Empirical corrections provide an alternate way to correct for CTE losses of observations of point sources. Noeske et al (WFC3 ISR 2012-09
) studied star clusters NGC6791 and 47 Tuc and examined how the measured fluxes of stars varied when particular stars were placed close to and far from the readout register. They documented the losses experienced by point sources as a function of (1) the electron flux of the point source, (2) the detector Y position of the source, (3) the background level, and (4) the date of the observation.
The most recent empirical corrections are described in WFC3 ISR 2015-03
, which also provides an analysis of the long term behavior of the UVIS CTE, using observations of star clusters.