4.6 WFC CCD Detector Charge Transfer Efficiency - CTE
4.6.1 The Issue
The ACS/WFC CCD detectors operate by the simple process of converting incoming photons into electron/hole pairs, collecting the electrons in each pixel, and then transferring those electrons across the detector array during the device readout. The transfer process moves each pixel's electrons down along the columns and then across in a transfer register to the amplifier located in the corner of the detector array.
When the detectors were manufactured, these transfers were extremely efficient (typically 0.999996 of each charge packet was transferred successfully from one pixel into the next), which means that slightly over 99% of the charge collected in a pixel would be delivered to the transfer register. Once in space, the flux of energetic particles such as relativistic protons and electrons damages the silicon lattice of the CCD detectors. This creates both "hot" pixels and charge traps. This radiation damage is cumulative and was unavoidable given current technologies for detector construction and shielding.
The charge traps degrade the efficiency with which charge is transferred from pixel to pixel during the readout of the CCD array. This is seen directly, as shown in Figure 4.28, in the "charge trails" that follow hot pixels, cosmic rays, and bright stars that can extend to over 50 pixels in length.
4.6.2 Improving CTE: Considerations Before Making the Observations
The simplest mitigation of the imperfect CTE is to reduce the number of charge transfers required for a given source to reach the readout amplifier on the CCD. If the source of interest is small (~10″ across or less), placing it close to the corner of the detector will result in greatly enhanced net CTE. APT now has a pre-defined aperture setting for this purpose called
CTE is a strong function of the signal level in the pixels through which a charge packet must pass on its way to the transfer register. Observations with very low background (< 20 e¯ for ACS) will suffer large losses for very faint sources. This is likely to be problematic for narrow band filters and extremely short exposures. In these cases, raising the background will greatly improve the CTE and thus the S/N of these sources. For users planning to stack multiple images to reach very faint limits, they should plan to achieve a background level of ~20 e¯ for ACS.
The background can be increased in several ways:
- Longer exposure times
- Selection of a broader filter
- Addition of internally generated photons (i.e., "post-flash").
ACS/WFC contains LED lamps configured to illuminate the side of the shutter blade that faces the CCD detector. Designed to provide fairly even illumination at low signal level, these lamps provide a "post-flash" capability. While the post-flash lamp can be used to increase the background in an image, it is recommended with reservations due to a 50% variation in the signal across both chips. See ACS ISR 2014-01 and the ACS website for more details.
4.6.3 Improving CTE: Post-Observation Image Restoration
The ACS team has developed and implemented a post-observation correction algorithm based upon the Anderson and Bedin (2010, PASP, 122, 1035) methodology. This empirical algorithm is based on a model for charge-transfer loss and release that reproduced the observed trails behind warm pixels. The correction software then uses an iterative forward-modeling process to estimate the source image from the observed trailed image.
The original version of the code worked very well for intermediate to high flux levels (> 200 electrons), but data were not available at the time to test it at lower flux levels. Several ISRs describe the original correction: ACS ISR 2011-01, ACS ISR 2012-03, and ACS ISR 2012-04.
Recently, we have taken calibration data that will allow a higher fidelity model at all flux levels. These data have been used to re-parameterize the model and make additional improvements (such as read-noise mitigation). The new model is now available as a part of the standalone pipeline and will soon be available via MAST. An ISR (Anderson 2017, in prep.) describing the new correction and evaluating its performance with on-sky tests is in progress and should be available before the end of 2017.
While pixel-reconstruction algorithms may do a good job removing trails behind stars, cosmic rays, and hot pixels, they have one serious and fundamental limitation: they cannot restore the lost S/N in the image. This limitation notwithstanding, the reconstruction algorithm provides the best understanding of the "original" image before the transfer, and also helps understand how the value of each pixel may have been modified by the transfer process. This algorithm is available in the ACS pipeline; standard calibrated products are now available both with and without this correction.
In general, we find that the correction is good to about 25% for stars with moderate signal to noise, so one can get a sense of the reconstruction error by determining the amplitude of the correction and taking 25% of that as the error. The upcoming ISR (ACS ISR 2018-04) should provide some empirical demonstrations of the efficacy of the correction.
The expected losses should be taken into consideration when one is deciding on the best CTE-mitigation strategy, which may involve taking fewer longer exposures to preserve S/N even with the increased cosmic-ray contamination.
ACS Data Handbook
- • Acknowledgments
- • Change Log
- Chapter 1: ACS Overview
- Chapter 2: ACS Data Structure
- Chapter 3: ACS Calibration Pipeline
- Chapter 4: ACS Data Processing Considerations
- Chapter 5: ACS Data Analysis