An on-line turbulence profiler for the AOF: on-sky results

The ESO's adaptive optics facility (AOF) is ending its commissioning at Paranal (Chile). It feeds two second-generation instruments of the VLT-UT4 telescope, HAWK-I and MUSE, with turbulence corrected wavefronts through the GALACSI and GRAAL modules. The main features of the AOF are its deformable secondary mirror with 1170 actuators and a laser asterism of 4 artificial stars that probe the atmosphere via four high-resolution Shack-Hartmann wavefront sensors (WFS), each with 40x40 subapertures. The system provides ground layer adaptive optics (GLAO) and laser tomography adaptive optics (LTAO) capabilities. In order to support the commissioning phases of the project, and later optimize and diagnose the operation of the system, a turbulence profiler has been developed and installed in SPARTA, the AOF real time controller (RTC). The profiler estimates two key turbulence parameters: the C_n ²(h) and the outer scale (L₀(h)) profiles and no limit on the number of the estimated layers exists, but for eight layers, the method takes about 2 minutes to yield a full characterization of the atmosphere. The maximum line of sight distance that the profiler can probed the atmosphere depends on the star separation defined for each operation mode: 3km for GRAAL; 14 km for GALACSI wide field and over 35km for GALACS narrow field mode. The remaining turbulence above these maxima (unseen turbulence from the undetected layers) are essential in the GRAAL mode and it is reliably estimated thanks to a novel method to determine the noise in the WFSs, which is mandatory for estimating this upper segment of the turbulence. The technique is also useful to alert about operational problems such as dome seeing and mis-registrations. The method is currently installed in the SPARTA RTC, providing continuous online estimations for the GALACSI (narrow and wide field modes), and for GRAAL mode. Results for several nights comprising hundreds of profiles show very good agreement with other independent measurements.