A summary of the in-orbit calibration activities are listed below. For details on the processes, please refer to the calibration paper - In-orbit Performance of UVIT and First Results and references therein. in-orbit calibration is an ongoing process. Latest values for the filters are available in the Instrument/Filters area. The latest release of the UVIT Calibration Database will be based on the most recent instrument parameters.
Sensitivity
A primary standard (white-dwarf Hz4) was used to calibrate the sensitivity, in the various ultraviolet filters, near the centre of the field. The magnitude system adopted for UVIT is the AB magnitude system (Oke 1974), and hence the zero-point magnitudes derived are in this system. The results shown in Table 2 are based on the spectrum of Hz4 from the CALSPEC database at STScI for NUV and FUV. The entries in this table match the estimates made from calibrations on the ground within ∼15% (see Subramaniam et al. 2016 for more details).
In the NUV the calibration was only done for 2 of the 6 filters, as the count rate for the other filters was estimated (for Hz4) as beyond the acceptable limit. Calibration for these would be done in the future with some other standard.
Spatial resolution and astrometry
Pointing of the satellite has slow drifts up to ± 1' with a rate up to 3"/s. Therefore, in the ultraviolet channels images are recorded at a rate of ∼29/s or higher to keep the blur in individual images to be « 1". The final images are made on the ground by shift and add algorithm. As many ultraviolet images have very low flux it is not possible to derive drift from the images taken at 29/s. In many fields, the NUV images can be used to find the drift by binning the images every one second. A more secure solution to find drift is to use images taken by the visible channel every ∼1 s. The final spatial resolution depends on the intrinsic quality of the images given by the optics and detector combination and any errors in correcting for the drift. The final FWHM of the images is found to be ∼1.2" for NUV and ∼1.5" for FUV – these are significantly better than the goal of 1.8".
Spectral calibration
Calibration for sensitivity of the gratings was done with Hz4 (a primary photometric standard), and the dispersions were calibrated with the planetary nebula NGC40. Only one of the two gratings in FUV has been calibrated. The present results for NUV are adversely effected by saturation and are only indicative of the effective areas and spectral resolution. More calibrations would be done in NUV with elimination of saturation, and the second FUV grating would be calibrated. The spectral resolution for NUV (first order) and FUV (second order) is ∼100.
Sensitivity variation
Transmission of the optics and quantum efficiency of the detectors can change with time. In particular, small contamination can lead to reduction in transmission of the FUV optics. Therefore, two bright stars in NGC188 are repeatedly observed every month to get a measure of any variation in sensitivity. So far all the results indicate that any reduction in sensitivity of FUV, with CaF2 filter, is less than 5% over a period of ∼7 months in the orbit. This gives confidence that any contamination, if at all, of the optics is not occurring at a high rate.
Background
The background in FUV and NUV arises from geocoronal lines, zodiacal light and Galaxy. The contribution of geo-coronal lines vary with time of the day and with the level of solar activity. The other contributions depend on the direction of view. In addition to these, a significant contribution is found from the cosmic ray interactions (seen as bunches of large number of photons in about 3 frames every second) which give a background ∼120 counts/s. In the FUV it can be a major background in dark fields. This background can be minimized by rejecting the frames with number of photons more than a threshold, e.g. average number observed per frame plus 3-sigma.
Artifacts in the images
The optics are not expected to give any ghosts which are outside the PSF of point sources. However, we do see faint streaks in some NUV images which are inferred as due to some star near the edge of the field. With more observations we expect to develop a better understanding of these streaks.
