Transcript Turbulence Layers - Facultad de Ciencias

3D Spectroscopy
Francisco Müller Sánchez
Instituto de Astrofísica de Canarias
La Laguna, España
 Introductory Review and Observational Techniques
 Science motivation for 3D Spectroscopy
 Instrumentation
 Preparation of Observations and Principles of Data Reduction
 Data Analysis
Examples of SINFONI-AO: prototypical merger NGC6240
velocity
flux
1”
stars
molecular gas
ionised gas
Analysis of data cubes
Introductory Review and Observational techniques
Beckers 1993:
ARA&A 31, 13
Hardy 1998:
Adaptive Optics for Astronomical Telescopes
Antichi 2009:
ApJ, 695, 1042
Kissler-Patig 2005: Science perspectives for 3D spectroscopy
also
Sterne & Weltraum articles 1994 (Hippler, Kasper, Davies, Ragazzoni)




Classical observational techniques
Strengths of 3D Data
Concepts of Adaptive Optics
Instrument techniques used to achieve 3D Spectroscopy
Ancient Mayan Photometry
- An eclipse table that
predicts times when
eclipses may occur.
- A Venus table that
predicts the times when
Venus appears as
morning star and the
other apparitions of the
planet.
- A Mars table that
records the times when
Mars goes into
retrograde motion. A
second Mars table that
tracks the planet's motion
along the ecliptic has
recently been identified.
Photometry
The venerable photographic plate and its more recent version, the CCD,
provide objective information in two dimensions concerning the
brightness, I(x,y), of an extended object or area of sky.
Spectroscopy
Spectroscopy
Long-slit spectroscopy enables us to split the light that reaches us not
just from a point but also from an entire line of points into its constituent
colours, thereby providing us with information in two dimensions position along the slit, x, and colour, lambda:I(x,lambda)
Why not combine them?
3D spectroscopy attempts to get closer to the
fundamental goal of astronomical observing
techniques, which is to record the direction,
wavelength, polarization state and arrival time
for every incoming photon over the largest
field of view. In fact using 3D spectroscopy,
the wavelength and the incoming direction in a
2D field of view are recorded in a (x,y,λ) data
cube, in contrast with standard techniques
which either do imaging over a 2D field, or
spectroscopy along a 1D slit.
3D spectroscopy yields datacubes
Color
Scanning Spectrophotometers (Fabry-Perot interferometer)
The FPI can be used to obtain
monochromatic images over a full twodimensional field of view with spectral
resolutions comparable to those of
grating spectrographs. In a Fabry-Pérot
the distance between the plates can be
tuned in order to change the
wavelengths at which transmission
peaks occur in the interferometer.
Imaging Fourier Transform Spectroscopy
Imaging Spectroscopy
Scanning Long-Slit Spectroscopy
Energy-Resolving Detectors
Tantalum superconducting tunnel junctions Peacock et al. 1998
Integral-Field Units
Concept of integral-field spectroscopy
Don’t confuse IFS with MOS (Multi Object Spectroscopy)
LUCIFER at the LBT
Three ways of doing IFS
Lenslets (TIGER Approach)
Example of lenslet IFU: SAURON @ WHT
Fibers (ARGUS approach)
Example of fibers IFU: INTEGRAL @ WHT
Slicers
Example of Image slicer IFU: SINFONI @ VLT
SINFONI - made @ MPE
Strengths of 3D data 1
- No slit losses: high system efficiency
- Less time consuming
- More accurate radial velocity determination
- Background estimate can be obtained
simultaneously
- Kinematics of crowded regions
- It doesn’t suffer from changes of several
exposures
Strengths of 3D data 2. Einstein’s cross section
Atmospheric Turbulence
quantified using the structure function
D( r ) 
 f (r' )  f (r'r )2
for Kolmogorov statistics, the
refractive index structure function is
Dn ( r )  C N2 r 2 / 3
for a wavefront propagating through the atmosphere,
the phase structure function is

 2 
5/ 3
D ( r )  2.91  sec  r  CN2 dh
0
  
2
van Karman model includes inner (~1cm) & outer (~30m) scales
Atmospheric Turbulence
CN2 is refractive index
structure constant.
The integral of CN2 is
Fried’s parameter
r0    C dh 
 0


3 / 5
2
N
and variance of wavefront
aberrations is just
D
  1.030 
 r0 
5/ 3
2
Turbulence limits the
resolution of a telescope
to λ/r0 instead of λ/D.
CN2 at Mt Graham
(LBT site)
Everything depends on CN2
coherence length
r0
where
coherence timescale
r0  0.1856 / 5 sec     CN2 dh 
 0

3/ 5

3 / 5
 0  0.314r0 Vwind
Vwind   CN2 v 5 / 3dh
 0

where


0
CN2 dh

3/ 5
assumes Taylor’s frozen flow hypothesis
isoplanatic angle
 0  0.314r0 / H
H  sec    C h dh
 0

where
2
N
5/ 3


0
C dh

2
N
3/ 5
Impact of a Perturbed Wavefront
parallel light
rays can be
focussed

Point
focus
how well spatial frequencies are
transferred through the optical system
light rays
affected by
turbulence
 blur
resulting shape of a point source
Modal Decomposition
Most common & simplest for a circular aperture are Zernike modes.
For an annular aperture, Karhunen-Loève modes are better.
coma & trefoil
A simple adaptive optics system
open & closed loop images
Neptune (Keck, NGS)
star (Calar Alto, LGS)
Shack Hartmann Sensor
(developed in 1900 by J.Hartmann)
Measures first derivative of wavefront (gradients)
Displacement of spots is proportional to the wavefront tilt
Many algorithms possible for centroiding
Easy to extend to very high order systems
Divides pupil into subapertures
Shack Hartmann Sensor
Piezo Actuator Mirrors
incoming wavefront will be flat
when it reflects off the mirror
349 actuator DM
thin flexible (glass) mirror
reference block
piezo actuators which
contract & lengthen
when voltages are
applied
wiring on back side
Curvature Sensor
(developed in 1994 by F.Roddier)
A few things to bear in mind
amplitude of aberration
- AO works better at longer wavelengths (dependence of r0 on λ6/5)
e.g. consider a phase change of 250nm with respect to 500nm optical light
and 2.2μm near infrared light. So at longer wavelengths, coherence length
is greater & timescales are longer
- One can measure in optical & correct in infrared (absolute phase change is same)
- AO systems have to run fast (bandwidth ~1/10 of the frame rate)
prediction would be great…
time
Residual Wavefront Variance & Strehl Ratio
coherence length
 2fitting ~ 0.2944 j 
3 2
D r0 5 / 3
for large j (number of Zernike modes)
coherence timescale
5/ 3
2
 timedelay
   0 
isoplanatic angle

total wavefront variance
2
2
2
2
 total
  2fitting   angle
  timedelay
  noise
 ...
Strehl ratio
SR ~ exp   2 
2
angle
  0 
5/ 3
ratio of peak intensity to that
for a perfect optical system
Sodium & Rayleigh Laser Guide Stars
sky coverage few % with NGS but ~50% with LGS
(most coverage in galactic plane; almost none at galactic pole)
Keck
MMT
VLT
Starfire Optical Range, Calar Alto, Lick, MMT, Keck, VLT,
Subaru, Gemini North, WHT, Palomar 200”, Mt Wilson 100”,
(LBT, Gemini South)
A few issues with Laser Guide Stars
1. laser technology
2. elongation of spot due to finite thickness of layer
3. variations in height of sodium layer
4. need for tip-tilt star
on-axis LGS spot
off-axis LGS spot
height (km)
sodium density
time (min)
MultiConjugate Adaptive Optics
MAD strehl maps
reference stars
1 star & 1 DM
high turbulence layer
low turbulence layer
telescope
3 stars & 2 DMs
DM2
DM1
WFSs
one wavefront
sensor per star
MultiConjugate Adaptive Optics
Classical MCAO needs
multiple guide stars
(e.g. Gemini South MCAO
needs 5 LGS & 3 NGS).
reference stars
high turbulence layer
This is computationally complex
Instead, one can use the
layer oriented approach,
with LGS or NGS.
LINC-NIRVANA on the LBT
uses pyramid sensors to
co-add the light from many
faint stars on the detector;
but note that the strehl ratio
is expected to be limited &
vary a bit over the field
low turbulence layer
telescope
DM2
DM1
WFS1
WFS1
one wavefront sensor
per deformable mirror
Examples of LGS-AO: interacting galaxies IRAS 09061-1248
K-band image of these interacting galaxies
shows the vast amount more detail that
LGS-AO can reveal
UKIRT (archive)
NACO-LGS/VLT
Examples of LGS-AO: prototypical merger NGC6240
2µm continuum
1”
Komossa et al. 2003
Tecza et al. 2000
Examples of LGS-AO: prototypical merger NGC6240
velocity
flux
1”
stars
molecular gas
ionised gas
Examples of LGS-AO: high redshift galaxies
Future perspectives: FRIDA @ GTC
Future perspectives: SERPIL @ LBT
Multiple IFS: KMOS @ VLT
Multiple IFS: KMOS @ VLT
IFS @ ELT
Outlook for tomorrow’s lecture
 Science perspectives for IFS
 Galactic astronomy
 The Galactic Center
 Nearby AGN
 Quasars and high-z galaxies
Bimorph Mirrors
2 layer piezo
ceramic which
bends when a
voltage is
applied
incoming wavefront will be flat
when it reflects off the mirror
thin glass
mirror
continuous
electrode
control electrodes
bimorph mirror for Gemini,
showing the zones
Realistic Expectations
Extreme AO (e.g. “planet finders”) aims for >90% strehl at K… but with bright stars
AGN are not particularly bright (fainter than typical limit of R~15mag), and tend to
be fuzzy with a relatively bright background.
Off-axis correction is usually not an option.
LGS performance can vary from 0.1” resolution to ~20% Strehl at K.
One can do much better than the seeing limit, but don’t expect perfect
performance every time; and beware of spatial & temporal variations
600nm
2.2µm
Circinus
Galaxy
5”
no bright point source for AO
reference; and bright background.
5”
with an IR-WFS (i.e. NACO)
Multiple Layers of Turbulence
with 2 turbulent layers,
on- and off-axis
wavefronts are
different
Turbulence Layers
adapted from Rigaut 2000
Multiple Layers of Turbulence
with 2 turbulent layers,
on- and off-axis
wavefronts are
different
and cannot be
corrected with a
single DM
Turbulence Layers
Deformable mirror
adapted from Rigaut 2000
Multiple Layers of Turbulence
with 2 turbulent layers,
on- and off-axis
wavefronts are
different
and cannot be
corrected with a
single DM
Turbulence Layers
Deformable mirrors
but they can be
corrected with
multi-conjugate
DMs
adapted from Rigaut 2000