-Stefano Castertano & Henry Ferguson
There are two small but significant systematic calibration uncertainties that can affect stellar photometry with the WFPC2 under certain conditions. These have been described in the WFPC2 Handbook, in previous STANs, and in Holtzman, et al. (1995). The first is the charge-transfer efficiency (CTE) problem which causes some signal to be lost when charge is transferred down the chip during readout. This has the effect of making objects at higher row numbers (more charge transfers) appear fainter than they would be if they were at low row numbers. The second is an apparent difference in the measured flux from stellar sources in short exposures (~200s) vs. that measured in long exposures (~1000s). We briefly describe these effects here, and give WWW links to detailed reports of on-going studies.
For sources exposed to more than 1000 DN with backgrounds typical of wide-bandpass optical filters CTE effects introduce systematic errors up to 0.03 mag in photometry, which can be largely removed by the photometric ramp (or row number) correction previously described elsewhere. For sources exposed to only a few hundred DN the systematic effects are potentially larger. For exposures with significant background (>50 electrons, i.e. >600s in F555W) the error should be 0.05 mag or less. For exposures with little background (exposures <100s, or narrow band filters) the errors are harder to quantify in the existing calibration observations, but may be in the range ~0.05-0.1 mag at 100 DN. For sources exposed to a few 10's of DN, CTE may introduce photometric errors of up to 0.2 mag, in cases where the background is very low (i.e. very short exposures and narrow band filters). It is also apparent that this effect should be largely removed by a preflash of ~160 electrons (although the overhead time and photon noise associated with preflashing is signifcant).
While the CTE effect shows a strong dependence on row number, the magnitude offset in short vs. long exposures persists even at low row numbers, and results in faint targets being too faint by up to ~0.05 mag in short exposures. This offset appears to be significantly smaller in small apertures (2 or 3 pixel radius). An empirical correction for this effect is discussed, which consists of adding ~2 electrons to each pixel in the target aperture.
The STScI WFPC2 group is working to understand these problems and provide a calibration. In the meantime, it was thought useful to make available some of the analysis done in the last few months. The pages referenced below show a partial analysis of frames taken with different exposure times and different preflash levels. The descriptions are culled from viewgraphs and are a bit terse. These pages will evolve over the next few months as our understanding progresses.
We thank Peter Stetson, Jeremy Mould, Jon Holtzman, and John Trauger for sharing their insights into these effects.
Please send comments or questions to the undersigned. (5/31/96)