Abstract
The Aperture Masked Interferometry (AMI) mode on JWST-NIRISS is implemented as a 7-hole, 15% throughput, non-redundant mask (NRM) that operates with 5-8% bandwidth filters at 3.8, 4.3, and 4.8 microns. We present refined estimates of AMI's expected point-source contrast, using realizations of noise matched to JWST pointing requirements, NIRISS detector noise, and Rev-V JWST wavefront error models for the telescope and instrument. We describe our point-source binary data reduction algorithm, which we use as a standardized method to compare different observational strategies. For a 7.5 magnitude star we report a 10-a detection at between
8.7 and 9.2 magnitudes of contrast between 100 mas to 400 mas respectively, using closure phases and squared visibilities in the absence of bad pixels, but with various other noise sources. With 3% of the pixels unusable, the expected contrast drops by about 0.5 magnitudes. AMI should be able to reach targets as bright as M=5. There will be significant overlap between Gemini-GPI and ESO-SPHERE targets and AMI's search space, and a complementarity with NIRCam's coronagraph. We also illustrate synthesis imaging with AMI, demonstrating an imaging dynamic range of 25 at 100 mas scales. We tailor existing radio interferometric methods to retrieve a faint bar across a bright nucleus, and explain the similarities to synthesis imaging at radio wavelengths. Modest contrast observations of dusty accretion flows around AGNs will be feasible for NIRISS AMI. We show our early results of image-plane deconvolution as well. Finally, we report progress on an NRM-inspired approach to mitigate mission-level risk associated with JWST's specialized wavefront sensing hardware. By combining narrow band and medium band Nyquist-sampled images taken with a science camera we can sense JWST primary mirror segment tip-tilt to lOmas, and piston to a few nm. We can sense inter-segment piston errors of up to 5 coherence lengths of the broadest bandpass filter used ( 250-500 0m depending on the filters). Our approach scales well with an increasing number of segments, which makes it relevant for future segmented-primary space missions.
