Transcript Document

Adaptive Optics in the VLT and ELT era
Beyond Basic AO
François Wildi
Observatoire de Genève
Page 1
Adaptive Optics wavefront errors reminder
• The residual wavefront error is the quality criterion in AO
• The wavefront error depends on:
– The number of degrees do freedom (i.e. +/- nb of
actuators) of the deformable mirror.
– The lag (delay) in the control system
– The noise in the wavefront sensor which depends on
the guide star magnitude
– The size of the field of view
– Side effects like WFS non-ideality, NCPA, disturbances
like vibrations
Dependence of Strehl on l and number of DM
degrees of freedom
 
5/3
S  exp  2  exp  0.28 d / r0  


r0 l   r0 0.5  m l / 0.5  m 
6 /5
5/3
2



d
 0.5  m 

S  exp 0.28 

 

l
 r0 0.5  m 


Deformable mirror fitting error only
• Assume bright
natural guide
star
• No meas’t error
or iso-planatism
or bandwidth
error
Reminder #1: Dependence of Strehl on l and
number of DM degrees of freedom (fitting)
• Assume bright
natural guide
star
Decreasing fitting error
Deformable mirror fitting error only
• No meas’t error
or iso-planatism
or bandwidth
error
Basics of wavefront sensing
• Measure phase by measuring intensity variations
• Difference between various wavefront sensor schemes
is the way in which phase differences are turned into
intensity differences
• General box diagram:
Guide
star
Turbulence
Telescope
Wavefront sensor
Optics
Detector
Reconstructor
Computer
Transforms aberrations into
intensity variations
Types of wavefront sensors
• “Direct” in pupil plane: split pupil up into subapertures
in some way, then use intensity in each subaperture to
deduce phase of wavefront. REAL TIME
– Slope sensing: Shack-Hartmann, pyramid sensing
– Curvature sensing
• “Indirect” in focal plane: wavefront properties are
deduced from whole-aperture intensity measurements
made at or near the focal plane. Iterative methods take a lot of time.
– Image sharpening, multi-dither
– Phase diversity
Shack-Hartmann wavefront sensor
concept - measure subaperture tilts
f
CCD
Pupil plane
Image plane
CCD
WFS implementation
• Compact
• Time-invariant
How to reconstruct wavefront from
measurements of local “tilt”
Effect of guide star magnitude
(measurement error)
Because of the photons statistics, some noise is
associated with the read-out of the Shack-Hartmann
spots intensities
 S2 H
 6.3 

 SNR 
2
2

 6.3  
2
S  exp   S  H   exp   
 

 SNR 
1
SNR 
N photons
Assumes no fitting error or other error terms
Effect of guide star magnitude
(measurement error)
Assumes no fitting error or other error terms
bright star
Decreaing
measurement error
dim star
Reminder #3: Strehl vs l and guide star
angular separation (anisoplanatism)
    5/3 
2
  exp     
S  exp   iso
   0  
r0
 0   l 6 /5
h
5/3
2
 


 0.5  m 
S  exp   

 

l
   0 (0.5  m) 

Reminder #3: Strehl vs l and guide star
angular separation (anisoplanatism)
Anisoplanatism side effect:
• Because correction quality falls off rapidly looking
sideways from the guide star AND because faint stars
cannot be used as guide stars,
Only a very small part of the
sky is accessible to natural
guide star AO systems!
Sky coverage accounting for
guide star densities
LGS coverage ~80 %
Tip/tilt sensor magnitude limit
Hartmann sensor magnitude limit
Galactic
latitude
NGS coverage 0.1 %
Isoplanatic angle 0
Isokinetic angle k
(Temporary) conclusion on isoplanatism:
• With 0.1% sky coverage, classical AO is of limited use
for general astronomy.
• This is perticularly true for extra-galactic astronomy,
where the science object is diffuse, often faint and
cannot be used for wavefront sensing.
AO’s great divide
High precision
ExAO
Wide field
LTAO (high coverage)
GLAO
MCAO
MOAO
ExAO in a nutshell
• Like classical AO but more of the same
• The wavefront error minimized on axis
– Large number of degrees do freedom (i.e. +/- nb of
actuators) of the deformable mirror.
– Minimal lag (delay) in the control system
– Low noise in the wavefront sensor: Bright guide star
– “No” field of view
– WFS non-ideality fought with spatial filter, NCPA
measured and corrected, disturbances like vibrations
countered with advanced signal processing
High contrast imaging
Diff. Pol.
Highest contrast observations
require multiple correction
stages to correct for
1. Atmospheric turbulence
visible
coronagraph
XAO
2. Diffraction Pattern
infrared
coronagraph
3. Quasi-static instrumental
aberrations
XAO, S~90%
SDI
Coronagraph
IFS
Differential Methods
NCPA compensation
Use of phase diversity for NCPA correction on Vis. path
-Strong improvement of
bench internal SR (45 ->
85 in Vis)
- various optimisations
still to be performed
1
NCPA compensation for IR path
Ghosts
No
compensation
NCPA compensation
320 modes estimated, 220
corrected
Implementation
CPI
De-rotator HWP2
Focus 1
HWP1
ITTM
PTTM
Polar Cal
Focus 2
DM
Focus 4
NIR ADC
VIS ADC
DTTS
VIS corono
Focus 3
ZIMPOL
WFS
NIR corono
IRDIS
DTTP
IFS
Sky coverage and Wide field in a nutshell
• To circumvent the sky coverage problem, several ways
have been devised and are actively pursued:
1. Laser Tomography Adaptive Optics (LTAO)
Laser guide stars are used to probe the atmosphere and
project it in the science object direction
2. Ground Layer Adaptive Optics (GLAO)
Laser guide stars are used to probe the atmosphere but only
the ground layer is corrected
3. Multi-Conjugate Adaptive Optics (MCAO)
Laser guide stars are used to probe the atmosphere and
turbulence is projected and corrected in several layers
4. Multi Object Adaptive Optics (MOAO)
Laser guide stars are used to probe the atmosphere and
turbulence is projected in several directions. Each direction
has one (or several DM’s)
LASER TOMOGRAPHY AO
In LTAO, the atmosphere is probed by multiple
Wave Front Sensors to form a model of the
atmosphere. This model is used to compute
the wavefront distorsion in a perticular
direction and therefore calculate a correction
in that direction.
It allows a good correction in a direction that
lacks a good natural guide star at the expense
of system complexity
Field is not increased!
Proper use of the system
requires several wavefront
sensors to perform
Tomography
Altitude Layer
(phase
aberration = +)
Ground Layer = Pupil
(phase aberration = O)
Tomography = Stereoscopy
WFS#1
WFS#2
WFS Set-up and LTAO reconstruction
Turb. Layers
#2
Atmosphere
UP
#1
Telescope
DM corrects
#1 + #2 in red
direction
WFS
GROUND LAYER AO
In GLAO, the atmosphere is probed by multiple
Wave Front Sensors to form a model of the
atmosphere. Only the ground layer is
extracted form the model and used to feed
back a correction mirror conjugated to the
ground.
It allows a correction of the atmospheric
wavefront error that happens in the common
path of all objects at the expense of system
complexity
Field is very large but performance is limited
Performance expected from GLAO (Gemini)
WFS Set-up and GLAO reconstruction
Turb. Layers
#2
#1
Telescope
WFS
DM corrects #1
Atmosphere
UP
LASER GUIDE STARS VS NATURAL GUIDE STARS
Tomography can also be performed with
natural guide stars BUT:
• Requires planning the NGS for each
observation
• Quality is not constant due to NGS geometry
and flux distribution
• Requires movable wave front sensors
Solution unanimously discarded today
Multi Conjugate Adaptive Optics
• To increase the isoplanatic patch, the idea is to design
an adaptive optical system with several deformable
mirrors (DM), each correcting for one of the turbulent
layer
Each DM is located at an image of the corresponding
layer in the optical system. (By definition, the layer and
the DM are called conjugated by the optical system).
What is multiconjugate? Case without
Turbulence Layers
Deformable mirror
What is multiconjugate? Case with it
Deformable mirrors
Turbulence Layers
Multiconjugate AO Set-up
Turb. Layers
#2
Atmosphere
UP
#1
Telescope
WFS
DM#2 DM#1
Effectiveness of MCAO: no correction
Numerical simulations:
• 5 Natural guide stars
• 5 Wavefront sensors
• 2 mirrors
• 8 turbulence layers
• MK turbulence profile
• Field of view ~ 1.2’
• H band
Effectiveness of MCAO: classical AO
Numerical simulations:
• 5 Natural guide stars
• 5 Wavefront sensors
• 2 mirrors
• 8 turbulence layers
• MK turbulence profile
• Field of view ~ 1.2’
• H band
Effectiveness of MCAO: MCAO proper
Numerical simulations:
• 5 Natural guide stars
• 5 Wavefront sensors
• 2 mirrors
• 8 turbulence layers
• MK turbulence profile
• Field of view ~ 1.2’
• H band
MCAO Performance Summary
Early NGS results, MK Profile
No AO
Classical AO
1 DM / 1 NGS
320 stars / K band / 0.7’’ seeing
165’’
MCAO
2 DMs / 5 NGS
Stars magnified for clarity
The reality…: GEMINI MCAO Module
LGS source
Science ADC simulator
NGS source
simulator
DMs
shutters
TTM
Beamsplitter
NGS
WFS
NGS
ADC
Diagnostic WFS
LGS WFS
LGS zoom corrector
Example of MCAO Performance
•
•
•
•
•
•
•
13x13 actuators system
K Band
5 LGSs in X of 1 arcmin on a
side
Cerro Pachon turbulence
profile
200 PDE/sub/ms for H.Order
WFS
Four R=18 TT GS 30” off axis
(MCAO)
One R=18 TT GS on axis(AO)
MCAO Performance
1
Classical LGS AO
MCAO
Strehl
1
0
Surface plots of Strehl ratio over a 1.5 arc min FoV.
13x13 actuator system, K band, CP turbulence.
Average Strehl (triangles)
• Robustness
• Sensitivity to noise1 is fairly better than with AO
Prop noise AO / Prop noise MCAO  sqrt( NGS )
• Predictive algorithms possible ?
4
.5
+
+
+
2
+
+
+
0
Profile number
Strehl St. dev across FoV % (+)
Other nice features of MCAO
Generalized Fitting
(Finite number of DMs)
Geometry of the problem
dact
Generalized Anisoplanatism
(Finite number of Guide Star)
Additional error terms are
necessary to represent laser guide
star MCAO. Tomography error
arises from the finite number and
placement of guide stars on the
sky. Generalized anisoplanatism
error results from the correction
of the continuous atmosphere at
only a finite number of conjugate
layer
altitudes.
Generalized Fitting
(Finite number of DMs)
Error [rd2]  (.h)5/3
Design Criteria e.g. Error balanced  hmax(,dact)
DM Spacing = 2 x hmax
dact
FoV [arcmin]
hmax [m]
NDM/GS
0.5
1
3000
3
0.2
1
1200
5-6
0.2
10
120
50
Generalized Anisoplanatism
(Finite number of Guide Star)
• Turbulence altitude estimation error
• OK toward GS, but error in between GS: Strehl “dips”
100”
FoVDM
= 70”
• Maximum FoV depends upon
pitch.
• Example for 7x7 system
Generalized Anisoplanatism goes down
with increasing apertures
2D info only
3D info
3D info
2D info only
Aperture
MCAO Pros and Cons
PROS:
• Enlarged Field of View
– PSF variability problem drastically reduced
• Cone-effect solved
• Gain in SNR (less sensitive to noise, predictive
algorithms)
• Marginally enlarged Sky Coverage (LGS systems)
CONS
• Complexity: Multiple Guide stars and DMs
• Other limitations: Generalized Fitting, anisoplanatism,
48
aliasing
MULTI OBJECTS ADAPTIVE
OPTICS
• In certain case, the user does not want to (or need to)
have a fully corrected image. He/she might be satisfied
with having only specific locations (i.e.) objects
corrected in the field.
• An AO system designed to provide this kind of
correction is called a Multi Objects Adaptive Optics
system
• MOAO are the systems of choice to feed spectrographs
and Integral Field Units in the ELT era.
–MOAO
• Up to 20 IFUs each with a DM
• 8-9 LGS
• 3-5 TTS
MOAO for TiPi (TMT)
MEMSDMs
Flat 3-axis
steering mirrors
OAPs
Tiled
MOAO
focalplane
4 of 16 d-IFU
spectrograph
units
Key Design Points for AO
Key points:
• 30x30 piezo DM placed at M6, providing partial
turbulence compensation over the 5’ field.
• All LGS picked off by a dichroic and directed back to
fixed LGS WFS behind M7. Dichroic moves to
accommodate variable LGS range.
• The OSM is used to select TT NGS and PSF reference
targets.
• MEMS devices placed downstream of the OSM to
provide independent compensation for each object:
16 science targets, 3 TT NGS, PSF reference targets.
LASER GUIDE STARS
Laser guide star AO needs to use a faint
tip-tilt star to stabilize laser spot on sky
from A. Tokovinin
Effective isoplanatic angle for image
motion: “isokinetic angle”
• Image motion is due to low order modes of turbulence
– Measurement is integrated over whole telescope
aperture, so only modes with the largest
wavelengths contribute (others are averaged out)
• Low order modes change more slowly in both time and
in angle on the sky
• “Isokinetic angle”
– Analogue of isoplanatic angle, but for tip-tilt only
– Typical values in infrared: of order 1 arc min
Sky coverage is determined by
distribution of (faint) tip-tilt stars
• Keck: >18th magnitude
1
Galactic latitude = 90°
Galactic latitude = 30°
271 degrees of freedom
5 W cw laser
0
From Keck AO book
LGS Related Problems: Null modes
• Tilt Anisoplanatism : Low order modes > Tip-Tilt at
altitude
–  Dynamic Plate Scale changes
• Within these modes, 5 “Null Modes” not seen by LGS
(Tilt indetermination problem)
 Need 3 well spread NGSs to control these modes
• Detailed Sky Coverage calculations (null modes
modal control, stellar statistics) lead to
approximately 15% at GP and 80% at b=30o
• Additional error terms are necessary to represent laser
guide star MCAO. Tomography error arises
• from the finite number and placement of guide stars on
the sky. Generalized anisoplanatism error results from
the correction of the continuous atmosphere at only a
finite number of conjugate layer altitudes
LGS WFS Subsystem needs constant refocussing!
• Trombone design accomodates LGS altitudes between
85-210 km (Zenith to 65 degrees)
• Astigmatism corrector present / Will study Coma
corrector
TMT.IAO.PRE.06.03
62
TMT MIRES (proposal)
Concept
Overview
LGS
trombone
system
TMT.INS.PRE.06.02
63
3. NGS WFS
• Radial+Linear stages with encoders
offer flexile design with min.
vignetting
• 6 probe arms operating in
“Meatlocker” just before focal plane
• 2x2 lenslets
EEV CCD60
• 6” FOV - 60x60 0.1” pix
Flamingos2 OIWFS
TMT.IAO.PRE.06.03
64
Issues for designer of AO systems
• Performance goals:
– Sky coverage fraction, observing wavelength, degree
of compensation needed for science program
• Parameters of the observatory:
– Turbulence characteristics (mean and variability),
telescope and instrument optical errors, availability of
laser guide stars
• AO parameters chosen in the design phase:
– Number of actuators, wavefront sensor type and
sample rate, servo bandwidth, laser characteristics
Effects of laser guide star on overall AO
error budget
• The good news:
– Laser is brighter than your average natural guide star
» Reduces measurement error
– Can point it right at your target
» Reduces anisoplanatism
• The bad news:
– Still have tilt anisoplanatism
– New: focus anisoplanatism
– Laser spot larger than NGS
tilt2 = (  / tilt )5/3
FA2 = ( D / d0 )5/3
meas2 ~ ( b / SNR )2