For UVIS data, the calibrated flt/flc files of scanned data can be used directly for analysis, just as for staring-mode observations.

10.3.1 Astrometry using UVIS Scanned Observations

There are several existing papers in the literature that illustrate the use of UVIS scanning-mode data for astrometric measurements (e.g., Riess et al. 2021, Riess et al. 2014, Casertano et al. 2016). In these investigations, the authors have been able to achieve an astrometric accuracy of ~30 micro arcseconds in the measurements of trigonometric parallaxes, more than 10 times the precision achievable with staring mode observations. The interested reader is directed to such references for more details about the data analysis process.

10.3.2 Photometry of Bright Targets using UVIS Scanned Observations

HST program 14878 was a Cycle 24 calibration program intended to demonstrate the photometric repeatability of spatial scans of bright, isolated stars with WFC3/UVIS. Analysis of two identical visits showed that the photometric repeatability was ~0.1% r.m.s. (WFC3 ISR 2017-21), an improvement of more than a factor of 5 over the traditional results using staring mode. Sensitivity losses for UVIS spatial scans are relatively flat (~0.1-0.2%/year), independent of wavelength on both UVIS CCDs, and show no evidence of contamination (WFC3 ISR 2022-04).

As an example of the analysis procedure, the steps performed on program 14878 are summarized below (the aforementioned report contains additional details).

• Calibrated (*_flt.fits) products, processed with the calwf3 calibration pipeline, were retrieved from the MAST archive. Vertical scans and corner subarrays were used in this program to mitigate CTE losses.
• Cosmic rays (CRs) were removed using an algorithm originally developed for CR rejection in STIS CCD images. This algorithm identifies cosmic ray hits in the scanned images and replaces them with an interpolated value from the surrounding ‘good’ pixels.
• Images were sky subtracted. The sky region is defined as all pixels excluding a 10-pixel border around the perimeter of the subarray, and a conservatively large 400 × 75 pixel rectangular aperture around the source center. The pixel values in the sky region are sigma clipped (iteratively) and the mean of the remaining sky pixels is subtracted from the science array to remove the sky.
• The pixel area map was applied. This step can be important if the position of the scan on the subarray drifts significantly between visits.
• A rectangular aperture was centered on the scan using photutils.RectangularAperture. For higher precision, the overlap of the apertures with the science array was set to the ‘subpixel’ mode to allow for subpixel centering of the scans. Finally, photutils.aperture_photometry was used to sum up source counts.