2007MNRAS.378.1229T

We present an extension of the spectral deconvolution (SD) method to achieve very high contrast at small inner working radii. We apply the method to the specific case of ground-based adaptive optics fed integral field spectroscopy (without a coronagraph). Utilizing the wavelength dependence of the Airy and speckle patterns, we make an accurate estimate of the point spread function that can be scaled and subtracted from the data cube. The residual noise in the resulting spectra is very close to the photon noise from the starlight halo. We utilize the technique to extract a very high signal-to-noise ratio H- and K-band spectrum of AB Doradus (AB Dor) C, the low-mass companion to AB Dor A. By effectively eliminating all contamination from AB Dor A, the extracted spectrum retains both continuum and spectral features. The achieved 1σ contrast is 9 mag at 0.2arcsec, 11 mag at 0.5arcsec, in 20 min exposure time, at an effective spectral bandwidth of 5.5 nm, proving that the method is applicable even in low-Strehl regimes.

The SD method clearly demonstrates the efficacy of image slicer based integral field units in achieving very high contrast imaging spectroscopy at small angular separations, validating their use as high-contrast spectrographs/imagers for extreme adaptive optics systems.