NGAO SDR Agenda
9:00
9:10
9:20
9:45
9:50
10:00
17:30
18:00
Welcome & Introductions (Armandroff)
Charge (Lewis)
Review Panel closed session (Hubin)
Re-entry for non-Panel participants
Comments from Chair (Hubin)
NGAO Report
General Discussion & Questions (Hubin et al.)
End
KAON 584: NGAO SDR Presentation
1
Keck AO has been a tremendous success thus far
Crab Nebula
LGS
AO
Kuiper Belt objects
Galactic Center
Merging galaxies
Substellar binaries
2
AO Science Productivity
Refereed Keck AO Science Papers by Year & Type
• 175 refereed AO science
papers
25.0
Solar System
Galactic
20.0
Number of Papers
– 38 LGS AO
– 18 Interferometer
Extra-galactic
15.0
10.0
5.0
Refereed Keck AO Science Papers by Year & Type
0.0
2000
30
2001
2002
2003
2004
2005
2006
2007
IF
25
Year
LGS
Number of Papers
NGS
20
15
10
5
0
2000
2001
2002
2003
2004
Year
2005
2006
2007
3
Most recent AO upgrade, NGWFC, resulted in
significant performance gains
Laser Guide Star
Natural Guide Star
60+% Strehl for R=14 NGS
4
WMKO is committed to raising
needed NGAO funds
• Funding plans will be reviewed by CARA Board
• Goodwin, Armandroff, Bolte & Kulkarni are charged with
NGAO fundraising
• 2/3 private support
– Very active Advancement Office at WMKO
– MOSFIRE: about 50% private support
• 1/3 federal support
– All relevant funding opportunities
– ExoPlanet Task Force recommendation: “Implement next-generation
high spatial resolution imaging techniques on ground-based
telescopes (AO for direct detection of young low mass
companions)”
5
Welcome Attendees
Reviewers: Norbert Hubin (Chair - ESO), Brent Ellerbroek (TMT),
Bob Fugate (NMT), Andrea Ghez (UCLA), Gary Sanders (TMT),
Nick Scoville (CIT)
SSC: Jean Brodie (UCSC), Tom Greene (NASA), Mike Liu (UH),
Chris Martin (CIT), Jerry Nelson (UCSC)
TSIP: Robert Blum, Mark Trueblood
Directors: Taft Armandroff (WMKO), Hilton Lewis (WMKO), Shri
Kulkarni (CIT)
NGAO Participants:
CIT: Antonin Bouchez, Rich Dekany, Anna Moore, Viswa Velur
UCSC: Don Gavel, Renate Kupke, Chris Lockwood, Claire Max, Liz
McGrath, Marco Reinig
WMKO: Sean Adkins, Erik Johansson, David Le Mignant, Chris
Neyman, Peter Wizinowich
6
Thank you for your role
in making Keck NGAO
a success!
Review Panel Report Questions
1. Assess the impact of the science cases in terms of the competitive landscape
in which the system will be deployed.
2. Assess the maturity of the science cases & science requirements and the
completeness & consistency of the technical requirements.
3. Evaluate the conceptual design for technical feasibility & risk, & assess how
well it meets the scientific & technical requirements.
4. Assess whether the design can be implemented within the proposed schedule
& budget.
5. Evaluate the suitability & effectiveness of the project management,
organization, decision making & risk mitigation approaches, with an emphasis
on the next project phase (preliminary design) and also with respect to the
entire project.
6. Provide feedback on whether the overall strategy will optimize the delivery of
new science.
7. Gauge the readiness of the project to proceed to the preliminary design phase.
8
NGAO SDR Agenda
9:00
9:10
9:20
9:45
9:50
10:00
17:30
18:00
Welcome & Introductions (Armandroff)
Charge (Lewis)
Review Panel closed session (Hubin)
Re-entry for non-Panel participants
Comments from Chair (Hubin)
NGAO Report
General Discussion & Questions (Hubin et al.)
End
9
NGAO System Design Review
Report
Peter Wizinowich, Rich Dekany, Don Gavel, Claire Max
for NGAO Team: S. Adkins, B. Bauman, A. Bouchez, M. Britton,
J. Bell, J. Chin, R. Flicker, E. Johansson, R. Kupke, D. Le Mignant,
C. Lockwood, E. McGrath, D. Medeiros, A. Moore,
C. Neyman, M. Reinig, V. Velur
System Design Review
April 21, 2008
NGAO SDR Agenda
10:00
10:05
11:15
11:30
12:00
12:30
13:30
14:00
14:45
15:15
15:45
16:00
16:40
Introduction & Presentation Approach (Wizinowich)
Science Cases & Science Requirements (Max)
Break
Requirements (Wizinowich)
Design (Gavel)
Lunch
Design Q&A (Gavel)
Performance Budgets (Dekany)
Project Management (Wizinowich)
Risks (Wizinowich)
Break
Cost Estimate (Dekany)
PD Schedule & Budget (Wizinowich)
+ Phased Implementation
17:20 Conclusion (Wizinowich)
17:30 General Discussion & Questions (Hubin et al.)
18:00
End
SCRD
SRD,FRD
SDM
SDM
SDM
SEMP
Risk KAONs
SEMP
SEMP
11
Tomorrow’s Agenda
8:30
9:30
10:00
11:30
Review Panel closed session (Hubin)
Questions for NGAO EC as needed
Review Panel closed session
Review Panel draft report (Hubin)
To Directors & NGAO EC
12:15
13:00
Lunch
End
12
Presentation Approach
•
The agenda topics were selected to correspond to the major System
Design deliverables and the 7 topics in the Review Panel Charge.
– With input from the Panel Chair.
•
Each session in the agenda is organized to:
– Provide a brief overview to address the specific charge & associated
questions.
– Provide answers to major questions from the reviewers.
– Provide time for additional reviewer questions & team responses.
•
Assumptions:
– People have read the System Design materials.
– Reviewers have read our responses to their questions.
•
89 questions received & answered.
– We will not need to use this meeting to bring people up to speed.
13
Science Case &
Science Case Requirements
Charges 1 & 2: Science Cases
• Charge 1: “Assess the impact of the science cases in
terms of the competitive landscape in which the system will
be deployed.”
– “Are the science cases given in the Science Case Requirements
document complete & compelling?”
• Charge 2: “Assess the maturity of the science cases &
science requirements ...”
– “Are the science requirements clear, complete & compelling?”
• NGAO Team response:
– NGAO will provide the WMKO community with an extremely
competitive & complementary facility.
– The science cases addressed to date are complete and
compelling, and the science requirements are well defined.
• Some requirements will be further developed during PD.
15
We categorized science cases into 2 classes
1. Key Science Drivers:
– These push the limits of AO system, instrument, and
telescope performance. Determine the most difficult
performance requirements.
2. Science Drivers:
– These are less technically demanding but still place
important requirements on available observing
modes, instruments, and PSF knowledge.
16
“Key Science Drivers”
(in inverse order of distance)
1.
High-redshift galaxies
2.
Black hole masses in nearby AGNs
3.
General Relativity at the Galactic Center
4.
Planets around low-mass stars
5.
Asteroid companions
17
“Science Drivers”
(in inverse order of distance)
1.
Gravitationally lensed galaxies
2.
QSO host galaxies
3.
Resolved stellar populations in crowded fields
4.
Astrometry science (variety of cases)
5.
Debris Disks and Young Stellar Objects
6.
Giant Planets and their moons
7.
Asteroid size, shape, composition
18
Conclusions from Science Cases
• Our scientists want a high performance AO system that will enable a
wide variety of science cases
• They want it to open up new vistas of both wide and narrow field
science at shorter wavelengths and higher sky coverage
• We determined that these science goals could best be met by using
new technologies rather than modest extension of existing ones
– Scaling existing technologies did not meet the desired science
performance (KAON 461)
– Any new Keck AO system will be expensive, and hence should have a
commensurately large payoff
• Keck has an excellent history of world leadership in AO
– First high-order AO systems on 8-10 m telescopes
– First operational laser guide star
• High payoff at modest risk are consistent with Keck’s approach to
science and instrumentation
19
Charge 1: “Assess impact of science cases in
terms of competitive landscape...”
• Other ground-based observatories
• JWST & ALMA
• TMT
20
NGAO in the world of 8-10 m telescopes:
Uniqueness is high spatial resolution, shorter ’s, AO-fed NIR d-IFS
Table 1. Next-Generation AO Systems Under Development for 8 - 10 meter Telescopes
Type
Telescope
High-contrast
Subaru
High-contrast
VLT
High-contrast Gemini-S
GS
Next-Generation AO Systems
for 8 to 10 m telescopes
N/LGS Coronagraphic Imager Hi(CIAO)
Capabilities
Dates
Good Strehl, 188-act curvature,
4W laser
2008
NGS
Sphere (VLT-Planet Finder)
High Strehl
2010
NGS
Gemini Planet Imager (GPI)
Very high Strehl
2010
Wide-field
Gemini-S
5 LGS
MCAO
2Õ FOV
2009
Wide-field
Gemini
4 LGS
GLAO
Feasibility Study Completed
?
Wide-field
VLT
4 LGS
HAWK-I (near IR imager) +
GRAAL GLAO
7.5' FOV, AO seeing reducer,
2 x EE in 0.1''
2012
Wide-field
VLT
4 LGS
MUSE (24 vis. IFUs) +
GALACSI GLAO
1' FOV; 2 x EE in 0.2" at 750nm
2012
Narrow-field
VLT
4 LGS
MUSE (24 vis. IFUs) +
GALACSI GLAO
7.5Ó FOV, 10% Strehl @ 750 nm
2012
•
Most 8-10 m telescopes plan either high contrast or wide field AO
•
Only the VLT has a narrow-field mode (7.5” FOV, 10% Strehl @ 750 nm)
21
JWST will push major advances in:
• End of the Dark Ages
• Assembly of galaxies
• Birth of stars, protoplanetary systems
• Properties of planetary systems including our own
Our goal is to position NGAO to build on,
and complement, JWST discoveries
22
Competitive Landscape: JWST
•
JWST advantages
– JWST will have better sensitivity than NGAO (low
backgrounds)
– Diffraction limited imaging between 2.4 and 5 m
– Multiplexed slit spectroscopy (x 100)
• But only 1 IFU
– Maximum spectral resolution R = 2700
•
Keck NGAO advantages
– Better spatial resolution than JWST at
wavelengths below 2 m
• JWST pixels under-sample the diffraction limit at
these wavelengths
–
–
–
–
Spectroscopy at spatial resolutions < 0.1”
Multi-IFU spectroscopy
Spectroscopy at spectral resolutions R > 2700
Higher resolution imaging at wavelengths < 2 m
23
Competitive Landscape: ALMA
•
Millimeter and sub-millimeter wavelengths
(0.35 - 9 mm)
•
Typical spatial resolutions ~ 0.1”
•
Resolutions for widest arrays as low as
0.004” at the highest frequencies
•
ALMA science: regions colder and more
dense than those seen in the visible and
near-IR by NGAO
•
Keck NGAO observations of H2 and atomic
hydrogen near IR emission lines:
characterize warmer outer regions of the
disks and molecular clouds seen by ALMA,
at similar spatial resolution
•
Keck NGAO and ALMA observations
complementary for:
–
Spatially resolved galaxy kinematics, z < 3
–
Debris disks and young stellar objects
24
Complementarity with TMT
• TMT (2017?)
– TMT has significant spatial resolution &
sensitivity advantages over NGAO
– NGAO d-IFS has spatially resolved
spectra & higher spatial resolution than
TMT’s IRMS; available a generation
before IRMOS
– NGAO proving MOAO, variable
asterisms, Point-and-Shoot
sharpening, MEMS DM’s, to TMT’s
benefit
25
NGAO with multiplexed IFU:
a real complement to TMT
• TMT IRMS: AO multi-slit (MOSFIRE) fed by MCAO
– Slits: 0.12” and 0.16”, Field of regard: 2 arc min
– Lower backgrounds: 10% of sky + telescope
• NGAO with multiplexed deployable IFUs
– MOAO  better spatial resolution than MCAO over full field
– Better spatial resolution: 0.07” is current spec.
– Higher backgrounds:  30% of sky + telescope (but much better
than current AO system)
•
•
•
TMT IRMS strengths: lower backgrounds, higher sensitivity
NGAO d-IFU strengths: higher spatial resolution, 3D information,
wide field performance
NGAO d-IFU a pathfinder for TMT IRMOS
26
Charge 2: “Assess the maturity of the science
cases & science requirements ...”
• Science Cases fully described in the Science Case
Requirements Document (SCRD, KAON 455)
• Here: Choose one “Key Science Driver” and walk
through the requirements process with you
Galaxy Assembly and Star Formation History
•
•
•
•
Broad scientific goals
Major sub-cases
How requirements were derived
Remaining issues
27
Galaxy Assembly and Star Formation History:
Focusing our Analysis
• “High Redshift Galaxies” has very wide scope
– z > 6:
Finding and characterizing galaxies
– 3 < z < 6: Morphologies, colors
JWST will
excel here
– 1 < z < 3: Internal kinematics, structure at time of peak star formation
• To define “Key Science Driver” we focused on 1 < z < 3
– 1 < z < 3 epoch: spatial resolution of 10-m telescope has strong impact
• Prominent emission lines redshifted to J, H, K bands
• Sufficient signal-to-noise to spatially resolve internal kinematics, star
formation rates, metallicity gradients using spatially resolved spectroscopy
Internal
velocities
Star
Formation
High Redshift
Galaxy
Bulge
Metallicity
star
cluster Supernova
Spiral Arm
28
What is happening to galaxies at 1  z  3?
•
At z ~ 1 – 3, galaxies accumulate most of their stellar
mass, rate of major mergers peaks.
•
This activity transforms irregular galaxies into the
familiar Hubble sequence of the local universe.
•
Studying these galaxies in detail is key to
understanding galaxy formation and the buildup of
structure in the universe.
– Global properties of these galaxies are being well
studied.
– Little is known about internal kinematics or small-scale
structure, mode of dynamical support, spatial
distribution of star formation.
– Is star formation due to rapid nuclear starbursts during
major mergers? To circum-nuclear starbursts caused
by bar-mode or other gravitational instabilities? Or to
consumption of gas reservoirs in stable rotationallysupported structures?
29
Substantial benefit from observing
many of these galaxies at once
Type of Object
SCUBA sub-mm galaxies to 8 mJy
Old and red galaxies with 0.85 < z < 2.5
and R < 24.5
Density per
sq arc min
0.1
2
Field galaxies w/ emission lines in JHK
windows, 0.8 < z < 2.6 & R < 25
> 25
Center of rich cluster of galaxies at z > 0.8
> 20
All galaxies K < 23
> 40
•
In survey mode, could make good use of as many as 20-25 IFUs at once
•
In more focused mode, typical science paper will study a sub-category of
these galaxies. Multiplexing factors of 6-12 fit many subcases.
30
Many Sub-Cases, Galaxies at 1  z  3
• Kinematic evolution from random sub-pieces to organized rotation
• Patterns of star formation (nuclear first? rings? uniform? ...) and their
trends with redshift
• Dependence of star formation rate on current merger activity and/or
existence of close companion galaxies
• How does status as Active Galactic Nucleus influence star formation
pattern and rate?
• Does status as Active Galactic Nucleus correlate with recent merger
activity? Existence of close companion galaxies?
• Sub-classes of targets will be selected using ongoing large surveys
(e.g. COSMOS, GOODS, ...)
Goal is to derive science requirements to jointly optimize as many
of these sub-cases as possible
31
Requirements Shared by Most Sub-Cases
•
Spatially resolved spectroscopy (2 spatial dimensions)
– e.g. to distinguish ordered rotation from discrete sub pieces, to see patterns of
star formation or metallicity
– Size of field for each galaxy? “Typical” galaxy is 1 arc sec; want additional real
estate in order to measure sky background or to accomodate larger galaxies
when needed. Chose 1” x 3” as minimum field size for IFUs.
•
High sky coverage fraction: ≥ 30%
•
Multiplexing to maximize science return per hour of observing
– Multiplexing factor of N is equivalent to N Keck Telescopes
– Requirement: Target sample size of ≥ 200 galaxies observable with ~10 nights
of allocated telescope time. (More on next slide)
•
Spectral bands: J, H, K with spectral resolution 3000-4000
– Major emission lines redshifted into JHK (H and [NII], [OII], [OIII])
– Spectral resolution chosen to look between the OH night-sky lines
• Choose lowest resolution that does this, to preserve faint-object sensitivity
32
Sensitivity Requirement is the
Hardest to Jointly Optimize
•
Overall requirement: spectra of ≥ 200 high-z
galaxies in 10 nights of observing time
 Must be able to observe ≥ 20 galaxies per
10-hour night (see table) to SNR ≥ 10
•
Choice of pixel / spaxel scale is key, for
galaxies with at least some fuzzy structure
– Extended H emission, low surfacebrightness disks, largest galaxies, ...
– For these, larger pixels/spaxels are better for
SNR. Optimum at 0.1”/px or more.
– But of course larger pixels/spaxels are worse
for spatial resolution
•
For smaller galaxies at 1  z  3, or those
that have point-like substructures, pixel
scales  0.05” are best
Integration time to
reach SNR ≥ 10
implies
this
minimum
IFU
Multiplicity
0.5 hrs
1
1 hr
2
2 hrs
4
3 hrs
6
6 hrs*
12
10 hrs*
20
* Not desirable
33
Sensitivity Requirement and Pixel/Spaxel Size, continued
• Recap:
– Sensitivity will depend on pixel/spaxel size
– Different sub-cases of the 1 ≤ z ≤ 3 science case optimize at different
pixel/spaxel sizes
– Large galaxies with diffuse Ha emission: 0.1” / px or more
– “Galaxies” consisting of several point-like star-forming knots: ≤ 0.05”
• Compromise for the deployable IFU: pixel/spaxel scale = 0.07”
– Narrow-field high-resolution IFU (OSIRIS-like) will have variable scales.
For example OSIRIS goes down to 0.02”
• Implications for the deployable IFU:
– Can meet 200-galaxy requirement with 0.07” spaxels, background due to
AO less than 30% of unattenuated (sky + telescope) between OH lines
– Yields 2-3 hr integration times (see next slide), min 4-6 deployable IFUs
34
Requirement on AO background:
Example of analysis logic
•
Tint = 3 hours  AO contribution to background = 30%, 6 IFUs
•
Then 70% throughput  cool AO system to -15C
•
Calculations will be refined for PDR, now that optical design is defined
35
Each Science Driver has a “Requirements Table”
• Summarizes requirements
discussed in text and figures
• Formatted for input into
System Requirements
Document and into the
Contour Database of
Functional Requirements
• Example: part of the
Requirements Table for HighRedshift Galaxies
36
Science Team Tasks During PD Phase
• Expand upon goals of “Science Drivers”, and finish documenting the
AO performance necessary to achieve these goals.
• Generate a Science Requirements Summary Matrix that rolls-up the
most demanding requirements for each part of the architecture
• Develop detailed observing scenarios for each “Key Science Driver”
to define pre- and post-observing tools and observing sequences.
• Detailed science simulations of “Key Science Drivers” to assess the
required level of PSF accuracy, stability, uniformity, and knowledge
as a function of position and time. Implications for:
– achievable astrometric and photometric accuracy
– achievable contrast ratio
– morphological studies
37
Science Team Tasks During DD-FSD Phases
• Develop a “Design Reference Mission” for at least two “Key Science
Drivers”
– Simulate expected on-sky performance.
– End-to-end simulation: planning tools, observing proposal, observing
sequences, science operations, PSF models, analysis tools, data
products.
– Integrate tasks and deliverables from throughout the NGAO Work
Breakdown Structure to ensure they work together and provide a
seamless observing process that meets all specifications.
• Design Reference Mission will help ensure that commissioning runs
smoothly, to advance to full-scale science operations as quickly as
possible and maximize the scientific return of NGAO.
38
Additional Science Team Efforts
• Continued discussions with Keck community to ensure that science
case requirements remain consistent and up-to-date with changing
methodology, advancing AO system design, and maturing
instrument concepts.
• Input from observers to improve planning tools, observing practices,
support, and efficiency.
• Feedback regarding NGAO science opportunities that complement
other ground-based AO and space-based facilities, and that take
advantage of the uniqueness space provided by NGAO at Keck.
39
Reviewer Q & A
• MCAO/MOAO Trade-offs
• Contrast requirements and capabilities
• PSF requirements and analyses
40
MCAO/MOAO Tradeoff: Key Science Drivers
• Four of the five Key Science Drivers use very narrow fields:
–
–
–
–
Black hole mass in nearby AGNs
General Relativity in the Galactic Center
Extrasolar planets around nearby stars
Minor planet multiplicity
( 5 arc sec field)
(10 arc sec field)
(5 arc sec field)
(3 arc sec field)
The fifth Key Science Driver, Galaxy
Assembly and Star Formation History,
needs wide fields and high sky coverage.
In all Science Cases, infrared tip-tilt stars
need to be AO-corrected, for high sky
coverage.
More on this later
41
All narrow-field Key Science Drivers are
within one isoplanatic angle
Seeing:
r0
0
Challenging
Median
Good
Excellent
0.5 m
14 cm
16 cm
18 cm
22 cm
0.85 m
1.2 m
1.6 m
2.2 m
4.1”
6.2”
8.7”
12.7”
5.1”
7.7”
11.3”
16”
5.5”
8.3”
11.7”
17.2”
7.6”
11.4”
16.2”
23.7”
• Nearby AGNs, extrasolar planets, multiplicity of minor planets
use 0.85    1.6 m, field radii  3 arc sec
• Galactic Center uses  ~ 2.2 m, field radius  5 arc sec
These cases don’t need MCAO or
MOAO for the science field
42
But most narrow field science cases need
MCAO or MOAO for high sky coverage
•
Laser tomography needs 3 natural stars for tip-tilt and other low modes
•
For high sky coverage, these tip-tilt stars must be AO-corrected (can
use fainter stars which are more plentiful)
•
Drives towards infrared tip-tilt stars, since these will have better AOcorrection than visible ones
•
For AO correction of widely spaced tip-tilt stars, must have laser
asterism extending to relatively large radius
– TMT NFIRAOS: 2 arc min technical field for tip-tilt sensors, lasers on 1.2 arc
min diameter circle.
– NGAO: 2.5 arc min technical field for tip-tilt sensors; point 3 lasers directly at
tip-tilt stars; science lasers variable up to 2.5 arc min diameter field.
•
Can be done either using overall wide-field MCAO correction, or putting
MOAO units within each tip-tilt sensor.
– This decision is independent of whether science field uses MCAO or MOAO.
43
Science Drivers (not “Key”):
Which ones need wide science field?
• Narrow Field Science (< isoplanatic angle, don’t need
MOAO or MCAO except for tip-tilt)
–
–
–
–
–
–
–
–
QSO Host Galaxies
Gravitational lensing of galaxies by galaxies
Some of the narrow-field astrometry science
Debris disks
Young Stellar Objects
Size, shape, composition of minor planets
Gas giant planet moons
Uranus and Neptune
• Wider Field Science
–
–
–
–
–
Can be done by
mosaicing
smaller fields
Gravitational lensing by clusters
Some wide-field astrometry science cases
Resolved stellar populations in crowded fields
Imaging of Jupiter and Saturn disks and rings
Imaging of Uranus and Neptune rings
44
Science Drivers (not “Key”):
Which ones need wide science field?
• Narrow Field Science (< isoplanatic angle, don’t need
MOAO or MCAO except for tip-tilt)
–
–
–
–
–
–
–
–
QSO Host Galaxies
Gravitational lensing of galaxies by galaxies
Some of the narrow-field astrometry science
Debris disks
Young Stellar Objects
Size, shape, composition of minor planets
Gas giant planet moons
Uranus and Neptune
• Wider Field Science
–
–
–
–
–
Potentially
benefits most
from MCAO
Gravitational lensing by clusters
Some wide-field astrometry science cases
Resolved stellar populations in crowded fields
Imaging of Jupiter and Saturn disks and rings
Imaging of Uranus and Neptune rings
45
Can NGAO meet its contrast goals?
• Science Case:
Planets around
nearby low-mass
stars
Giant planet (2x mass
of Jupiter)
Brown dwarf
1/30 mass of
Sun (hidden
behind occulting
mask)
Simulations by Bruce Macintosh
46
and Chris Neyman
Can NGAO meet its Contrast Goals?
Target Sample 1: Old field brown dwarfs to 20pc
Requirement: H=14, H=10 at 0.2” (2MJ at 4 AU)
Target Sample 2: Young (<100Myr) brown dwarfs & low-mass stars to 80pc
Requirement: J=11, 1MJ: J=11, 2MJ: J=8.5
a. (minimum) J=8.5 at 0.1”
b. (nominal) J=11 at 0.2”
c. (goal)
J=11 at 0.1”
Target Sample 3: Solar-type stars in Taurus and Ophiuchus, and young
clusters at 100-150 pc.
Requirement: J=10-12, 1MJ: J=13.5, 5MJ: J=9
a. (difficult goal) J=13.5 at 0.07”
b. (goal) J=9 at 0.07”
47
Simulations of Contrast Performance
•
•
Numerical simulation
inputs:
– Keck pupil
– 7 layer turbulence
model, median to
good conditions
– 36x36 subaps
– Measurement
errors due to spot
elongation &
fratricide
– 1 kHz frame rate
– 5 LGS at 11”
– Tip/Tilt error (3
NGS, J=16 at 30”)
– Static telescope
errors - 65 nm
Occulting spot sizes
NGAO, no coronagraph
Coronagraphs
No treatment of noncommon-path errors
yet
48
Target
Sample 1
Target
Sample 2a
Target
Sample 2b
Target
Sample 2c
Target
Sample 3b
Target
Sample 3a
49
Conclude that Science “Requirements” (but only one
“Goal”) can be met for Exoplanets science case
Target Sample 1: Old field brown dwarfs to 20pc
Requirement: H=14, H=10 at 0.2” (2MJ at 4 AU)
 at 8 
Target Sample 2: Young (<100Myr) brown dwarfs and low-mass stars to
80pc
Requirement: J=11, 1MJ: J=11, 2MJ: J=8.5
a. (minimum) J=8.5 at 0.1”
b. (nominal) J=11 at 0.2”
c. (goal)
J=11 at 0.1”
 at 8 
 at 5 

Target Sample 3: Solar-type stars in Taurus and Ophiuchus, and young
clusters at 100-150 pc.
Requirement: J=10-12, 1MJ: J=13.5, 5MJ: J=9
a. (difficult goal) J=13.5 at 0.07”

b. (goal) J=9 at 0.07”
? at 5 
50
What are the requirements for
PSF stability and knowledge?
• In System Design phase, we stated requirements in terms of
photometric and astrometric accuracy
• These in turn will be only achievable with specific levels of PSF
stability, uniformity, and knowledge
– “Stability” refers to temporal uniformity
– “Uniformity” refers to spatial uniformity (specify over what field)
– “Knowledge” -- no matter what the actual stability and uniformity, how
well do you know the PSF that pertained during a specific science
exposure?
51
PSF-Related Plans During Preliminary Design Phase
• During Preliminary Design phase we will:
– Develop a set of quantitative measures of “PSF Knowledge”
• Different science cases are sensitive to different aspects of the PSF
– Translate our photometry and astrometry requirements into specific
requirements on PSF knowledge
– Develop an astrometry error budget for specific science cases
– Work with ongoing projects at CfAO etc. to develop methods of deriving
PSFs from atmospheric measurements plus with telemetry from the
laser tomography system
• CfAO projects under David Le Mignant & Ralf Flicker, and Matthew Britton
• Ongoing work at other observatories (VLT, Gemini, NSO)
• Main issues? Based on experience at CFHT and on simulations, we expect
to do well if the laser tomography AO system at tip-tilt system are operating
at high signal to noise. For cases where the SNR is low, need “plan B”.
• Monitor PSF stars at other places in field, etc.
52
NGAO SDR Agenda
10:00
10:05
11:15
11:30
12:00
12:30
13:30
14:00
14:45
15:15
15:45
16:00
16:40
Introduction & Presentation Approach (Wizinowich)
Science Cases & Science Requirements (Max)
Break
Requirements (Wizinowich)
Design (Gavel)
Lunch
Design Q&A (Gavel)
Performance Budgets (Dekany)
Project Management (Wizinowich)
Risks (Wizinowich)
Break
Cost Estimate (Dekany)
PD Schedule & Budget (Wizinowich)
+ Phased Implementation
17:20 Conclusion (Wizinowich)
17:30 General Discussion & Questions (Hubin et al.)
18:00
End
SCRD
SRD,FRD
SDM
SDM
SDM
SEMP
Risk KAONs
SEMP
SEMP
53
Requirements
Charge 2: Requirements
• Charge 2: “Assess … the completeness & consistency of
the technical requirements.”
– “Is a clear flow down established from the science requirements to
the technical requirements?”
– “Are the technical requirements clear, complete & documented?”
• NGAO Team response:
– The science, system and technical requirements are well
documented in a requirements database.
– The flow down has clearly driven the technical requirements.
– Additional review is required in the preliminary design to ensure
that the technical requirements and flow down documentation is
complete and consistent.
55
56
Science Requirements Summary
57
Science Requirements Flow-Down (1/2)
1. Dramatically improved performance at NIR wavelengths.
a. Improved IR sensitivity.
•
•
High Strehls required over narrow fields. Flow-down derived from WFE
performance budget & assumptions about how error terms can be met.
Lower backgrounds. Need to cool AO system & operating temperature driven
by high redshift galaxy science.
b. Improved astrometric, photometric & companion sensitivity performance.
•
•
Improved IR sensitivity required (see above).
Improved PSF stability & knowledge. Required PSF stability & knowledge,
astrometric budget & PSF tools TBD during PD.
2. Increased sky coverage.
•
Wide field required.
•
•
Drove architecture to incorporate a wider field of regard for tip-tilt stars than
needed for d-IFS science alone.
Ability to use faint NGS.
•
Drove architecture to incorporate MOAO correction of tip-tilt stars to achieve
high infrared Strehl ratio.
58
Science Requirements Flow-Down (2/2)
3. Efficient extragalactic target surveys.
a. Science instrument.
•
•
•
Efficient acquisition of spectral & imaging data drove IFS selection.
Multiple targets over 2′ dia field & survey efficiency drove multiple heads.
Need to adapt to observation field drove deployable heads.
b. Sensitivity.
•
•
•
Required image resolution allowed EE requirement requiring fewer actuators
than for narrow field science.
This, & requirement to AO correct tip-tilt NGS over a wide field, drove
MC/MOAO choice. Maximizing performance over narrow non-contiguous fields
led to MOAO.
Low backgrounds need drove cooled AO enclosure.
4. AO correction in red portion of visible spectrum.
•
Drove requirements to transmit to visible science instruments & to share
visible light with LGS & NGS WFSs via dichroics.
5. Science instruments that facilitate the range of science programs.
•
•
Drove science instrument selection & concept.
Drove providing locations for these science instruments.
59
Database Tool for Requirements
Management
• NGAO will have a few thousand requirements.
• Team needs to keep science, system and subsystem requirements
consistent.
• Searched for an affordable software tool & selected Contour by Jama
Software, Inc. Key features:
–
–
–
–
–
–
web-based
multi-user tool
no client software required
easy to use and configurable to meet our needs
affordable
user community includes Intel, Amgen, and Lockheed-Martin, as well as
smaller companies and startups
60
Key Requirements Data Contained in One Place
Short name for
easier searching
Requirements document
section: easier to organize
final documents
Organized by
SRD section
Rationale and traceability
61
Key Requirements Data Contained in One Place
Link to other database
elements (traceability)
Add notes and
email change
notices
Attach documents or
web links
Compare versions
62
Reviewer Q & A
64
NGAO SDR Agenda
10:00
10:05
11:15
11:30
12:00
12:30
13:30
14:00
14:45
15:15
15:45
16:00
16:40
Introduction & Presentation Approach (Wizinowich)
Science Cases & Science Requirements (Max)
Break
Requirements (Wizinowich)
Design (Gavel)
Lunch
Design Q&A (Gavel)
Performance Budgets (Dekany)
Project Management (Wizinowich)
Risks (Wizinowich)
Break
Cost Estimate (Dekany)
PD Schedule & Budget (Wizinowich)
+ Phased Implementation
17:20 Conclusion (Wizinowich)
17:30 General Discussion & Questions (Hubin et al.)
18:00
End
SCRD
SRD,FRD
SDM
SDM
SDM
SEMP
Risk KAONs
SEMP
SEMP
65
NGAO System Design
Charge 3&7: Design
• Charge 3: Evaluate the conceptual design for technical
feasibility …, & assess how well it meets the scientific &
technical requirements.
– Does the conceptual design appear feasible?
• Charge 7: Gauge the readiness of the project to proceed to
the preliminary design phase.
– Is the technical design sound?
– Is the design concept & architecture adequately documented?
• NGAO Team response:
– A feasible and optimal conceptual design has been developed and
documented to meet the science and technical requirements.
67
NGAO Block Diagram
Wide field
science
Wavefront
sensing
Narrow field science
68
NGAO Asterisms
Narrow Field Science
Wide Field Science
69
Flowed-Down Key Architectural Features (1/2)
•
Laser tomography to measure wavefronts & overcome cone effect.
•
Variable radius LGS asterism to maximize performance for each science field
& changing atmospheric turbulence profiles.
•
LGS projection from behind secondary to minimize perspective elongation.
•
Nasmyth platform location for sufficient space & stability.
•
Cooled AO system to meet background requirements.
–
•
Alternate approaches including adaptive secondary considered.
K-mirror rotator at AO input to keep field or pupil fixed.
–
–
AO system would need cooling even without a rotator.
Provides most AO/instrument stability.
•
Wide-field (150" dia.) relay to LGS wavefront sensors, TT sensors, & d-IFS.
•
Conventional (5 mm pitch) DM to transmit wide field.
•
Low-order (20 actuators across pupil) DM for wide-field relay
–
–
–
Limits size.
Permits closed loop AO on LGS WFSs & keeps them in smaller, more easily
calibrated linear range.
Reduces requirement on downstream open-loop correction.
70
Flowed-Down Key Architectural Features (2/2)
•
Open loop MOAO-corrected NIR TT sensors to maximize sky coverage.
–
–
–
–
MOAO maximizes delivered Strehl over narrow fields.
Open-loop correction applies tomographic reconstruction result to point in field.
NIR sensing since AO will sharpen NGS image & provide better TT information.
2 TT sensors & 1 TT-focus-astigmatism sensor provides optimum correction.
•
Open loop MOAO-corrected d-IFS heads to meet ensquared energy
requirement over required field of regard.
•
Open loop MOAO-correction to narrow field science instruments to reduce
cost (compared to alternate large relay &/or MCAO architectures).
•
MEMS DM’s for MOAO-correction.
–
–
–
•
Very compact. Lab demonstrated to accurately go where commanded.
Small, modest cost 32x32 MEMS DM’s provide TT sensors & d-IFS correction.
64x64 element MEMS provides AO correction to narrow field science instruments.
High order AO relay to feed narrow field (30“ dia.) science instruments.
–
–
–
–
Science instruments fed by this relay only require a narrow-field.
Narrow field facilitates single MEMs DM for all instruments.
Science instruments include NIR & visible imagers & OSIRIS.
Larger, 60" diameter, field to the NGS WFS.
71
AO System Optical Layout
72
Science Instruments
•
•
•
Provide imaging and spectroscopy over the full NGAO passband
Support both narrow field and ‘wide’ field relays
Use simple, heritage based designs where possible
•
Imagers
–
–
–
–
–
–
–
Near-IR 0.97 to 2.4 µm, Visible 0.7 (0.62) to 1.05 µm
30" diameter FOV
Well sampled (3 pixel) diffraction limited images
Coronagraph
Selection of filters
Integral field option (2" x 2") for the visible wavelength range
Low (well controlled) background
73
Science Instruments
•
Spectrographs
– Near-IR single object integral field spectrograph (IFS)
•
•
•
•
0.97 to 2.4 µm, diffraction limited spatial sampling (20 to 35 mas)
Up to 4" x 4" FOV desired
R ~3,000 to 4,000 (goal of ~15,000)
Use OSIRIS initially (0.32" x 1.28" FOV at 20 mas) and develop new
instrument in a later phase
– Visible single object IFS
•
Provided by integral field unit in visible imager
– Near-IR deployable multi-channel IFS (d-IFS)
•
•
•
•
•
1 to 2.4 µm, spatial sampling optimized for 50% EE (50 to 70 mas)
Six channels, 1" x 3" FOV, total FOR 120"
R ~4,000
Close packed mode (goal of 0.5" gaps between channels in one dimension)
Direct imaging mode (through slicer)
74
LGS Facility Layout
75
System Configurations
Science Cases
drive the AO
system
configurations
(i.e., required
hardware &
control)
#
Key Science Drivers
1 Minor planets as remnants of early Solar System
Survey
Orbit Determination
2 Planets around low-mass stars
Survey
Spectra
3 General Relativity at the Galactic Center
Astrometry
Radial Velocities with dIFS
Radial Velocities with OSIRIS
4 Black hole masses in nearby AGNs
5 High-redshift galaxies
#
Science Drivers
1 Asteroid size, shape, composition
2 Giant Planets and their moons
Imaging
Spectroscopy
3 Debris disks and Young Stellar Objects
Imaging
Spectroscopy
4 Astrometry in sparse fields
5 Resolved stellar populations in crowded fields
6 QSO host galaxies
Imaging
Spectroscopy
7 Gravitationally lensed galaxies by other galaxies
Imaging
Spectroscopy
8 Gravitationally lensed galaxies by clusters
Imaging
Spectroscopy
9 Backup Science
Faint NGS science
Seeing-limited science with acquisition camera
NGS Configuration
Primary Secondary
LGS Configuration
Primary Secondary
4e
5a
4a-d,4f
4ef,5b
4c-f,6c-f
6c-f,4c-f
4a
1
6ac
Vis spectra
1
4a
6a
4ab
6ab
5a, 7a
4c
6ce
5a
4e
4ace
6ace
5a, 7ace
4a-f
6a-f
4ace
4ace
4ace
6ace
4ace
6ace
5a
4ace
1
9ab
76
System Configurations
#
Configuration
LGS Science Modes
1
d-IFS
4a
4b
4c
4d
4e
4f
NIR Camera
5a
5b Visible Camera
#
Science
Science Field
Rotator
λ
Diameter Mode
Projected
LGS
Asterism
LGS
Asterism
Rotation
≤ 120"
≤ 30"
≤ 2"
≤ 30"
≤ 2"
≤ 30"
≤ 2"
≤ 30"
≤ 2"
Wide
≤ Medium
Narrow
≤ Medium
Narrow
≤ Medium
Narrow
≤ Medium
Narrow
Yes for field
Yes for field
No, fixed
Yes for field
No, fixed
Yes for field
No, fixed
Yes for field
No, fixed
z-K
K
H
z-J
0.7-1µm
LOWFS LOWFS 1st relay PSF
& PSF
+
Truth Monitor
ADC Tweeter Sensor +Tweet
LGS Science
1 Tracking
4a
4b
4c
4d
4e
4f Tracking
5a
5b Tracking
Modes
3
3
1 on-axis
3
1 on-axis
3
1 on-axis
3
1 on-axis
Yes
Option
Unlikely
Option
Unlikely
Option
Unlikely
Option
Unlikely
Option
Option
Unlikely
Option
Unlikely
Option
Unlikely
Option
Unlikely
d-IFS
Fixed field
Fixed field
Fixed pupil
Fixed field
Fixed pupil
Fixed field
Fixed pupil
Fixed field
Fixed pupil
2nd
NGS
Interfero
Relay
WFS
meter OHANA Tweeter Dichroic
Woofer
LGS
dichroic
LGS
WFSs
Yes
In
Yes
In
Yes
In
9
9
6
9
6
9
6
9
6
NGS
Field
Visible Steering 2nd FSM
ADC
Tracking Position
NGS
WFS
InterferPost Relay ometer
1 Dichroic
Fold
Out
JH transmit
/ K reflect
J transmit /
H reflect
H transmit
/zJ refl
JH transmit
/ vis reflect
2nd relay Visible
Truth
Imager
Sensor Dichroic
Visible
Imager
ADC
NGS LOWFS/T
Acquis WFS Diff
Fold Tracking
Out
Out
Out
Out
Out
Out
Visible
Imager
NIR
ADC
No
Option
No
Option
No
Option
No
Option
No
OSIRIS
NIR
Fold OSIRIS Camera
Yes
Yes
Yes
Out
Out
Option
Track or
out
Option
TWFS
TWFS
Yes
on-axis
Yes
on-axis
Yes
on-axis
Yes
on-axis
Out or IR
trans
Vis reflect
Track or
out
Track or
out
Out
Yes
Yes
77
Flowed-down Key Control Features
• AO controls fully integrated with observatory
• Integrated control of AO, laser, telescope offloads
• Three-stages of science operations support:
– Observation planning tools
– On-sky operations sequencer
– Post-processing tools, e.g. incorporating AO diagnostics, PSF
calculation, etc.
• Archiving system
• etc…
• Separate safety control system
78
Science Operations Design
• Classical observing model: astronomers are in charge
– Model-of-choice for our community and the Observatory.
– Built-in flexibility to switch between NGAO mode and instruments.
– Back-up program is the responsibility of the astronomer.
• Within these constrains, the science operations design optimizes.
observing efficiency: 80% open shutter time for high-z galaxies
– Pre-observing tools: selection of guide stars, performance and SNR
prediction, planning and saving the observation sequences.
– Operations tools integrating NGAO, telescope and instruments, allowing
for parallel command and multi-system coordination.
– Dithering/offsetting/centering using internal steering optics, that do not
require to open/close AO loops and offset telescope.
• Quality of the final data product:
– Use of WFC and ancillary data for monitoring atmospheric conditions
and image quality (SR, EE, photometry, etc).
– Data archiving for calibration and science products.
– PSF calibration, including PSF reconstruction from telemetry.
79
Science Operations Design
Pre-observing tools
GUIs and high-level operations tools
Multi-system Command
Sequencer
Subsystem Command Sequencer
80
Control System Block Diagram
• Highly distributed control system
• Client/Server relationships
between components, with master
sequencer
• Communication paths identified
• Communication protocols TBD
81
Non-Real-Time Control Elements
• Motion and bench automation control
–
–
–
–
–
–
–
–
–
Field derotator
Calibration source in/out
Dichroic changers
LGS WFS assembly
LOWFS pickoff assembly
Acquisition pickoffs
Laser constellation configuration
Laser pointing and centering
Etc.
• Device control
–
–
–
–
Sensors configuration control (HOWFS, LOWFS, …)
Device power control (DMs, cal sources…)
Environmental control (temperature, humidity, particulates, cooling)
Laser diagnostics sensors and power/environment control
Within a uniform
motion control
architecture and
design approach
• Standardization of
motors/servos
• Uniform specs for
electronic drives
• Electronics
location
requirements
• Reliability
specifications
82
Flowed-Down Real-Time Control Features
• Performs high speed tomography of the atmosphere
–
–
–
–
Up to 9 LGS, 3 LOWFS (2 tip/tilt, 1 tip/tilt/focus/astigmatism) sensors
Up to 11 DMs (woofer, narrow field tweeter, 6 d-IFS tweeters, 3 LOWFS tweeters
Multiple atmospheric layer tomography
Up to 2kHz frame rate
• Incorporates external parametric information
– Cn2 (atmospheric) profile
– Sodium layer profile
– Wind profile
• Flexible with science observing mode
– Variable LGS constellation
• Optimize for narrow and wide field
– Arbitrary tip/tilt star locations and magnitudes
• “Point and shoot” option
– NGS narrow field
83
Real-time Control Architecture
• Massively parallel processor (MPP) implementation is the only
feasible approach
• All aspects of the RTC algorithms have been analyzed and mapped
to MPP implementation
Guide star
height
Wavefront
Wavefront
Sensors
Wavefront
Sensors
Wavefront
Sensors
Sensors
Image
Image
Processors
Image
Processors
Image
Processors
Processors
Centroid algorithm
r0, guidstar brightness,
Guidestar position
Kolmogorov
spectrum
Tomography
Unit
Cn2 profile
Layer heights
Wavefront
Wavefront
Wavefront
Sensors
DM
Sensors
Sensors
Projection
Image
Image
Processors
Image
Processors
DM
Processors
Fit
DM field position
Image
Image
Processors
Image
Processors
Deformable
Processors
Mirrors
Actuator
influence
function
RTC top level of parallelization
84
Reviewer Q & A
•
Start with control block diagrams requested by Fugate.
85
LGS Wide Field Tomography Woofer/Tweeter MOAO
Feeding Deployable Integral Field Spectrographs (d-IFS)
f-fW-fT
f-fW
f
+
Scienc
d-IFS
(6)
Scienc
-
+
LGS
WFS (9)
+
fW
LOWFS
(3)
LOWFS
Tweeter & TT (3)
Tomog
Recon
Tweeter
Tweete
&Tweete
TT (6)
eeInst
Inst
fT
rr
Truth
WFS
Deployable on the field
Woofer &
TT
86
LGS Narrow Field Woofer/Tweeter
Feeding IFU, Vis and NIR Imagers
f-fW-fT
f-fW
f
+
-
+
LGS
WFS (9)
+
fW
Tomog
Recon
Tweeter
& TT
Science
Inst
fT
Truth
WFS
LOWFS
(3)
LOWFS
Tweeter & TT (3)
Woofer &
TT
87
Point and Shoot Option
Feeding IFU, Vis and NIR Imagers
f-fW-fT
f-fW
f
+
-
+
LGS
WFS (6)
-
LGS
WFS (3)
fW
+
Tomog
Recon
Tweeter
& TT
Science
Inst
fT
Truth
WFS
WF
Reconstructor
LOWFS
(3)
LOWFS
Tweeter & TT (3)
Woofer &
TT
88
NGS Narrow Field Woofer/Tweeter
Science
Inst
f
f-fW +
+
-
fW
f-fW-fT
-
fT
NGS
WFS
Tweeter
& TT
WF Reconstructor
Woofer &
TT
89
Brent Ellerbroek's "Big Three" Questions
1. Is the requirement for order 64x64 LGS wavefront sensing realistic,
and how is it driving the design?
•
•
Yes, it is realistic. To be discussed in the performance section.
Drove us to a cascaded relay (or alternatively a split relay).
2. Can the MOAO/MCAO tradeoff be quantified further in terms of the
performance - science - technical risk - programmatic risk - cost
metrics that are defined in the report?
•
See answers in design, performance, risk & cost sections.
3. Is it too ambitious to develop a single AO system design for both the
narrow field and wide(r) field applications?
•
•
•
We initially proposed a single large relay.
Five architectures were evaluated during the system design.
The large relay received a low ranking for several reasons including
size and cost.
90
Q&A: 64x64 Wavefront Sensing
Is 64x64 LGS wavefront sensing realistic, and how is it driving the design?
•
Drives the design in the following ways:
• DM clear aperture sets the 2nd relay beam size
• 64x64 LGS & NGS WFS needed for high Strehl science
• All the LGS WFS will need to accommodate the wavefront spatial
sampling (at least 256x256 assuming a square grid CCD)
• See performance section for more
•
32x32 fall-back impact:
•
•
•
•
Reduces 2nd relay beam size from 20 to 10 mm, making space for
instrument switchyard smaller & mechanically more difficult to design.
Relaxes requirements on number of LGS & NGS CCD pixels
DM less costly
WFS risks: 240x240 PN sensor available. Baseline 256x256 CCID-56
under development.
91
Q&A: Addressing MEMS Risk
4K Device (engineering grade) has been delivered to GPI
DM will be mounted
on a tip/tilt stage
Mount design
concepts under
consideration
92
Cooling the Boston Micromachines MEMS
does not appear to be a problem
•
Source: Steve Cornelisson, Boston Micromachines
•
Deflection vs. voltage is temperature independent, +20 C to -30 C
•
For Nusil adhesive, no temperature effect on rms surface figure, +24 C to 30 C
93
Q&A: MOAO vs MCAO
Q: MOAO/MCAO trade study not adequately quantified. MCAO should
be considered further, particularly if (i) fold mirror at an appropriate
conjugate & (ii) order 64x64 wavefront correction not practical due to
laser power. Impact on WFE budget if increased DM projection error
for MCAO traded against MOAO implementation errors?
• KAON 499 multi-parameter scoring system
used instead of KAON 452 metrics for the
architecture decision.
• MOAO benefits are pretty clear from Figure 
• Science cases require significantly reducing
the generalized anisoplanatism error, driving an
impractical number of MCAO DMs.
• DM projection & open loop go-to error are
characterized & are significantly smaller (~30
nm for go-to, ~50 nm for implementation  now
verified on-sky)
94
Q&A: MEMS Calibration
Q: What is your plan to calibrate the MEMs control in open loop?
BMC 144 actuator MEMS successfully calibrated at LAO. The open-loop
performance of this same mirror has been confirmed on-sky in the Villages
experiment. A similar approach will be taken with the 32x32 & 64x64 MEMS.
95
Q&A: RTC Challenge
Q: NGAO is clearly a challenging system from the point of view of Real
Time Computer. … concerned by the fact that FPGA does not offer
large flexibility during optimization of the AO system. … new ideas or
new ways to control these systems will significantly evolve … flexibility
in the SW implementation should be considered ...
•
•
•
•
MPP design approach based on breaking down problem into basic key
algorithms & allowing a maximum of flexibility in combining building blocks.
Design will allow either of the presently proven stable LTAO algorithms:
Fourier-Domain Pre-conditioned Conjugate Gradient Back Projection
Tomography, and V-cycle Multi-grid Spatial Domain.
Design will allow full flexibility in number of modeled atmospheric layers,
number of subapertures in wavefront sensors, number of DMs, DM
architecture (MOAO or MCAO), a-priori Cn2 model, asynchronous WFS frame
rates, etc.
The needed compute power scales with "problem size" (e.g. number of layers
or number of subapertures) but the MPP architecture can track this with
additional identical FPGA-populated boards & maintain the overall system
throughput rate.
96
NGAO SDR Agenda
10:00
10:05
11:15
11:30
12:00
12:30
13:30
14:00
14:45
15:15
15:45
16:00
16:40
Introduction & Presentation Approach (Wizinowich)
Science Cases & Science Requirements (Max)
Break
Requirements (Wizinowich)
Design (Gavel)
Lunch
Design Q&A (Gavel)
Performance Budgets (Dekany)
Project Management (Wizinowich)
Risks (Wizinowich)
Break
Cost Estimate (Dekany)
PD Schedule & Budget (Wizinowich)
+ Phased Implementation
17:20 Conclusion (Wizinowich)
17:30 General Discussion & Questions (Hubin et al.)
18:00
End
SCRD
SRD,FRD
SDM
SDM
SDM
SEMP
Risk KAONs
SEMP
SEMP
97
Performance Budgets
Charge 2&3: Performance
•
Charge 2: “Assess … the completeness & consistency of the technical
requirements.”
– “Are the performance & error budgets complete & consistent with the
science requirements?”
•
Charge 3: “Evaluate the conceptual design for technical feasibility &
risk, & assess how well it meets the scientific & technical
requirements.”
– “Does the performance predicted for the conceptual design meet the
scientific and technical requirements given in the System Requirements
document?”
– “If the predicted performance of the conceptual design does not meet the
scientific or technical requirements are there adequate plans for
addressing these deficiencies as the project continues?
•
NGAO Team response:
– The NGAO team has developed and is utilizing a capable set of well
anchored error budget and simulation tools to understand and evaluate
performance, and to optimize the design.
•
Additional improvements to these tools are planned for the PD.
– The predicted performance meets the science requirements.
•
But not all the high contrast goals.
– Astrometry performance tools need to be developed during PD.
99
Performance Budgets and Tools
KAON #
Detailed
Detailed
System
System
Simulation
Simulation
471
471
471
471
501
501
497
497
480
480
474
474
---




KAON #
Performance
Budget
Performance
Budget
Residual
wavefront
error
Residual
wavefront
error
Ensquared
energy
Ensquared
energy
Transmission
background
radiation
Transmission
and and
background
radiation
High-contrast
observations
High-contrast observations
Astrometric
precision
Astrometric
precision
Photometric
precision
Photometric precision
Polarimetric precision
Polarimetric precision
Efficient
Efficient
Spreadsheet
Spreadsheet
Tool
Code
Tool
oror
Code



--1
1
2,3
2,3
2,3
2,3
--
11
PDPD
-----
5
5
--
PD
PDPD
--
--
Analytical
Key Design
Analytical
Key Design
Relationships
Drivers
Relationships
Drivers
Documented
Identifed
Documented
Identifed










-- -4
4














Efficiency Budget
Efficiency Budget
Observing efficiency
463
Observing
efficiency
System
uptime
System uptime
463
PD
PD
Overall summary of SD activities
Overall summary of SD activities
Legend
Notes
Notes
PD
PD

x

x
491

x
x PD
--
Legend
--
PD
PD PD
PD
--
491
Mature capability; will continue to evolve as required
Initial capability;
willwill
continue
to develop
in PD
Mature
capability;
continue
to evolve
asphase
required
No current capability; will develop during PD phase
Initial capability; will continue to develop in PD phase
Not planned
No current capability; will develop during PD phase
Not planned
1 Further analysis of high-contrast performance requires coordination with the NIR and visible
coronagraph instruments.
1 Further
analysisvia
of detailed,
high-contrast
performance
requires PSF
coordination
with the
NIR and visible
2 Will support
Monte Carlo
simulation-based
libraries (LAOS,
Arroyo)
coronagraph
3 Used for ainstruments.
posteriori performance verfication; we do not plan to use for requirements flowdown
4 PD
phasevia
support
for Keck
interferometer
only
2 Will
support
detailed,
Monte
Carlo simulation-based
PSF libraries (LAOS, Arroyo)
5
During
the
DD
phase,
we
may
implement
a
prototypical
simulation
AOfor
observing
sequences.
3 Used for a posteriori performance verfication; we do not plan to of
use
requirements
flowdown
4 PD phase support for Keck interferometer only
5 During the DD phase, we may implement a prototypical simulation of AO observing sequences.
100
NGAO Performance
Observation
TT
reference
LGS
asterism
dia.
TT
error
(mas)
Sky
Coverage
HO
WFE
(nm)
Eff.
WFE
(nm)
H
Strehl
/ EE
K
Strehl
/ EE
Io
Science
Target
NGS
2.7
NGS
104
112
83%
90%
KBO Companion
Survey
Field
Star
11”
4.7
10%
154
175
64%
78%
Exo-Jupiters
w/ LGS
Science
Target
11”
2.4
N/A
152
157
69%
82%
Galaxy /
Galaxy Lensing
Field
Star
11”
9.5
30%
159
226
47%
66%
High-Redshift
Galaxies
Field
Star
51”
9.3
30%
204
257
55%
63%
Galactic Center
IRS 7
11”
3.0
N/A
177
184
61%
76%
All but 1 case assume 100W
High-redshift galaxies 150W
101
Reviewer Q & A
102
Q&A: WFE Budget Tool
•
WFE budget tool treats a wide variety of physical effects at appropriate
levels of detail for our major design decisions. It includes:
– Estimates based on first principles (e.g. DM fitting error)
– Estimates based on real optical measurements
•
e.g. static & dynamic telescope errors
– Estimates based on parametric models grounded in more detailed standalone numerical codes (Monte Carlo simulations)
•
e.g., background model, LGS tomography (3 independent codes compared),
LOWFS architecture & sky coverage
– Key interactions between systems
•
•
e.g. LOWFS NGS sharpening, LGS WFS degradation by Rayleigh backscatter
WFE budget tool has been anchored against:
– Independent Keck AO WFE budget
– On-sky NGS & LGS performance of Keck & Palomar AO systems
103
Q&A: LGS WFS CCD noise has moderate performance impact
(in part due to high SNR for good WF measurement)
Gal / Gal Lensing Performance vs. LGS WFS CCD noise
(for 100W, 3+3 LGS WFS, 64x64 subaps, 4x4 pixels/subap, simple
1800
1600
250
1400
200
1200
1000
170 nm requirement
150
800
100
600
400
50
Optimal LGS WFS Frame
Rate [Hz]
Total equivalent WFE [nm]
300
High-order WFE
Total WFE
Optimal frame rate
200
0
0
0
2
4
6
8
RON [e-, rms] @ optimal rate
10
12
These curves are for read noise
Independent of frame rate - we
usually link these through a
detailed CCD noise model
Benefit of optimal centroiding algorithms not shown
105
Q&A: IR LOWFS noise has modest performance impact
(in part due to flexure, other TT error terms)
Gal / Gal Lensing Performance vs. LOWFS read noise
12
700
10
600
500
8
400
6
300
4
200
2
100
0
0
0
2
4
6
8
10
12
14
Optimal LO WFS Frame
Rate [Hz]
Total TT Error [mas]
(for 100W, 3+3 LGS WFS, 2 TT + 1 TTFA, 2x2 pixels/subap, 30% sky, simple centroiding)
TT Error [mas]
Optimal frame rate
16
RON [e-, rms] @ optimal rate
For this science case, 15 mas TT requirement met across range
of LOWFS read noises
(Tightest requirement (Gal Center) also has brightest IR TT star) 106
Q&A: Number of Subapertures / Rates
“150W… seems inadequate… for up to 9… beacons, 17cm subapertures, and up
to 2000 Hz frame rates.”
– Order 64x64 wavefront sensing/correction drives complexity of many other
systems…
•
True, but this combination of parameters is not an optimal performance point
– Optimum WFE is found at N = 64 & 1055 Hz
• For half this sodium return, optimum is N = 58 (19 cm) & 908 Hz
•
•
Design includes selectable pupil samples, N=16, 32, and 64
Design drivers / Rationale
– N=64 correction intended to reduce telescope fitting error and to provide a large
dark hole and fine control of residual PSF speckles
– N=64 WFS is optimal for bright NGS, less so for LGS (but N.B. future uplink AO)
– 2,000 Hz frame rate needed for bright NGS, high-contrast (where latency speckles
are more insidious then noise speckles), and outlying Greenwood frequencies
(where we accept larger WFE)
107
Keck Wavefront Error Budget Summary
Mode:
NGAO LGS
Instrument:
TBD
Sci. Observation:
KBO
Version 1.35
m
m
/D (mas)
Wavefront
Error (rms)
Science High-order Errors (LGS Mode)
Atmospheric Fitting Error
Bandwidth Error
High-order Measurement Error
LGS Tomography Error
Asterism Deformation Error
Multispectral Error
Scintillation Error
WFS Scintillation Error
48
30
107
37
22
22
13
10
nm
nm
nm
nm
nm
nm
nm
nm
43
17
25
40
15
38
34
0
13
16
1
30
30
15
15
nm
nm
nm
nm
nm
nm
nm
nm
nm
nm
nm
nm
nm
nm
nm
Parameter
64
100
150
9
0.50
30
0.34
Alloc
Subaps
Hz (-3db)
W
beacon(s)
m LLT
zenith angle, H band
Scint index, H-band
131 nm
Uncorrectable Static Telescope Aberrations
Uncorrectable Dynamic Telescope Aberrations
Static WFS Zero-point Calibration Error
Dynamic WFS Zero-point Calibration Error
Leaky Integrator Zero-point Calibration Error
Go-to Control Errors
Residual Na Layer Focus Change
DM Finite Stroke Errors
DM Hysteresis
High-Order Aliasing Error
DM Drive Digitization
Uncorrectable AO System Aberrations
Uncorrectable Instrument Aberrations
DM-to-lenslet Misregistration
DM-to-lenslet Pupil Scale Error
64 Acts
Dekens Ph.D
Alloc
Alloc
Alloc
Alloc
30 m/s Na layer vel
4.0 um P-P stroke
from TMT
64 Subaps
16 bits
Alloc
TBD Instrument
Alloc
Alloc
99 nm
Angular Anisoplanatism Error
Total High Order Wavefront Error
23 nm
164 nm
166 nm
1.5 arcsec
High Order Strehl
Error budget corresponding to reviewer’s questioned scenario:
108
150W SOR, 9 beacons, 2,000 Hz frame rate, N=64 WFS’ing, 4e9 Na
NGAO Major Error Terms
Depending on Subaperture Sampling
(for KBO science case, r0 @ 30 zen =14.7 cm, 100W SOR power, median LGS spot size ~1.71",
3 Sci Ast + 3 PnS Ast WFS's @ fixed 1000 Hz)
120
rms WFE [nm]
100
80
Telescope Fitting Error
Measurement Error
60
Atm Fitting Error
RSS Sum
40
20
0
0
10
20
30
40
50
60
70
80
Number of Actuators Across Telescope Diameter
109
Q&A: MOAO PnS vs. MCAO for TT sharpening
• “What is the quantitative impact on sky coverage if the PnS lasers
are eliminated from the system?”
– “What is the impact if MOAO is replaced with MCAO?”
• Both questions are directed to the benefits of superior sharpening of
IR WFS NGS’s
• Assumptions
– NGAO IR WFS NGS MOAO PnS sharpening model
• Interior to the Science Asterism
– No anisoplanatism, constant tomography error
• Exterior to the Science Asterism
– Tomography error transitions smoothly to single-LGS focal anisoplanatism error
– NGAO IR WFS NGS MCAO sharpening model
• Interior to the Science Asterism (which must be expanded to increase NGS sharpening)
– No anisoplanatism, increased constant tomography error
• Exterior to the Science Asterism
– Fall off as normal single-conjugate anisoplanatism:  ~ ((-sci_aster) / 0_eff)5/6
110
NGS Sharpening Model
(for KBO science case, r0 @ 30 zen =14.7 cm, 100W SOR power, median LGS spot size ~1.71",
N=32 actuators, 38 nm Go-To errors (MOAO), 1060 Hz,
3 Sci Ast 3 on 5" radius + 3 PnS Ast at NGS (MOAO) OR "5+1" Sci Ast on 21" radius (MCAO))
0.60
J-band Strehl Ratio
0.50
0.40
MOAO
MCAO
0.30
0.20
Extending sci asterism further pushes ‘roll-off’ edge
outward at cost of greater tomography error in the
science direction
0.10
Going from 5 to 21” in this example increased sci
direction tomography error from 54 nm to 70 nm
0.00
0
10
20
30
40
50
60
70
Off-axis distance [arcsec]
111
Q&A: MOAO PnS vs. MCAO for TT sharpening
Design Drivers / Rationale
• Sky coverage is important for NGAO science (KAON 455)
• Tip/tilt error is a concern at Keck (N.B. conservative NGAO wind shake model)
– Experience shows K2 LGS FWHM often not diffraction-limited (KAON 489)
• d-IFS performance benefits from variable radius asterism (KAON 427)
• Quantitative benefit for Galaxy/Galaxy Lensing Science Case
– MOAO PnS provides ~3x higher J Strehl for NGS at distance of ~50”
– Higher TT Strehl reduces intrinsic TT error from 8.3 to 6.0 mas
• This is equivalent to an improvement in science H-Strehl of about 10% absolute
– Evaluation across other science cases still needed (PD phase)
112
Q&A: MOAO PnS vs. MCAO for TT sharpening
Design Drivers / Rationale (cont.)
• Better-corrected NGS PSF’s are operationally easier to handle
– More likely to obtain a diffraction-limited PSF core across a variety of
environmental conditions
– More consistent core expected to improve acquisition efficiency
• Not all science targets follow Spagna statistics
– c.f. GOODS-N/S, HDF N/S, Chandra DFS, Lockman Hole, COSMOS field
– Design for wider NGS field of regard than indicated by Spagna average
values
• Cost / Benefit of PnS architecture (vs. scalable but fixed geometry
asterism) is subject of a preliminary design study
113
Q&A: Sodium Layer Photoreturn
• Basic assumption (based on published SOR measured return)
– 150 ph/cm2/s/W
• Explicit assumptions
– All lasers contribution to science wavefront calculation
• Worst case Point-and-Shoot laser is ~75” off-axis. The “10% metapupil shear
height” at this angle is ~3 km, which is above ~80% of the turbulence in the MK
Ridge model. The claim therefore that the PnS LGS sample most of the save
volume as the science asterism, so their photons also count.
– During the PD phase, this will be confirmed using detailed LAOS simulations.
– Transmission(s)
• Up: LGSF 0.75 x Atm30 up 0.78 = 0.59
• Down: Atm30 down 0.78 x Telescope 0.61 x NGAOHOWFS 0.37 X HOWFS QE589
0.80 = 0.14
• Implicit assumptions
– SOR technology (or its equivalent) can be made available to NGAO
– NGAO lasers will be backpumped and return will be invariant across
different magnetic field lines
114
NGAO performance is robust to fluctuations in laser photoreturn
(from Na column density, laser return per W, or laser power)
NGAO Performance vs. Photoreturn
0.75
0.70
0.65
0.60
H-Strehl
H Strehl
0.55
N
N
N
N
0.50
0.45
=
=
=
=
64
32
64
32
KBO
KBO
Gal Gal Lens
Gal Gal Lens
0.40
For only 50W laser power,
performance pivots around relative
photoreturn = 0.5; not acceptable
0.35
0.30
0.25
0
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Relative photoreturn
(1 = baseline; 150 ph/cm2/s/W, 100W, 4e9 cm-2 Na)
At each data point, the frame rate that minimized WFE is used
(ranging from 425 fps to 1886 fps)
115
NGAO SDR Agenda
10:00
10:05
11:15
11:30
12:00
12:30
13:30
14:00
14:45
15:15
15:45
16:00
16:40
Introduction & Presentation Approach (Wizinowich)
Science Cases & Science Requirements (Max)
Break
Requirements (Wizinowich)
Design (Gavel)
Lunch
Design Q&A (Gavel)
Performance Budgets (Dekany)
Project Management (Wizinowich)
Risks (Wizinowich)
Break
Cost Estimate (Dekany)
PD Schedule & Budget (Wizinowich)
+ Phased Implementation
17:20 Conclusion (Wizinowich)
17:30 General Discussion & Questions (Hubin et al.)
18:00
End
SCRD
SRD,FRD
SDM
SDM
SDM
SEMP
Risk KAONs
SEMP
SEMP
116
Project Management
Charge 5: Project Management
•
Charge 5: Evaluate the suitability & effectiveness of the project
management, organization, decision making …, with an emphasis on
the next project phase (preliminary design) and also with respect to the
entire project.
– Does the performance of the project to date support the project’s approach
to management & decision making?
– Is the project’s proposed approach to management & decision making
likely to succeed? What modifications would be advantageous to assure
the success of the entire project?
•
NGAO Team response:
– SDR deliverables complete.
– 4% schedule & budget overruns (3 weeks & $50k).
– ~7% of planned work postponed or cancelled (equivalent to ~$70k).
– ~$120k used for higher than planned salary rates.
– Improved management/org structure defined for next phase.
– Structure will be further strengthened for detailed design
118
System Design Actuals vs Plan
FY07
FY08
(to 2/29)
FY08
Remain
Total
Plan
Plan –
Total
COO
261.6
72.1
20.9
354.6
314.9
-39.7
UCO
144.0
92.6
11.9
248.5
238.1
-10.4
WMKO
327.1
195.3
80.9
603.3
438.6
-164.7
6.2
7.0
0.0
13.2
57.4
44.2
Institution
Students
Contingency
103.9
Inflation
16.7
Total ($k) =
738.9
Plan ($k) =
818
351.6
1169.6
-79.1
129.2
50.0
Actual - Plan =
MS Project Schedule:
• SDR delayed from
3/31 to 4/21/08
• 91% work complete
thru SDR
367.0
113.7
1219.6
1169.6
-50.0
~$120k
impact
Institution
Plan
(hours)
Actuals
(to 2/29)
Actual
- Plan
Actual
Rate ($/hr)
Plan Rate
($/hr)
COO
3369
3581
212
85.29
87.56
UCO
3154
2651
-503
88.10
69.11
WMKO
7276
7539
263
65.80
56.21
Total (hrs) =
13799
13771
-28
119
Organization Structure
120
Decision Making
Organization structure assigns responsibilities/authority:
•
Clear requirements & interfaces should facilitate localized technical decision making.
•
Systems engineering flows down requirements, interface definitions & architecture, &
evaluates changes.
•
PM makes or delegates project decisions, in consultation with senior NGAO management.
•
Direction & consultation on major changes (funding, cost, schedule or requirements) &
priorities (schedule, budget &/or requirements) will be sought from WMKO Directorate.
•
Science consultation with SSC as needed.
Decision making falls into several categories:
•
Configuration control
–
–
–
•
Risks
–
•
Risks will be tracked by PM & Systems Eng. Decisions to retire risks early wherever possible.
Build versus Buy
–
–
•
Requirements & interface definitions will be under configuration control immediately, & designs
starting during Detailed Design.
Change Control Board reviews & approves changes to the above starting in Detailed Design.
Science & System Requirement changes additionally approved by Project Scientist & PM.
Need to determine any constraints imposed by the Directors.
Prefer to buy when we can at a level existing vendors have demonstrated their ability to deliver.
Reviews (provide input to the Directors & NGAO team for decision making)
–
–
Project reviews provide an opportunity to review project decisions at key milestone points.
Monthly report, SSC presentations & meetings with the Directorate.
121
Charge 6&7: Project Management
•
•
Charge 6: Provide feedback on whether the overall strategy will
optimize the delivery of new science.
Charge 7: Gauge the readiness of the project to proceed to the
preliminary design phase.
– Has the project adequately defined the objectives, work breakdown
structure & task plan for the next design phase?
•
NGAO Team response:
– Overall strategy optimizes science delivery.
– PD plan well defined.
•
Further improvements could be made starting with DD.
122
Work Breakdown Structure
123
Integration
& Test
+ Reviews
Milestones
22 mos
SDR
24 mos
18 months
6 mos
PDR
124
Integration
& Test
+ Reviews
AO Unavailable (12 mos)
Milestones
6 mos
6 mos
4 mos
125
Reviewer Q & A
126
Technical & Programmatic Risks
& Risk Mitigation
Charge 3&5: Risk
• Charge 3: “Evaluate the conceptual design for technical
feasibility & risk, & assess how well it meets the scientific
& technical requirements.”
• Charge 5: Evaluate the suitability & effectiveness of the
project management, organization, decision making & risk
mitigation approaches, with an emphasis on the next
project phase (preliminary design) and also with respect to
the entire project.
• NGAO Team response:
– The technical & programmatic risks have been identified and
ranked, and will be tracked.
– Risk mitigation approaches have been identified.
• Risk mitigation during PD funding limited.
128
Technical
Programmatic
Risk Overview
Likelihood
4
3
20
2
4
1,2
11,12
5-8
3
21-22 13-19 9,10
1
1
24
23
2
3
4
Consequences
4
5,6
3
10
7-9
3,4
1,2
2
1
5
1. Inadequate PSF calibration
2. Inadequate sky coverage
3. Required lasers unavailable
4.
5.
6.
7.
8.
9.
10.
5
Likelihood
5
WFE budget assumptions
Inadequate tomographic reconstruction
Astrometric performance
Tomographic computer HW architecture
Keck Interferometer needs
SW control complexity & instability
CCD availability
1
2
3
4
Consequences
1.
2.
3.
4.
NGAO funding
Required lasers unavailable
Rapid project ramp up
Growth in cost estimate
5.
6.
7.
8.
9.
Lack of full-time personnel
Management structure
Science instruments schedule
Funding schedule impact
Contract schedule slips
129
5
Technical Risk Evaluation
1. Inadequate PSF calibration
• Most impact on Galactic Center GR & narrow-field proper motion
astrometry, & detection of planets around low mass stars.
a) Collaborate on CfAO-funded PSF reconstruction effort
b) Produce a system-level design for PSF calibration
c) Investigate Mauna Kea atmospheric profiler collaboration
2. Inadequate sky coverage to support wavefront error budget
• AO corrected low order wavefront sensing using low noise NIR
detectors assumed.
a) NIR TT sensor to demo detector & AO correction. Investigate IRIS
LOWFS study collaboration
b) Lab &/or on-sky demo during DD.
3. Required lasers unavailable
• Sodium return inadequate. See programmatics risks.
130
Risk Mitigation with ongoing experiments at
LAO/Mt Hamilton with ViLLaGEs*
• Objectives
1.
2.
3.
Test MEMS deformable mirrors on-sky
Open-Loop AO
Uplink-corrected laser projection
• Implementation
–
–
–
AO system on Nickel 40” telescope
140-DOF BMC micro deformable mirror
CCD-39 WFS with simultaneous wavefront measurements
1. Uncorrected (Open loop) wavefront
2. Closed loop residual corrected wavefront
3. Sharpened tip/tilt star
*Visible Light Laser Guidestar Experiments
131
Open loop control requires
•
Predictable response of the DM
•
Absolute measurement of the wavefront with high
dynamic range
132
Villages open-loop
on-sky AO results
Strehl vs wavelength
On-sky Data and Prediction Using Error
Allocation Model
0.2
0.18
0.16
1”
[Montage of open-loop controlled
images in V, R, I, 900nm bands]
Strehl
0.14
0.12
0.1
0.08
0.06
0.04
900/40 nm
0.02
I band (800 nm)
0
500
600
700
800
900
1000
R band (600 nm)
Uncorrected
V band (500 nm)
wavefront error, nm
wavelength, nm
70
Error Allocation
60
50
40
30
20
10
0
Internal Calibrator
(650 nm)
133
Rejection Ratio of Temporal Power in the
Wavefront Residuals
Closed Loop
Open Loop
135
Open vs Closed Loop Conclusions
•
•
Open loop control has higher bandwidth
Closed loop control does better at low frequencies (consequence of
open loop cancelation accuracy)
• On sky, we noticed no difference in Strehl performance
between open-loop and closed-loop operation
Take-Home Points
• We now have on-sky experience with MEMS DMs
• Open-loop control appears to be working in a way that is
consistent with NGAO error budget models
136
Programmatic Risk Evaluation
1. NGAO funding
• Not a review item
2. Required lasers unavailable &/or too expensive
• Business model for affordable SOR or LMCT-type lasers?
a) Determine best laser availability solution
b) Evaluate impact of procuring less laser power
3. Rapid project ramp-up
• Rapid personnel ramp-up required after preliminary design
a) Produce a viable ramp up plan
b) Find funds to allow more people to be involved earlier
4. Cost estimate growth
a) Identify & exploit cost savings opportunities
b) Employ a design to cost approach
137
Laser Systems
• NGAO laser procurement issues
– Single frequency CW laser appears to produce the most return per watt
– All known systems require engineering development to be suitable for
NGAO deployment
– TMT also needs lasers, it appears NGAO requirements can be
harmonized with TMT requirements
– This creates a more realistic commercial opportunity for vendors
• NGAO laser procurement strategy
– Develop harmonized requirements by collaborating with TMT
– Issue an RFP for laser system development
• Start as early as possible with contract for laser engineering (detailed
design) phase
– Procure lasers in some way that allows vendor to build 6 x 50 watt, (or
12 x 25 watt) over some reasonable period for both NGAO and TMT
– Ensure reliable source of spare parts and long term support
138
Reviewer Q & A
139
NGAO SDR Agenda
10:00
10:05
11:15
11:30
12:00
12:30
13:30
14:00
14:45
15:15
15:45
16:00
16:40
Introduction & Presentation Approach (Wizinowich)
Science Cases & Science Requirements (Max)
Break
Requirements (Wizinowich)
Design (Gavel)
Lunch
Design Q&A (Gavel)
Performance Budgets (Dekany)
Project Management (Wizinowich)
Risks (Wizinowich)
Break
Cost Estimate (Dekany)
PD Schedule & Budget (Wizinowich)
+ Phased Implementation
17:20 Conclusion (Wizinowich)
17:30 General Discussion & Questions (Hubin et al.)
18:00
End
SCRD
SRD,FRD
SDM
SDM
SDM
SEMP
Risk KAONs
SEMP
SEMP
140
Cost Estimation
Charge 4: Cost Estimate
• Charge 4: Assess whether the design can be implemented
within the proposed schedule & budget.
– Are the plans for completion of the project, including the cost
estimate, schedule & budget to completion, sufficiently detailed?
– Is the methodology used to develop the cost estimates sound?
– Is the proposed budget to completion realistic?
– Is there sufficient management reserve (contingency) allocated in
the proposed budget to completion?
• NGAO Team response:
– A detailed and realistic bottoms-up cost estimate has been
prepared for the completion of the project and for each major
phase, including identification of contingency.
142
Cost Estimation Methodology (KAON 546)
• Cost estimation spreadsheets
– Based on TMT Cost Book approach, simplified for SD phase
– Prepared for each WBS element (~75 in all)
– Prepared for each of 4 phases
• Preliminary design, detailed design, full scale development,
delivery/commissioning
– Prepared by technical experts responsible for deliverables
– Process captures
•
•
•
•
•
•
•
WBS dictionary
Major deliverables
Estimates of labor hours
Estimates of non-labor dollars (incl. tax & shipping) & travel dollars
Basis of estimate (e.g. vendor quote, CER, engineering judgment)
Contingency risk factors & estimates
Descope options
– Standard labor classes, labor rates & travel costs used
143
Cost Estimate to Completion (FY08 $k)
WBS
WBS Title
PD
DD
FSD
D&C
Base
Cost
Contingency
Total
($k)
2
Management
874
1,232
1,594
657
4,356
318
4,674
11%
3
Systems Eng
811
1,004
478
193
2,485
401
2,886
7%
4
AO System Dev
730
2,208
9,742
3
12,683
3,849
16,533
39%
5
Laser System Dev
285
1,947
6,619
128
8,980
1,935
10,915
26%
6
Science Operations
166
756
646
1,568
233
1,801
4%
7
Tel. & Summit Eng.
95
424
1,049
19
1,587
344
1,932
5%
8
Telescope I&T
46
106
114
1,944
2,211
525
2,735
6%
9
Ops Transition
14
20
555
70
660
91
750
2%
3,021
7,697
Sub-Totals ($k)
20,797 3,015 34,530
7,697
42,227 100%
144
Cost Estimate to Completion (FY08 $k)
Cost Estimate (FY08 $k)
Labor
(PY)
Labor
NonLabor
Travel
SubTotal
Contin
-gency
Total
% of
NGAO
Budget
Preliminary Design
21.0
2,582
216
224
3,022
458
3,479
8%
Detailed Design
43.6
5,516
1,827
354
7,697
1,403
9,100
22%
Full Scale Develop
50.5
5,661
14,510
626
20,797
5,234
26,031
62%
Delivery/Commission
22.4
2,287
250
478
3,015
602
3,617
9%
Total =
138
16,045
16,804
1,681
34,531
7,697
42,227
100%
46%
49%
5%
100%
22%
122%
Phase
% =
145
Reviewer Q & A
146
Q&A: Cost Impact of MOAO/MCAO
for TT Sharpening
• “What is the impact of the MCAO/MOAO tradeoff on the cost
estimate, particularly if the order of correction were scaled back to
32x32 or 20x20?”
• Cost increments for MCAO vs. MOAO for purpose of NGS sharpening
– Savings ~$2,100K?
• Remove three LOWFS 32x32 MEMS: ~$500k hardware
• Reduce RTC requirements: ~$600k hardware
• Reduce LOWFS Assembly, I&T, RTC & Commissioning labor: ~$1,000k?
– Increases ~$1,700K - $2,600K?
• 20x20 or 32x32 9 km conjugate DM: $300k or $1200k.
• Increase RTC requirements: ~$400k hardware(?)
• Increase Optical Relay, I&T, RTC, Commissioning labor: ~$1,000k?
147
Q&A: Cost Impact of MOAO/MCAO
for Science
•
Achieving science goals with MCAO requires somewhat different approach
(KAON 452)
•
N = 64 x 64 actuators for narrow-field science path (KAON 499)
– DM with appropriate pitch and large stroke unavailable
– Result drives architecture toward our MCAO “Large Relay” option
•
Incremental cost comparison during architecture downselect indicated that
Large Relay was ~$2.6M greater cost than Cascaded Relay
•
MOAO would likely still be required for dIFS (KAON 471)
– 2’ circular field of regard suffers from generalized anisoplanatism, not present in
MOAO architecture
• Details have not been quantified, but using MOAO we barely meet the dIFS performance
requirements
• MOAO could be avoidable over ~60” field of regard
– d-IFS impact was not considered in earlier cost comparison
148
Schedule & Budget
Charge 4&7: Schedule, Budget & Resources
• Charge 4: Assess whether the design can be implemented
within the proposed schedule & budget.
• Charge 7: Gauge the readiness of the project to proceed to
the preliminary design phase.
– Are the resources identified for the next design phase sufficient to
address the scope of work?
• NGAO Team response:
– This section will only address the PD phase (the rest of the project
has been addressed in earlier sections)
– A realistic budget and schedule has been produced for the PD
phase.
– Excellent team for PD. Resources are sufficient.
• Will need more full-time effort starting with DD.
150
Preliminary Design Phase Schedule
•
•
•
PD phase tasks & hours from cost estimation entered in MS Project.
Personnel assigned to tasks.
Tasks scheduled in required sequence & in order to fit within FY
available budgets.
151
Preliminary Design Budget (FY08 $k)
Work (hours)
Institution
Cost ($k)
FY08
FY09
FY10
Total
FY08
FY09
FY10
COO
1116
4360
927
6403
107
419
88
614
UCO
1719
6407
1675
9801
113
444
118
675
WMKO
2542
11633
3030
17204
196
841
228
1264
Free (Max + WMKO)
292
1068
203
1563
0
0
0
0
Student/Postdoc
227
933
0
1160
9
37
0
46
5895
24401
5835
36131
425
1741
434
2600
Procurements ($k)
2
164
50
216
Travel ($k)
28
125
61
214
30
289
111
430
0
0
449
449
Total ($k) =
455
2030
994
3479
Available ($k) =
455
2000
1024
3479
0
-30
30
Labor Total =
Labor & Non-Labor Total ($k) =
Contingency ($k)
Available - Total ($k) =
Total
0
152
Preliminary Design Core Team
Name
EC
Fulltime
Inst
Role
%
Adkins, Sean
WMKO
Laser procurement, instrument interfaces
26
Bell, Jim
WMKO
AO enclosure & infrastructure
23
Britton, Matthew
COO
Wavefront sensor design, performance budgets
24
Dekany, Rich
COO
COO project management, systems engineering
52
EE / Programmer
UCO
Real-time control
78
Gavel, Don
UCO
UCO project management, technical overview
37
Non-real time controls & software, systems eng
89
AO optical design
25
Science operations tools, operations concept
93
Johansson, Erik
WMKO
Kupke, Renate
UCO
Le Mignant, David
WMKO
Lockwood, Chris
UCO
AO mechanical design
36
Max, Claire
UCO
Project Scientist, science requirements development
35
McGrath, Elizabeth
UCO
Postdoc for Project Scientist, science development
100
Morrison, Doug
WMKO
Non-real time control software
25
Neyman, Chris
WMKO
Systems engineering, laser & AO facility design
84
Wavefront sensor design
58
Velur, Viswa
COO
Wetherell, Ed
WMKO
Non-real time control electronics
34
Wizinowich, Peter
WMKO
Project manager, technical overview
59
Wavefront sensor design
28
Zolkower, Jeff
COO
153
Reviewer Q & A
154
Phased Implementation
Charge 6: Staged Implementation/Descopes
•
Charge 6: Provide feedback on whether the overall strategy will
optimize the delivery of new science.
– Are there possibilities for staged implementation or descopes that
are viable in terms of the science requirements?
•
Disclaimer: Not a System Design phase deliverable.
– Some initial thoughts presented here.
– Options can be considered during Preliminary Design.
– Staged implementation will be more expensive, but allows science return
as funding available.
156
One Phased Approach Option
Phase 1: Laser tomography
• 50W laser, fewer LGS
• fewer LGS WFS & LOWFS
• no MEMS or MOAO control
• no new instruments
• uncooled enclosure
 Higher SR from laser power &
reduced focal anisoplanatism
Add back based on science priority
RTC
157
Science Cases vs AO Capabilities & Instruments
Science Case
Galaxy Assly & Star Formation
Nearby AGNs
Galactic Center - Relativity
Galactic Center - Stellar Pop'ns
Extrasolar Planets
Minor Planet Multiplicity
QSO Host Galaxies
Gravitational Lensing by Galaxies
Gravitational Lensing by Clusters
Astrometry Science
Resolved Stellar Populations
Debris Disks
YSOs
Minor Planet Size, Shape, Comp.
Gas Giant Planets
Ice Giant Planets
PSF High Low
High Stab- Sky BackStrehl ility Cover. gnd d-IFS
NIR
IFU
NIR
Cam
Vis
Cam
narrow
narrow
narrow
Helps illustrate the science impact of deferrals/descopes
158
Reviewer Q & A
159
Review Panel Report Questions
1.
Assess the impact of the science cases in terms of the competitive landscape in which the
system will be deployed.
– Science section
2.
Assess the maturity of the science cases & science requirements and the completeness &
consistency of the technical requirements.
– Science, Requirements & Performance sections
3.
Evaluate the conceptual design for technical feasibility & risk, & assess how well it meets
the scientific & technical requirements.
– Design, Performance & Risk sections
4.
Assess whether the design can be implemented within the proposed schedule & budget.
– Cost & PD Schedule & Budget sections
5.
Evaluate the suitability & effectiveness of the project management, organization, decision
making & risk mitigation approaches, with an emphasis on the next project phase
(preliminary design) and also with respect to the entire project.
– Project Management & Risk sections
6.
Provide feedback on whether the overall strategy will optimize the delivery of new science.
– Project Management & Phased Implementation sections
7.
Gauge the readiness of the project to proceed to the preliminary design phase.
– Design, Project Management & PD Schedule & Budget sections
160
Conclusion
•
•
•
•
•
•
•
•
•
NGAO will provide the WMKO community with a powerful & unique
scientific capability.
The requirements are well understood & have been flowed down to
the design.
A feasible and optimal conceptual design has been developed.
The technical and programmatic risks are well understood.
A solid cost estimate has been developed.
The management structure, plan and personnel are in place for the
preliminary design.
The very significant scientific rewards offered by NGAO come with
significant technical & programmatic challenges.
We have a team capable of addressing these challenges and
delivering this powerful new scientific capability.
We are ready, willing & able to proceed with the preliminary design.
161
NGAO SDR Agenda
10:00
10:05
11:15
11:30
12:00
12:30
13:30
14:00
14:45
15:15
15:45
16:00
16:40
Introduction & Presentation Approach (Wizinowich)
Science Cases & Science Requirements (Max)
Break
Requirements (Wizinowich)
Design (Gavel)
Lunch
Design Q&A (Gavel)
Performance Budgets (Dekany)
Project Management (Wizinowich)
Risks (Wizinowich)
Break
Cost Estimate (Dekany)
PD Schedule & Budget (Wizinowich)
+ Phased Implementation
17:20 Conclusion (Wizinowich)
17:30 General Discussion & Questions (Hubin et al.)
18:00
End
SCRD
SRD,FRD
SDM
SDM
SDM
SEMP
Risk KAONs
SEMP
SEMP
162