Content

Introduction

Internal monitoring of the decay of parallel CTE of the ACS/WFC CCDs is accomplished with the extended pixel edge response (EPER) test. EPER images have extra-large overscans in which to accumulate trapped charge during readout of the lamp-illuminated active pixels. Comparing the amount of trapped charge in the overscans to the signal level of the illuminated pixels provides an estimate of CTE per pixel. EPER images of various signal levels have been obtained since ACS was installed on HST, which provides a long baseline over which to track CTE losses. Recent analysis of parallel CTE for ACS/WFC is described here (ACS ISR 2018-09). Future work includes an analysis of serial CTE in WFC from EPER data.

EPER images have extra-large overscans in which to accumulate trapped charge during readout of the lamp-illuminated active pixels. Comparing the amount of trapped charge in the overscans to the signal level of the illuminated pixels provides an estimate of CTE per pixel. EPER images of various signal levels have been obtained since ACS was installed on HST, which provides a long baseline over which to track CTE losses. The 

Modeling the CTE

Parallel CTE from EPER data has a power law dependence on signal level (s) and linear dependence on time (d) described by,

\(CTE(s,d) = 1.0 - (n+c(d-52335))s^p\)

with the variables and coefficients defined below:

  • CTE := Charge transfer efficiency measured from EPER image
  • s := signal level in electrons.
  • d := MJD date of the observations
  • n := intercept of the linear time dependence
  • c := slope of the linear time dependence
  • p :=  exponent of the power law dependence

 

The best-fit coefficients quoted below are from the most recent analysis of EPER data (ACS ISR 2018-09).

​​​​​

Epoch n c p
Pre-SM4 \( 6.8×10^{−5} ± 4×10^{−6}\) \(1.66×10^{−6} ± 4×10^{−8}\) \(-0.528 \pm 0.003 \)
Post-SM4 \(−1.3×10^{−4} ± 6×10^{−5}\) \(2.00×10^{−6} ± 3×10^{−8}\) \(-0.521 \pm 0.002\)

 

The rates of decrease of CTE as measured from EPER data for specific signal levels are given in the table below.

Signal level (e-) Rate of decrease of CTE per year
\(180\)

\(−4.9×10^{−5}\)

\(430\) \(−3.1×10^{−5}\)
\(1600\) \(−1.5×10^{−5}\)
\(3400\) \(−1.1×10^{−5}\)
\(7100\) \(−7.2×10^{−6}\)
\(42000\) \(−2.9×10^{−6}\)

 

Power law dependence of CTE on pixel signal
 Parallel CTE as a function of signal level (top) and residuals as a function of signal level (bottom) for programs that obtained data at more than two signal levels. Each set of data is labeled by the anneal date of the observations. The circles represent pre-SM4 data, and the squares and triangles post-SM4 data. Both WFC1 and WFC2 results are plotted with the same symbols. The color-matched curves are the best-fit power law models evaluated at the anneal date of the datasets. The solid curves represent pre-SM4 model fits and the dashed and dotted curves both represent post-SM4 model fits.
Linear time dependence of CTE
 Parallel CTE as a function of time (top) and residuals as a function of time (bottom) for programs that obtained data at the labeled signal levels. Both WFC1 and WFC2 results are plotted with the same symbols. The color-matched curves are the best-fit power law models evaluated at the average signal level of the datasets, separated according to pre- and post-SM4 observation dates. The gray shaded band is the time period during which ACS/WFC was offline.

 

LAST UPDATED: 06/17/2019

Please Contact the HST Help Desk with any Questions

https://hsthelp.stsci.edu.