New NICMOS Readout Sequences Chun Xu

New NICMOS Readout Sequences Chun Xu Space Telescope Science Institute

AbstractOn 12 June 2005, four new NICMOS multi-accum sequences (SPARS4, SPARS16, SPARS32, SPARS128) were implemented in the NICMOS flight-software. All of the SPARS sequences have two initial readouts at 0.303 and 0.606 seconds followed by equally spaced readouts. Observations with the new sequences were carried out in August 2005 (Proposal 10721). The observed dark currents are well matched by the dark current model used in the present pipeline.The advantages of having these new sequences is that observers may choose timing patterns that optimize the handling of cosmic rays with equally spaced reads without having the extra amplifier glow caused by the logarithmic initial readouts of the current STEP sequences.  This is especially desirable for fields with moderate to low dynamic range. Another advantage of these new sequences is a reduction in thermal variations that give rise to fluctuations in bias levels, called "shading".

IntroductionIn order to make better use of the pre-defined readout sample sequences, NICMOS group added four new readout “sparse sampling” (or SPARS) sequences, called SPARS4, SPARS16, SPARS32 and SPARS128, in addition to the existing sequences SPARS64 and SPARS256. The new sequences replace the rarely used MIF sequences. The SPARS readout sequences are recommended for observations of fields where only faint objects are present. Another advantage of using SPARS readout sequences is that these sequences primarily sample the readout time equally so the data can be better calibrated (off the pipeline if necessary) as far as the shading is concerned. The SPARS readout sequences consist of 2 rapid initial readouts at times 0.303s and 0.606s, followed by multiple readouts at equally sampled time series. Table 1 lists all the time sequences for these new SPARS readouts.

Table 1. New Sequences and Readout Time (seconds)

SPARS4 0.303 0.606 3.996 7.997 11.998 15.999 20.000 24.001 28.002 32.004 36.005 40.006 44.007 48.008 52.009 56.010 60.011 64.012 68.013 72.014 76.015 80.016 84.017 88.018 92.019

SPARS16 0.303 0.606 15.996 31.999 47.999 63.996 79.997 95.997 111.997 127.997 143.998 159.998 175.998 191.998 207.999 223.999 239.999 255.999 272.000 288.000 304.000 320.000 336.001 352.001 368.001

SPARS32 0.303 0.606 32.002 64.001 96.000 127.999 159.999 191.998 223.997 255.996 287.996 319.995 351.994 383.993 415.992 447.992 479.991 511.990 543.989 575.989 607.988 639.987 671.986 703.985 735.985

SPARS128 0.303 0.606 127.996 255.996 383.996 511.996 639.997 767.997 895.997 1023.997 1151.998 1279.998 1407.998 1535.998 1663.999 1791.999 1919.999 2047.999 2176.000 2304.000 2432.000 2560.000 2688.001 2816.001 2944.001 Data Reduction and Analysis

The observations of dark current took place in August 2005. All the observation were scheduled in SAA free orbits to avoid excessive cosmic ray persistency, with the three cameras operating in parallel. The data are calibrated through standard pipeline. The resulted ima files off the pipeline contain practically all the dark images at the corresponding readouts. These darks should be matched exactly by the model dark images used in the pipeline if the models are correct. To perform such comparison, we generate corresponding model dark images with calnica in iraf by setting keyword darkcorr to “PERFORM” in the raw files and setting writedark option to “yes” in calnica. We then compare the generated model dark images with the observed darks (i.e., ima files off the pipeline) extension by extension. Figure 1 shows the observed darks, model darks and their differences side by side, extension by extension, for all 3 cameras and all 4 new readout sequences. We only list first 9 out of 25 readouts to save the space. This figure shows visually how well the model matches the observed darks.

Figure 2 plots the detailed differences between the observed darks and model darks.

Figure 1. The comparison between the observed darks, the model darks and their differences. The 3 panel images are for 3 cameras, 4 columns in each panel are for 4 sequences. In each column, there are 3 sub-columns, which are observed darks, model darks and their differences. The smoothed difference images infer that the model and observed darks are well matched. Note that the there are some quadrant by quadrant offsets in the difference images, these offsets are caused by incomplete removal of the pedestal.

Figure 2, The mean value (in DN) of the difference between observed and model darks are shown for each readout in a full 25 read sequence.  Different SPARS sequences are color coded as indicated.  In general, the model darks are consistent to better than +/- 10 DNs compared to observed darks.  Differences are due to incomplete removal of pedestal and time-dependent effects related to the bus voltage and aft-shroud temperature.

Acknowledgements: The author would like to thank Louis E. Bergeron, Ilana Dashevsky and Keith Noll for their valuable comments.