Transcript Recent GLAO experiments on the UH88”

Ground Layer Adaptive Optics (GLAO)
Experiment on Mauna Kea
Doug Toomey
February 2013
The Imaka
Project
This talk is about an experiment in support of a larger project called
IMAKA



IMAKA is a project to build a wide field imager for large telescopes
that has improved image quality by using a type of adaptive optics
called Ground Layer Adaptive Optics (GLAO)
Adaptive optics involves the correction of atmospheric induced optical
aberrations (seeing) using electronically controlled deformable mirrors
IMAKA achieves corrections over much larger fields than present
instruments by only fixing atmospheric aberrations close to the
ground
Imaka Science




Examples of the projects this is useful for are:
Galaxy Formation and Evolution
Finding and Studying Kuiper Belt Objects
Resolving Stellar Populations of Nearby Galaxies
Traditional
Adaptive optics
Traditional adaptive optics systems
are limited in field of view because
as the field increases the star light
travels through different paths in the
upper atmosphere.
This limits corrected fields to about
1 arcminute
Ground Layer
Adaptive optics
Experiments on Mauna Kea in
Hawaii have shown that the
turbulence is primarily found in two
layers. One at or near the ground
and one in the upper atmosphere.
Ground Layer
Adaptive optics
By just correcting this lower layer a
much larger field can be corrected
and still produce a useful
improvement reducing the image
size from 0.5 to 1.25 arcseconds
down to 0.3 to 0.4 arcseconds
The ground turbulence aberrations
and the telescope dome induced
aberrations are removed.
Proof
`Imaka has the goal to reach the “freeatmosphere” seeing over large fields of view
(e.g. 10’s of arcminutes to a degree)
How could we prove that this
technique would work on a large
telescope.

CFHT funded us last year to
perform an experiment on the
UH 2.2 meter and the 3.6 meter
CFHT telescopes, to try to
verify, on-the-sky, the level of
performance we can achieve.

Do we really see large
correlations of the wavefronts
over these large angles when
we look thru the telescope?

Tests on the UH 2.2 meter
Our foundation: Site studies with numerous optical turbulence profilers:
SLODAR, LOLAS, LunarShabar, MKAM (MASS/DIMM)
Each of these studies was done “outside” - e.g. not thru one of the big
telescopes.
Question: Is there something fundamentally different arising within the
telescope enclosures?
We started on the UH 2.2 meter telescope since we could get more time
The Experiment

Approach:

Following Baranec (2007) - On Mt. Hopkins, in support of MMT
GLAO, observe a constellation of stars with multiple wavefront
sensors to measure the phase correlations and estimate the GLAO
PSF over a 2’ FOV.
➡ We would use five wavefront sensors, on a constellation of stars
covering 0.5 degree on UH88” (Cassegrain) and 1 deg on CFHT
(prime).
Wavefront sensor
Imaging the telescope
entrance pupil (the primary
mirror) onto a 2d array of
lenses (Lenslet array) turn the
2 meter telescope into 400 9
cm telescopes.
Shack Hartmann wavefront
sensor
Measuring the star position
from each sub-aperture
measures the wavefront tilt in
each sub aperature
Schedule
mWFS
experiment
‣
GL and dome seeing - direct
measure of the correlation of
wavefronts over one degree
✓
UH88: Prototype WFS - July `12
✓
UH88: 5 WFSs 0.5 deg - September
`12
‣ CFHT: 6 WFS 1.0 deg - Dec/Jan `13
mWFS/UH
88
• Using the optics for
the UH8k camera
replacing the focal
plane with our five
WFSs
• First two runs were
in September.
• Observed two
different
constellations
What the data
looks like...
This is a combination of SHWFS spots from three
WFSs. Each WFS is a different color in the image.
What the data

looks like...
When the wavefronts are correlated the spots are white
When the wavefronts are uncorrelated the spots separate
into distinct colors
What can we learn from this
data
What the data tells us...
These are the slope cross-covariance maps...
A layer at the ground
moving with the ground
wind speed?
What the data tells us...

From the data we can extract a number of quantitative measures:

Simplest:




Total phase variance (seeing)
We can extract the vertical profile (where in altitude the seeing
comes from the ground up to about 600 meters)
What portion of the seeing is common to all wavefront sensors (
the GLAO correctable part of the aberations)
An estimate of the image size that a GLAO system could
achieve
amplitude (nm)
What the data tells us...
time step (50Hz)
What the data tells us...
Next
steps...


Working our way thru an incredibly rich data set
Next phase is a GLAO demonstrator on the 88” proposal with NSF
pending
The Experiment Team

Mark Chun (UH) PI

Olivier Lai (CFHT)

Tim Butterley (Durham University)

Doug Toomey (MKIR)

Kevin Ho and Derrick Salmon (CFHT)

Yutaka Hayano/Shin Oya (Subaru) - contributing DM/electronics for demonstrator

Simon Thibault (Lavel University) - optical design of demonstrator

Christoph Baranec (CalTech) - real-time controller from RoboAO