Abstract

The goal of adaptive optics (AO) systems for astronomical telescopes is to remove the distortions caused in optical images by the turbulent atmosphere above the telescope. In this work, the Frozen Turbulence Predictor (FTP) has been developed based on the frozen flow model of atmospheric turbulence for use in AO systems. The FTP is designed to predict tropospheric turbulence a few milliseconds after atmospheric data has been collected by the system.

By implementing the FTP in an AO system, time delay errors inherent to the system can be greatly reduced. Amongst 36 atmospheric models of one, two, and three layers of turbulence, the FTP was able to reduce the time delay error by an average of 68%. In favorable cases the FTP was able to reduce the time delay error by over 90%. While other linear atmospheric predictors have shown similar reduction in time delay errors, the advantage of the FTP is that it requires very little data to make predictions. This allows predictions to be performed very rapidly and within the time limits of the frozen flow model of turbulence.

An additional benefit of the FTP is that it solves for the turbulence profiles in multiple turbulent layers using a single wavefront sensor. This is important for AO systems that attempt to compensate turbulent layers individually. Using two and three layer turbulence models, the ability to resolve each of the input layers is demonstrated. It is shown that the ability of the FTP to predict turbulence is tied to its ability to reconstruct the layers in the turbulence.