PARTIAL ADAPTIVE COMPENSATION AND PASSIVE INTERFEROMETRY WITH LARGE GROUND-BASED TELESCOPES

The possibilities for diffraction-limited imaging with large ground-based optical telescopes have been investigated assuming the availability of a modest adaptive optics system. Using numerical simulations, we have examined the degree of partial wave front compensation required for useful diffraction-limited direct imaging as well as the performance improvement expected for both filled-aperture speckle imaging and nonredundant mask interferometric imaging in the presence of partial adaptive correction. A partially compensated point-spread function will have a significant diffraction-limited central core if the residual rms wave front error is less than 2 rad. In this regime (the ''sharp-core regime'') deconvolution is likely to permit diffraction-limited direct imaging. However, if the residual rms wave front error is more than 2 rad (the ''speckle regime''), conventional passive interferometric methods will be required to attain diffraction-limited resolution. To study the improvement in passive interferometry we have investigated the effect of a low-order adaptive optics system, correcting the first 21 terms in the Zernike expansion of the wave front perturbations, for the case of D/r0 = 25. By using the attenuation of the monochromatic speckle transfer function to define an effective measure of r0 we find that for Kolmogorov atmospheric turbulence characterized by a value of r0 of 0.2 m, the 21 term Zernike corrected value of r0 is 0.29 m. The quantitative improvement in the power spectrum transfer function at finite bandwidth can primarily be accounted for by this increase in effective ro which reduces the redundancy of baselines caused by both the finite bandwidth and the spatial extent of the pupil. For optical aperture synthesis imaging, the higher value of r0 permits larger subapertures to be utilized which in turn permits wider bandwidths to be employed without geometric attenuation of the power spectrum. The derived values of r0 can be used to estimate the improvement in the coherence time tau0 and the overall signal-to-noise ratio of the imaging procedure. These results predict that the signal-to-noise ratios of the power spectrum and bispectrum are increased by factors of approximately 3 and approximately 5, respectively, for both filled-aperture speckle imaging and optical aperture synthesis imaging.