NICMOS HgCdTe detectors are subject to image persistence resulting from overexposure of bright sources or to cosmic ray hits. The persistence signal is an excess dark current which is an additional source of noise during data reduction. In addition, the detectors are bombarded by charge particles during passage of HST through the SAA which deposit energy in nearly every pixel of the arrays. Starting with the activation of NICMOS following installation of the NICMOS Cooling System (NCS) during HST Servicing Mission 3B, a pair of ACCUM mode darks will automatically be scheduled following SAA passage and before the first science observation. These darks can be used to map the persistent afterglow and be used during data reduction to remove a significant fraction of the persistence signal for observations obtained close in time to the SAA passage.
Persistence is the residual image (signal) in images subsequent to overexposures of bright sources or cosmic ray hits. This excess dark current decays exponentially with a time scale of about 160 +/- 60 seconds. However, there is also a long, roughly linear tail to the decay such that persistence from very bright sources remains detectable for as much as 30 to 40 minutes after the initial exposure.
Cosmic ray persistence adds non-Gaussian, spatially correlated noise to images. During Cycle 7 and 7N, observations obtained following passage of HST through the South Atlantic Anomaly (SAA) were severely impacted by persistence from charge particles which mimic cosmic ray persistence. Excess charge was deposited into nearly every pixel of the arrays. This spatially correlated noise significantly degraded the quality of NICMOS data taken less than 30 minutes after an SAA passage (Najita, Dickinson, and Holfeltz 1998).
Placed into a low-earth orbit by the space shuttle, HST orbital period is about 96 min. with an orbital inclination of 28.5° and eccentricity of 0.00172. There are 8-9 SAA passages every day in consecutive orbits. This means there are 8-9 SAA impacted orbits in a row followed by 5 to 6 orbits in a row that are not SAA impacted. Normally no science is obtained during SAA passage; i.e., the NICMOS cameras are transitioned from OPERATE (on) to SAAOPR (off). Each instrument is assigned its own SAA contour to allow scheduling of power on and off during SAA passage. The STScI Instrument Teams can change the contour for their respective instrument without affecting other instruments. Analysis of data obtained during SMOV (June 1997) was used to define SAA contour Model 23 as the most appropriate for NICMOS observations (Daou and Calzetti 1997).
NICMOS observations are obtained outside of contour Model 23.
During normal operation at the start of an exposure, the NICMOS detectors are commanded to drop out of autoflush mode and to run the pixel reset pattern three times. It takes ~0.6 seconds to complete this task. No data is saved and this task is transparent to the user. The array is then read out to determine the amount of remaining charge on the array, a bias frame. This read is the first read in a MULTIACCUM exposure and is saved. It is called the "zeroth-read". For MULTIACCUM mode, each successive read of the camera is saved. During OPUS pipeline processing, the zeroth-read is subtracted from each MULTIACCUM readout and the difference images are individually calibrated. For ACCUM mode, the NREAD parameter determines the number of reads (N) that are read and averaged. The detector is read N times, and these reads are averaged to form the initial read. The process is repeated to form the final read. The initial read is subtracted from the final read, and the resulting image is sent to the ground. The initial and final reads are not sent to the ground.
The HST schedulers normally resume scheduling observations as soon as the instruments exit their respective SAA contour. Following SAA passage, the NICMOS cameras are transitioned from SAAOPR to OPERATE. Depending upon the orbit, whether or not the SAA passage occurred during bright Earth or Earth occultation (shadow), science observations could resume almost immediately following transition to OPERATE. The detectors are commanded to autoflush mode upon transition to OPERATE mode.
Charge particle induced persistence decays pseudo-exponentially. Therefore, the longer the time since exiting the SAA, the greater is the decrease in the residual noise. To facilitate the reduction in the residual noise, a series of post-SAA dark exposures will be automatically scheduled following SAA passage and before any NICMOS science observations.
A software tool was created that will automatically schedule six ACCUM darks, two per camera, following every SAA passage for which there is an NICMOS observation before the next SAA passage. These ACCUM darks will have proposal ID's of 8790-8795. The scheduling of the six SAA ACCUM darks is transparent to the NICMOS observer.
Post-SAA Dark Product:
Post-SAA darks an be identified by the target name "POST-SAA-DARK". The dark exposures will complete OPUS pipeline processing with a minimum of calibration switches set to perform. calnicb will generate a product that is currently a simple average of the two dark exposures and the product filename will be ipppssoot_saa.fits. The associated post-SAA dark exposures to an observation will be automatically retrieved with the science data whenever a post-SM3B NICMOS science observation is retrieved from the HST Archive.
New NICMOS Keywords:
Four new keywords have been added to the headers of the science data. These keywords contain information about the last exit from the NICMOS SAA contour (SAA-EXIT, SAA_TIME) and the product filename of the post-SAA dark exposures closest in time to the science exposures (SAA_DARK, SAACRMAP). The values of the SAA_EXIT and SAA_TIME keywords will be used by the OPUS pipeline to identify the filenames of the post-SAA darks closest in time to the respective NICMOS observation, and these filenames are written into the header keywords. If no post-SAA dark is appropriate for a science observation, the value of these respective keywords will be set to `N/A'. The new keywords are presented in the following example:
/ POST-SAA DARK KEYWORDS SAA_EXIT='2000.171:12:34:49'/time of last exit from SAA contour level 23 SAA_TIME= 1529 / seconds since last exit from SAA contour level SAA_DARK= 'N/A ' / association name for post-SAA dark exposures SAACRMAP= 'N/A ' / SAA cosmic ray map file