The MIRI/MRS Library

Context. The Medium Resolution Spectrometer (MRS) of the Mid-InfraRed Instrument (MIRI) on the James Webb Space Telescope (JWST) is affected by interferometric spectral fringing, due to scattering within the detector and dichroic layers. The amplitude of these oscillations on the spectrum can be up to 30%. Correcting them is non-trivial, since the depth and phase of the fringes depend strongly on the illumination pattern and the way the pixels sample it. By default the JWST pipeline uses static fringe flats to divide out the fringes. These flats are representative for a spatially homogeneous extended source, but not for point sources. The significant residuals in the data are removed by using a self-calibrating correction step which can alter physical features in the spectra in a non-systematic way.Aims. We build on our corrections from Paper I (, A&A, 688, A226) in this series, to derive a library of detector-based fringe flats for unresolved sources in a nine-point mosaic around all nominal MRS point source dither positions. We provide users with either an absolute or interpolated fringe flat that can correct the fringes without the need for self-calibration, hence mitigating the risk of altering astrophysical features of interest.Methods. We used the data of 10 Lac from the Cycle 2 calibration programme 3779 to create the library of fringe flats. By removing the continuum and spectral features from the data at the detector-plane level, each of the nine mosaic points around the eight dither positions resulted in a pointing specific fringe flat. By assessing the difference in response between the individual pointings, we found correction factors to bring all the spectra to the same level, and used these to derive a single spectrophotometric calibration curve per band.Results. The library of fringe flats is able to reduce the remaining power of the fringe frequencies on the detector by up to two orders of magnitude compared to the current pipeline flats tailored to extended sources. This improvement carries over to the residuals in the cube spaxels, where the contrast is reduced from >10% to <1−2%. This becomes less apparent after extracting a spectrum from the cube, where in channel 1 averaging of fringe phases in the current pipeline case can reduce its residual contrast. The spectrophotometric calibration curves have a root-mean squared variation of less than a percent in all bands except bands 4B and 4C, while channels 2 and 3 have a stability within 0.5%. Sources taken without target acquisition (TA) fall outside the mosaic grid, but our correction improves the defringing depending on the source location.Conclusions. The improvements in fringe residual found are significant on the detector and spectrum-level. The corrections derived here are directly compatible with the current JWST pipeline infrastructure, and work best for unresolved sources observed with TA in one of the nominal point-source dither patterns.