Visible AO at Palomar using a sodium laser beacon

The movies below show dynamical simulations of PSF's at visible wavelengths delivered by the Palomar AO system guiding on a sodium beacon. The Cerro Pachon vertical turbulence model was assumed in these simulations, and two different sets of layer weights in the model were chosen to yield 1.3 arcsecond seeing and .65 arcsecond seeing. Wind speeds in the 6 layer model varied from 5 m/s on the ground to 30 m/s in the troposphere. The simulation performed wavefront sensing at 589 nm using a 16x16 Shack Hartmann system, and drove tip tilt and deformable mirrors to perform adaptive compensation. The simulations assume a 500 Hz control loop rate, which is set by the current Chicago sum frequency laser repetition rate. The resulting on-axis PSF's were computed at 500 and 1000 nm.

The first movie below shows the uncompensated and compensated PSF's evolving in time due to the dynamical atmosphere and adaptive optics system. This movie shows both uncompensated (left) and compensated (right) PSF's at 500 nm (top) and 1000 nm (bottom) computed under 1.3 arcsecod seeing conditions. Each of these PSF's is shown over a 2 arcsecond field of view. In these conditions, a diffraction limited core is clearly visible at 1000 nm, and the seeing halo is reduced at 500 nm. The second movie shows the compensated PSF's at 500 nm (top) and 1000 nm (bottom) under .65 arcsecond seeing conditions. At both wavelengths a diffraction limited core is apparent.

Uncompensated and compensated PSF's in 1.3 arcsecond seeing
Compensated PSF's in .65 arcsecond seeing

Here is a time history of the Strehl ratios for these simulations.

These simulations account for fitting error, servo error, and focal anisoplanatism error. They do not account for optical aberrations in the telescope, AO system, and instrument. Previous experience indicates that these aberrations contribute about 140 nm of wavefront error, and will further degrade the PSF's shown here. Extrapolating the current performance to the visible and adding in the effects of focal anisoplanatism imply a Strehl ratio of about 20% at 1000 nm. This is roughly consistent with adding 140 nm of wavefront error in quadrature to the Strehl ratio derived from simulation.