HgCdTe detector arrays like those in NICMOS are subject to image persistence. When pixels collect a large amount of charge, they will tend to “glow” for some time after the end of the exposure.
The persistent signal appears as an excess dark current and decays exponentially with a time scale of about 160
±60 seconds (different pixels show different decay rates), but there is also a long, roughly linear tail to the decay such that persistence from very bright sources remains detectable as much as 30 to 40 minutes after the initial exposure. Subsequent exposures can therefore show residual images. With NICMOS, this can happen under a number of circumstances. Exposures of bright astronomical targets can leave afterimages which appear in subsequent images taken within the same orbit. If you are observing bright objects you should be aware of this potential problem: dithered exposures may contain “ghosts” of bright stars from previous images. It appears that all sources of illumination leave persistent afterimages, but under typical conditions they are most noticeable for sources which have collected 20000 or more ADU during the previous exposure. There is little that can be done to avoid this. If observations are well dithered, then the persistent afterimages can usually be recognized and masked during data processing when combining the images to form a mosaic. This, however, is not done by the standard calibration pipeline.
More insidiously, during regular passages of HST through the South Atlantic Anomaly, the arrays are bombarded with cosmic rays, which deposit a large signal in nearly every pixel on the array. The persistent signal from these cosmic rays may then be present as a residual pattern during exposures taken after the SAA passage. This appears as a “mottled,” “blotchy” or “streaky” pattern of “noise” (really signal) across the images, something like a large number of faint, unremoved cosmic rays. These persistent features cannot be removed by the
MULTIACCUM cosmic ray processing done by the standard pipeline because they are not transient. Rather, they are a kind of signal, like a slowly decaying, highly structured dark current.
Cosmic ray persistence adds non-Gaussian, spatially correlated noise to images and can significantly degrade the quality of NICMOS data, especially for exposures taken less than 30 minutes after an SAA passage. Count rates from moderately bad cosmic ray persistence can be of order 0.05 ADU/second, with large pixel-to-pixel variations reflecting the spatial structure of the signal. The effective background noise level of an image can be increased by as much as a factor of three in the worst cases, although 10% to 100% are more typical. This “noise” is primarily due to the spatially mottled structure in the persistence, not the added Poisson noise of the persistence signal itself. Because HST passes through the SAA many times a day, a large fraction of NICMOS images are affected by cosmic ray persistence to one degree or another. Observations of bright objects are hardly affected, since the persistent signal is usually quite faint. Similarly, short exposures are not likely to be badly affected because the count rate from persistence is low and may not exceed the detector readout noise. But deep imaging observations of faint targets can be seriously degraded. The NICMOS
ISR-98-001 (Najita
et al. 1998) presents a detailed discussion of this phenomenon and its effects on imaging observations.
Starting in Cycle 11, a pair of ACCUM mode NICMOS dark exposures are scheduled after each SAA passage. This provides a map of the persistent cosmic ray afterglow at a time when it is strongest, and has just begun to decay. Experiments using Cycle 7 NICMOS data have shown that it is possible to scale and subtract such “post-SAA darks” from subsequent science exposures taken later in the same orbit, and thus to remove a significant fraction of the CR persistence signal. Doing so comes at the cost of adding some additional pixel-to-pixel Gaussian noise, as the readout and dark current noise from the darks is added in quadrature to that from the science exposures, albeit with a multiplicative scaling that will be < 1, since the CR persistence signal decays with time. Most science programs benefit from the use of post-SAA darks, leading to a significant improvement in the quality of NCIMOS data taken after SAA passages. Software for implementing this correction has been created (Bergeron,
NICMOS ISR-2003-010). It is now part of the NICMOS data reduction pipeline. It is also available as a stand-alone routine called SAAclean within the nicmos package of STSDAS (Barker,
NICMOS ISR-2007-001) Feedback on SAAclean can be sent to
help@stsci.edu.
In addition, observers can plan observations to further minimize the impact of cosmic ray persistence, should it occur. Taking images with as many independent dither positions as possible is one good strategy (which can help in many ways with NICMOS imaging). Without dithers, the persistent pattern will stay fixed relative to the astronomical targets (although its intensity will decay), and co-adding successive exposures will just reinforce the contamination. Dithered images will move the targets relative to the persistence so that it adds incoherently when the data are summed. With well-dithered data (at least three positions), one can also take advantage of the MultiDrizzle software (Koekemoer
et al., 2002 HST Calibration Workshop, p337) and associated software in the STSDAS
dither package to identify and mask the worst effects of persistence (as described in the
HST Dither Handbook v2.0, Koekemoer
et al. 2002). The
HST Data Handbook reports more details on how to handle images affected by cosmic ray persistence.
