HARMONI: A single field, wide
band, integral-field spectrograph
for the E-ELT
Matthias Tecza
for the HARMONI team, including
Niranjan Thatte (PI), Fraser Clarke,
Roland Bacon, Santiago Arribas,
Evencio Mediavilla, Gary Rae, Roger Davies
HARMONI: A single field, wide
band, integral-field spectrograph
for the E-ELT
HARMONI is a proposed optical-NIR,
adaptive optics assisted, integral field
spectrograph designed to exploit the earlylight capabilities of the European ELT
Phase A study: March ‘08 — Dec ‘09
E-ELT instrument road map
High-z Kinematics
High-z LAE
Stellar Outflows
Sandro D’Odorico, Marseille Nov’06
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
INSTRUMENT
OBS.
MODES
Rhoads et al. 2005
FOCUS /
AO
Exoplanets
(Coronagraphy)
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
WAVE
RANGE
(µm)
FIELD
Genzel et al. 06,
Förster Schreiber et al. 06
PIXEL
SIZE
Morse, Davidson
(mas)
/d
PROMINENT
SCIECE
CASES
REF
STUDY
Chauvin et al. 2004
DL, NIR Imager
imaging
Nasm./LTAO ,
MCAO
0.9-2.5
>30”
4
wide, n.
bands
~ all
ONIRICA
@ OWL
Narrow Field
Spectrograph
spectroscopy
Nasm./SCAO
, LTAO
0.6-2.5
1”/
10”:
20 / 50
3000,
20000:
~ all
Not
studied
High Resolution
Vis Spectrograph
spectroscopy
Coude/ GLAO
0.4 -0.8
Point
source
=
150000
C2, C7
CODEX
•
Proven instrument concept delivering high quality science
S3, S9
in deep
imaging
surveys (e.g.
JWST),
which>=is
ELT
NIR MOS
Spectroscopy
Grav. Inv./
0.8-2.5
5’ only
30 - possible
50
3000,with anC4,
C10
MOAO
10000:
• Narrowmultiplex.20
field-of-view
matched to early AO capabilities
— nearlimited Grav.
over
field >30” 10 - 30
NIR MOSdiffraction
,DL
Spectroscopy
Inv.aorsmall
0.8-2.5
3000 ,
G4, G9
multiplex 20
Nas/MCAO
20000
• Single object mode rather than survey mode (à la MUSE)
• Imager
Oxford imaging
pre-studyNas
oforan
- Sep ‘07)
MIR
IF/ instrument
Mar-20
30”concept
6 - 20 (Apr
w-n bands,
S3, S9, S5,
Planetary Imager
imaging,
Nasm/ EXAO 0.6-1.75
~2” V
>=
>15
•
Early
spectroscopic
follow
up
faint
sources
discovered
Spectrograph
spectroscopy
~4” H
Nyquist
–
(+limited
£70k
award
spectroscopy
HARMONI, Matthias Tecza
SCAO or
from
STFC
LTAO
EPICS
WFSPEC,
MOMSI
MOMSI
MIDIR
G9, C10
RAS, 9th May 2008
3
Single-Field Wide-Band Spectrograph (SFWBS)
Primary mode
Optional modes
- Integral Field
- Medium Spectral Resolution
- NIR
- GLAO & LTAO/MCAO
- Visible-red wave bands
- High Spectral Resolution
SPIFFI/SINFONI
-
SWIFT
NIR: 1.0-2.5µm
≈ 2000 spectra
R ≈ 2000 - 4000
0.25” - 0.025”
8”x8” - 0.8”x0.8”
VLT/SINFONI AO module
HARMONI, Matthias Tecza
-
RAS, 9th May 2008
I/z: 0.65-1.0µm
≈ 4000 spectra
R ≈ 4000
0.235” - 0.08”
22”x10” - 7.5”x3.5”
Palomar 5m / PALAO
4
Principle of the Image Slicer
(used in SINFONI, GNIRS, NIFS)
2nd mirror stack
preserves pupil of input beam
input from telescope
Light loss at
boundary
slicer
stack
located in telescope
focal plane
output to spectrometer
HARMONI, Matthias Tecza
RAS, 9th May 2008
6
Image Slicer with de-magnification
slicer entrance pupil
from pre-optics
re-imaged
telescope pupil
flat slicer mirrors
flat pupil
mirrors
f=d
d
HARMONI, Matthias Tecza
re-imaged slicer stack
lens mosaic
• creates tele-centric exit slit
• de-magnifies slicer stack
RAS, 9th May 2008
7
SWIFT slicer
•
•
•
•
•
•
•
≈4000 spectra
44 slices, 91 pixel long
0.47mm slice width
≈4:1 de-magnification
Twin exit slits, ≈115mm length
Flat slice and pupil mirrors
Lenses for de-magnification
HARMONI, Matthias Tecza
RAS, 9th May 2008
8
Study-slicer
flat
pupil
mirrors
re-imaging mirrors
flat
slicing
mirrors
from telescope/AO
HARMONI, Matthias Tecza
Top view
RAS, 9th May 2008
9
Study-slicer
Side view
flat
pupil
mirrors
reimaging
mirrors
flat
slicing
mirrors
from telescope/AO
HARMONI, Matthias Tecza
RAS, 9th May 2008
10
Single slicer
• ≈4,000 spaxels
• 44 slices, 88 pixel long
• 1.3mm slice width
• 10:1 de-magnification
• Exit slit length ≈260mm
• Flat slice and pupil mirrors
• Mirrors for de-magnification
(cf lenses in SWIFT)
HARMONI, Matthias Tecza
RAS, 9th May 2008
11
Two slicers: 8000 spectra
HARMONI, Matthias Tecza
RAS, 9th May 2008
12
HARMONI, Matthias Tecza
470mm
• Four exit slits
• 4:1 aspect ratio on slicer due
to 2:1 anamorphic pre-optics
• 2:1 aspect ratio on sky
– 176 x 88 pixels
• Maximum spaxel scale of
50mas
– 8.8” x 4.4” FoV
• Smaller spaxel scales (eg.
5mas) through scale changing
pre-optics
230mm
Full slicer
RAS, 9th May 2008
13
Study-slicer design advantages
• No field splitting in pre-optics
– No re-imaging like MUSE
– High throughput
• Flat slicer and pupil mirrors
– Aberrations are equal for all slices
• All mirror design
– Fully achromatic for wide waveband coverage
• Very efficient use of detector real estate
– ≈95% spectrum packing factor
HARMONI, Matthias Tecza
RAS, 9th May 2008
14
Conceptual spectrograph design
• f/6 Collimator
–
–
–
–
1500mm focal length
120mm x 240mm beam
3 mirror design, 2 fold mirrors
700mm mirror segments
• Grating (VPH)
– 200mm x 250mm
• f/1.8 Camera (from KMOS)
–
–
–
–
–
420mm focal length
±5˚ field
6 lens design
Ø 150-300mm lenses
2 HAWAII2 detectors
• 3.5m x 2.5m x 1.0m
HARMONI, Matthias Tecza
RAS, 9th May 2008
15
Phase A Study Developments
• Addition of field splitter at telescope focus allows
simplification of optics and layout (4 separated slicers,
easier pre-optics design).
• Presently studying completely reflective pre-optics
solution for wide wavelength coverage, which would also
allow dichroic split at disperser.
• Doubling the number of detectors to 16x 2K2 detectors
(4K spectral length) seems feasible. Beyond that cameras
get very challenging.
HARMONI, Matthias Tecza
RAS, 9th May 2008
16
HARMONI in context
• No study of such an instrument for E-ELT carried out so
far, although IRIS is part of TMT first-light suite
• Currently no other visible wavelength spectroscopic
capability planned for E-ELT (except for CODEX)
• Preliminary ESO estimate of 9M€ hardware costs.
HARMONI, Matthias Tecza
RAS, 9th May 2008
17
Spectral Discovery Space
15000
13000
Resolving power
11000
9000
7000
HARMONI
EAGLE
5000
IRIS
3000
NIRSpec
MUSE
1000
0.4
1.0
1.6
2.2
2.8
3.4
4.0
5.2
4.6
wavelength [µm]
HARMONI, Matthias Tecza
RAS, 9th May 2008
18
Spatial Discovery Space
100
Mapping: 35,000 spectra
Field of View ["]
MUSE
90,000 spectra
EAGLE
10
NIRSpec, 900 spectra
Single IFU: 441 spectra
HARMONI, 16,000 spectra
1
IRIS, 4096 spectra
0.1
0
25
50
75
100
125
150
175
200
225
250
pixel size [mas]
HARMONI, Matthias Tecza
RAS, 9th May 2008
19
Phase A Study Plan
• Phase 1: March ‘08 — January ‘09
– We are conducting a scientific-technical trade-off to determine optimal
instrument parameters in line with the instrument science cases
developed by the science working group. At the end of phase 1 we will
arrive at a single instrument concept to take forward into phase 2
– Investigate scientific drivers for optional modes: high spectral resolution
(R≈15,000) and bluer minimum wavelength (0.6µm)
– 2 science team meetings to date
• Phase 2: February ‘09 — November ‘09
– We will detail the chosen instrument concept to an advanced conceptual
design, together with a suggested management pathway to realise a
complete instrument inline with E-ELT requirements.
• Study review: December ‘09
HARMONI, Matthias Tecza
RAS, 9th May 2008
20
Phase A Study Milestones
Milestone
Kick-off meeting
Sience team meeting 1
Consortium meeting 1
Science team meeting 2
Progress meeting 1
Consortium meeting 2
Progress meeting 2
Science team meeting 3
Progress meeting 3
Phase 1 review
HARMONI, Matthias Tecza
Date
1st April 2008
12th Mar 2008
7th May 2008
8th May 2008
19th May 2008
15th July 2008
15th Sep 2008
16th Sep 2008
15th November
15th January (tbc)
RAS, 9th May 2008
Location
Videocon
Oxford
Oxford
Oxford
Garching
Oxford
Oxford
Oxford
Garching
21
Baseline Instrument Specifications
Field of view
5-10″, likely 2:1 format; ~100x200 spaxels
Spatial pixel scales
At least 3: 50 mas, 4 mas, 15 mas(TBD)
Wavelength range
0.8-2.4μm, visible extension
Spectral resolution
≥ 4000, 20000(?)
Simultaneous λ
coverage
2K-4K spectra possible
At least single band at medium/high res; goal:
entire spectral range at once
IFU technology
Image slicer? (best fill factor on detector)
Throughput
>35% average, incl. detector Q.E. (similar to
SINFONI)
AO performance
GLAO: 3-5x gain in EE in 50 mas spaxel (abs.
value 3.7% at K with GLAO!)
LTAO: K-60%, J-20%, NGS-19th mag.
MCAO: K-50%(uniform), NGS-19th/20th
HARMONI, Matthias Tecza
RAS, 9th May 2008
22