Speckle reconstruction of photometric data observed with adaptive optics

Institut für Astrophysik, Freidrich-Hund-Platz 1, 37077 Göttingen, Germany e-mail: [kgp;msailer]@astro.physik.uni-goettingen.de

Received: 26 July 2005
Accepted: 3 April 2006

Abstract

To achieve the highest spatial resolution for ground-based observations one has to correct for degradations by the Earth's atmosphere. This can be done by on-line and post-facto techniques. Here we combine observations with Adaptive Optics (AO) and speckle reconstruction. As possible techniques we present two modified versions (methods B and C) of the Göttingen speckle masking code and describe their application to observations of a solar active region obtained with the Swedish 1-m Solar Telescope (La Palma) by means of AO. By comparing the results with those obtained from the conventional method (method A), the influence of AO on the post-facto reconstruction is revealed. One has to account for the field dependence of the AO correction and the fitting errors of the calculated surface of the deformable mirror with respect to the according wave fronts. Both new approaches enhance the contrast in the lockpoint region less than the conventional method. Method C uses simulated AO corrected speckle transfer functions (STFs). This method shows the most uniform image quality over the whole field of view and indicates the importance of the atmospheric model assumptions and the partial character of the WF correction by AO with appropriate WF error statistics. However, the consideration of field dependent speckle transfer functions in the classical speckle reconstruction (Method B) can serve as a first approach to avoid an intensity over-amplification around the AO-lockpoint region, although the rms of the contrast in the reconstructed images tends to rise slightly towards the outer parts of the field of view. Even the apparently small differences between the speckle transfer functions, used in the different reconstruction methods, cause the modulus of the estimated intensity differences to vary by up to 22%.