝑎𝑔
~ 100 available stars
2

The SIT’s can Angular Differential Imaging Case

help select the
+13 o + roll angle (≡ 0 at min solar angle)
right bright stars! 1
Bright
0o
Star

SA normal
+13 o
2
Solar angle
Target
Star 0o
3 Sun
-13 o
WFIRST CGI “Stanford” Meeting 8 - Nemati - Coronagraph
April 26, 2016 22
Operational Concept
High-Contrast Science with WFIRST

• Characterize Roughly a Dozen Known RV Planets

• Photometrically Discover New Planets Down to Mini-Neptunes
• Image Debris Disks and Exo-Zodi in Two Polarizations
• GO High-Contrast and Narrow Field Science
• Demonstrate technology for future Earth imager
Exoplanet Imaging Survey
• Encompasses 1 year of mission time
• Image previously discovered radial velocity planets in
WFIRST CGI range
• Perform searches for new planets in range from Super-
Earths to Jupiters
• Search for debris disks and characterize down to a few zodi
levels
• Opportunities for general observer science using imager or
IFS with our without the coronagraph
Current WFIRST Preparatory Science project and SITs will
develop detailed DRM simulations to optimize distribution of
mission time and observations.
 1000

Planet Mass (MEarth)

100

10
WFIRST/RV
WFIRST/New
Transit Spec
Sec. Eclipse
1 Direct Imaged

0.01 0.10 1.00 10.00 100.00

Semi-major axis (AU)

Figure credit Eric Nielsen

achievable contrast can be a misleading value, as the contrast
floors of the architectures here typically vary by a factor of
is the collecting area of the telescope, T all (electrons∕photon)
up to ∼100. Instead, we quote in Table 5 a common example
is the optical efficiency of the system, and T point is the fraction
target, 47 UMa c, which has a planet/star contrast of
of light contained within the FWHM of the PSF, from Sec. 4.2.
5.9 × 10−10 , at an angle of 242 mas, and can be detected by
Note that we count only that fraction of the starlight that falls
all three architectures. See Fig. 7 for a graphical simulation
within the FWHM of the image, as a simple approximation that
recognizes that the background speckles have shapes similar to
astrophysical point sources. The portion of the image outside
this boundary is not as likely to be useful in detection as it
would be if the background were flat.
The planet count rate npl is then given by the count rate of the
star times the contrast factor (planet∕star), from above, giving

EQ-TARGET;temp:intralink-;sec4.4;326;631 npl ¼ nstar × CðαÞðelec∕sÞ

in the area enclosed by the FWHM of the PSF.

4.5 Background Count Rate from Zodiacal Light

Fig.Contrast of of
6 Contrast speckles
specklesininthe
the dark holefor
dark hole foreach
each coronagraph, forScience
Fig. 8 imaging yield of yield
Science imaging known of RV planets
known RV planetswith HLC. SolidSolid
with HLC.
coronagraph,
a small for (0.4
residual jitter a small
masresidual
RMS) and jitter (0.4 mas
a large jitter (1.6symbols
RMS)
residual The
symbols aretotal
areRVRVbackground
planets count rate
detectable
planets detectable with from
with
HLC HLC inthe sky
ain15% is
a 15% theband
band sum of inten-
centered
masand a large
RMS), residual
at the designjitter (1.6 mas
wavelength of RMS),
550 nm. at For
the other
design wave-centered sities
at 565 nm, from
in less
at 565 speckles
nm, than a and
in less day
thanlocal
each, zodi
a day plus
in aeach,
singlein exo-zodi,
a singlein this
polarization, withsection.
a
lengths, the planet–star
wavelength of 550 nm.angleForis replaced by one that is
other wavelengths, signal
(Instrength
wavelengthpolarization,
the thiswith greater
study, we than
a signal ignore thescattered
strength worst-case
greaterand thanfloor (long-dash
diffracted
the worst-case line,floor
contributions
Fig. 7 planet–star
Simulated
scaled, WFIRST-AFTA
as suggested
angleby is the coronagraph
axis
by at theimage
top.isofwavelength
the star 47 1.6 mas RMS
(long-dash pointing jitter,
RMSand a postprocessing
jitter, and a factor of 1/10).
λ∕D
replaced one that fromline, 1.6
companion masstars.) pointing postprocessing
Ursa Majoris, showing two directly detected planets (from Ref. 1), The open symbols are for the additional detections possible with
scaled, as suggested by the λ⁄D axis at the top. factor of 1/10).
For aThe open symbols
representative are we
value, for thedefine additional
one detections
zodi to have a
for HLC, and with a field of view radius of 0.5 arc sec. A PSF reference the best-case floor (0.4 mas jitter, 1/30 factor). Typical relative uncer-
jitter amplitudes and frequencies, and Ref. 28 for a discussion ofpossible
has been subtracted, improving the raw contrast by a factor of 10. with
brightness
tainties the best-case
in contrast floor (0.4
of and separation mas
(Sec. 2.4)jitter, 1/30 factor).
are indicated.
the speckle simulations. We assume that the intrinsic jitter level Traub, et al., JATIS, 2016
 Key Coronagraph Technology
Coronagraph Design, Masks and Hardware (varies by type)
– Shaped Pupil (SP), SPLC, Hybrid Lyot, APLC, Vector Vortex, PIAA, PIAA/CMC, 4QPM
– Large central obstruction, spiders, segmented mirrors

Wavefront Estimation and Control (common to all)

Probes and Field estimation, Control Algorithms (EFC & Stroke Minimization),
Deformable Mirrors, Broadband control (with and without IFS), Low-Order Wavefront
Sensing and Control (LOWFSC)

Data Analysis and Planet Identification

PFS Subtraction (LOCI, ADI, KLIP), IFS data cube, Spectral Characterization

Mission Modeling and DRMs

Engineering and Instrumentation

Optical design, polarization, IFS, calibration and test, operations

Error Analysis
Polarization, finite stellar size, stability, thermal bending (low-order aberrations)
 CGI as a Technology Demonstrator

Coronagraph Design, Masks and Hardware (varies by type)

– Shaped Pupil (SP), SPLC, Hybrid Lyot, APLC, Vector Vortex, PIAA, PIAA/CMC, 4QPM
– Large central obstruction, spiders, segmented mirrors

Wavefront Estimation and Control (common to all)

Probes and Field estimation, Control Algorithms (EFC & Stroke Minimization),
Deformable Mirrors, Broadband control (with and without IFS), Low-Order Wavefront
Sensing and Control (LOWFSC)

Data Analysis and Planet Identification

PFS Subtraction (LOCI, ADI, KLIP), IFS data cube, Spectral Characterization

Mission Modeling and DRMs

Engineering and Instrumentation

Optical design, polarization, IFS, calibration and test, operations

Error Analysis
Polarization, finite stellar size, stability, thermal bending (low-order aberrations)
Starshade Rendezvous

A study is underway to identify needed modifications to make WFIRST “starshade ready”.

Exo-S study provided example of starshade architecture that could fly with
WFIRST, expanding the scope of technology demonstration and opening the
opportunity for imaging of Earth-size planets.
Technology Development

Ilya Proberezhskiy, JPL