Data processing for exoplanet detection using direct imaging

Abstract

This communication is devoted to data processing of images obtained using an extreme adaptive optics (AO) system and a coronagraph. Specific attention is given to the following degrading factors: the residuals of atmospheric turbulence after AO correction and the side eects of the coronagraph. Relying on a statistical modeling of the measurements a test based on short exposure images is proposed. This processing, which generalizes the dark-speckle technique, takes into account the local variance of the complex amplitude residuals and the deterministic response of the system (i.e. without atmospheric turbulence). 1. Assumptions and data model Let us start considering a high-flux regime. We adopt in the sequel the model derived in Aime & Soummer (2004). The complex amplitude resulting from the propagation of light through the atmosphere, to the telescope, and then through an adaptive optics (AO) system and a coronagraph can be decomposed into two terms in the focal plane. The first term is the deterministic response of the coronagraph, we denote it c(~) where ~ represents the position in the focal plane. The second term comes from the atmospheric turbulence residuals not entirely corrected by the AO system and propagated through the coronagraph, s(~). The complex value s(~x) is assumed to be circular Gaussian distributed: s(~) » Nc(0,ae 2 (|~|)). As a consequence, the instantaneous intensity i(~) is written |s(~)+c(|~|)| 2 . Moreover, a planet in ~ will basically shift i(~) by a deterministic