MIRI Imaging Calibration Status
The overall calibration status and estimated accuracy of the MIRI imager are described in this article; please also see the article on known issues affecting MIRI Imaging data.
On this page
Astrometric calibration
Words in bold are GUI menus/
panels or data software packages;
bold italics are buttons in GUI
tools or package parameters.
Photometric calibration
The MIRI imager experiences a time-dependent reduction in observed count rates at wavelengths longward of F1280W; this count rate loss is similar to the MRS, but of a significantly smaller magnitude. As indicated in Table 1, the observed reduction between the JWST commissioning period and May 2025 was about 26% at the longest wavelengths. The coronagraphic filters are assumed to have effects as interpolated between the nearest bracketing imaging filters. The shorter wavelength filters (F560W, F770W, F1000W, F1130W) do not show a detectable effect.
Table 1. MIRI imager count rate reduction between the JWST commissioning period and May 2025
| Filter | Reduction in count rate |
|---|---|
| F1280W | 3.1% |
| F1500W | 3.5% |
| F1800W | 9.0% |
| F2100W | 14.8% |
| F2550W | 25.7% |
The root cause of this issue is still under investigation, with a combined focus on the MRS and imager. Regular monitoring observations are being taken with the MIRI imager to continue measuring the photometric response, and to characterize the temporal trend alongside what has been observed in the MRS. As illustrated in Figure 1, the overall rate of loss is decreasing with time with an exponential plus line functional form.
Figure 1. MIRI imager count rate loss
Click on the figure for a larger view.
(Left) Imager relative count rates obtained using data from standard stars as a function of time for each of the nine imager filters. Most points correspond to observations of BD+60 1753. For the 15-micron and longer wavelength filters, observations from δ UMi and HD 37962 have been used to fill in a time gap in the first year between the BD+60 1753 observations. The green line represents an exponential plus line fit to the data, and the delta change between the start and latest measurement is given for each filter.
(Right) The residuals to the best fit model are shown with the 1-sigma standard deviation given above each set of residuals.
Users wishing to propose observations in filters F560W through F1800W can assume the SNR calculations in the ETC are sufficiently accurate (nominally, the SNR goes as the square root of the count rate). For the F2100W and F2550W filters, adding margins of 5% on the SNR is recommended to mitigate for the observed loss.
To compensate for this count rate loss in the JWST calibration pipeline, as of May 2025, this loss is taken into account in the pipeline using a model of the evolving response function. After this correction, uncertainties are of order 1% or better at all wavelengths (Gordon et al. 2025). The corrected telescope/instrument stability per band is given by the 1 sigma standard deviation in the right panel of Figure 1.