Bayesian retrieval algorithm to detect unresolved point sources in JWST/MIRI spectra using the detector-produced interferometric fringes

Abstract

Similarly to other spectroscopic instruments operating in the infrared wavelength range, the observations of the Mid-Infrared Instrument (MIRI) on-board the James Webb Space Telescope (JWST) are subject to fringing. The different layers in the detectors act as Fabry-P´erot etalons, causing fringes up to 30% in depth. The depth and phase of these detector fringes is not constant for all source geometries on the sky, and depends on the illumination of the MIRI pupil. This means that point sources will show a different fringe pattern from semi-extended sources and extended sources. In fact, it has been found that a smooth change in depth and phase occurs depending on what part of the point spread function (PSF) is sampled. Here, we aim to use the predictable change in fringe pattern to find evidence for the presence of an additional body in what is seemingly a single point source. To do this, we create a forward model of the PSF including fringes and insert this into a Bayesian retrieval loop. The Bayesian loop finds the coordinates on the detector for one or more point sources that best fits the data of two point sources added together. We find that the code is able to identify the along-slice coordinates of the two point sources, but there is less of a dependency on the fringes with the across-slice coordinate. Using commissioning and Cycle 1 data, we will be able to better characterise the fringes, and improve the forward models.