Speaker
Description
The Near Infrared Camera and Multi-Object Spectrograph (NICMOS) instrument aboard the Hubble Space Telescope was the highest resolution near-infrared imager in space before the James Webb Space Telescope, with an extensive archive of high-resolution and high-contrast surveys. The HST PSF is imperfectly known and strongly time-varying, due to breathing mode oscillations in the optical structure which alter the telescope wavefront, and, in the case of NICMOS, a misaligned cold mask. We construct a differentiable forward model of the NICMOS optics, which produces images from a parametric description of the on-sky source and optics. We apply this model to PSF calibrators to simultaneously retrieve the wavefront and pupil metrology, and to time series observations to characterise the breathing mode while simultaneously reconstructing a clean exoplanet transit curve directly from the model fit without further photometric calibration. We additionally demonstrate information-optimal spectral retrieval from on-sky broadband images for the first time. We retrieve low-resolution spectra at $R \sim 20$ of known brown dwarfs and white dwarfs consistent with ground-based measurement and can in principle retrieve such spectra for any point-like object. We revisit previous companion search campaigns, correctly treating the variable pupil metrology and retrieving low-resolution spectra of detected companions. Finally we perform calibrator-free deconvolution, discovering and producing a deconvolved image of a new debris disk around a T Tauri star. This disk is unusually large and bright, and shows evidence of shadowing from a known unresolved inner disk which we also detect from its low-resolution spectrum.
| I am the presenting author | Yes |
|---|