Performance Projections

Strehl and contrast with ExAO on a 25 m telescope

The Giant Magellan Telescope (GMT) is being constructed at Las Campanas Observatory (LCO). LCO is home to the twin 6.5 m Magellan Telescopes, Baade and Clay. Our team deployed the Magellan Adaptive Optics system (MagAO) to the Clay telescope in 2012, and we have since averaged nearly a month per semester observing with AO at LCO. We are now constructing a new "extreme" AO system for the Clay Telescope, funded by the NSF MRI program, called MagAO-X.

Figure 1. 7x 3000 actuator MEMS DMs. MagAO-X: The key parameters of MagAO-X are: a 2040 actuator MEMS deformable mirror (DM), with 48 actuators across the 6.5 m pupil (13.5 cm pitch); a pyramid wavefront sensor (PWFS) with an electron multiplying CCD (EMCCD) allowing us to operate at 3.6 kHz with very low noise; an integral coronagraph, coronagraphic wavefront control system, and science cameras on the same optical table with the DM and PWFS. The MagAO-X preliminary design review (PDR) is extensively documented here, and publications about the instrument are here.

The GMT consists of 7 circular 8.4 m mirrors. If we scale the 2040 actuator MEMS from a 6.5 m to 8.4 m, we find that a 3000 actuator MEMS will give the same actuator spacing (or pitch). This forms the basis of our concept for GMagAO-X: 7x 3000 actuator MEMS DMs, one per segment, will give the same Strehl ratio on the GMT as we expect to obtain with MagAO-X on the Clay. The cartoon below shows how this can be implemented optically, with a hexagonal prism to disassemble the pupil, and then a 2nd prism to reassemble after the DM segments.

Figure 2. An illustration of how we will implement a 21,000 actuator deformable mirror, providing 13.5 cm actuator pitch across the 25 m primary mirror. The pupil slicer prism forms an image of each segment of the primary mirror on its own 3000 actuator MEMS, and then reassembles the pupil. After two such devices (one for the low-spatial-order "woofer"), the sine condition will be met and true diffraction limited images will be made in the focal plane.


Strehl Ratio

The "Strehl Ratio" is the ratio of the brightness of the peak of an image of a star to the peak of an image of the same star with a perfect telescope (no atmosphere and no optics aberrations). An important thing to remember about AO systems is that Strehl ratio depends on DM actuator count and WFS speed, but not telescope diameter or collecting area. Because GMagAO-X will have the same actuator spacing as MagAO-X and use the same WFS detectors, it will obtain the same Strehl ratios on the same stars.

Below are performance predictions for MagAO-X@Clay from end-to-end simulations. For all of the details see this PDR document. These plots also illustrate the performance we would expect form our scaled-up concept for GMagAO-X on the GMT.

Figure 3. Strehl Ratio vs. guide star magnitude, at 900 nm and H-alpha (656 nm), for MagAO-X on the Magellan Clay telescope at LCO. These wavelengths are used here because they span the key science cases for MagAO-X. 50th percentile conditions are 18.7 m/s winds and 0.62" seeing. 25th percentile are 9.4 m/s winds and 0.5" seeing. Since GMT is at LCO these conditions are (more or less) the same, and since GMagAO-X maintains the same actuator pitch and system speed, we will obtain the same performance at GMT.
Figure 4. Example star images (point spread functions, PSFs) at H-alpha (656 nm), for MagAO-X on the Magellan Clay telescope at LCO. See above figure caption for description of conditions. The main difference on GMT will be a much higher angular resolution, meaning the X and Y axis numbers will scale by 6.5/24.5. There would also be differences in the fine structure of the images due to the aperture structure. Other than these details, these images illustrate the high image quality we expect from GMagAO-X at visible wavelengths.


Contrast

Figure 5. Raw contrast maps for a GMagAO-X-like system in 50th percentile conditions, at 800 nm. The left column shows results for a standard pure integrator (PI) law, and the right column shows results for a linear predictor (LP) control law. AO performance depends on star brightness. We expect to perfect predictive control on current systems (MagAO-X and SCExAO), and will employ it on future systems like GMagaO-X. Figure reproduced from Males & Guyon (2018).
Figure 5. Raw contrast profiles for a GMagAO-X-like system in 50th percentile conditions, at 800 nm. The dashed lines show results for a standard pure integrator (PI) law, and the solid lines are results for a linear predictor (LP) control law. AO performance depends on star brightness. Very deep raw contrasts (better than 1e-5 on an 8th mag star) are possible on a large telescope like GMT. Figure reproduced from Males & Guyon (2018).