|Space Telescope Science Institute|
We give here a summary of general recommendations for both Phase I and Phase II proposal preparation. Recommendations are based our experience with NICMOS. However, observers are strongly advised to read the technical sections that follow in order to develop an optimal observation strategy based on the demands of their individual scientific goals. Also the Advisories page maintained on the NICMOS WWW site should be consulted for updates.
• NIC1 (NIC2) offers diffraction-limited capabilities at J- (H-) band and longer wavelengths, while NIC3 offers high sensitivity (due to the lower angular resolution) with the largest field-of-view among the three cameras (see Section 2.3.2). In particular, NIC3 reaches fainter magnitudes than the other two cameras (for the same exposure time) for observations which are not limited by photon noise from source + background, i.e. where read-out or dark noise are significant. This is true for most observations of faint targets. However, the poor spatial sampling of NIC3 can limit its sensitivity for faint point sources, and also limits photometric accuracy for point sources.
• When choosing NIC3, proposers should be aware that this camera is slightly out-of-focus, with a typical loss of peak flux around 20% and a loss of encircled energy of about 10%–15% at 0.2" radius. Chapter 4 should be consulted for the detailed performance of the out-of-focus NIC3.
• The highest sensitivity gain relative to ground-based observations is at wavelengths shorter than 1.8 μm. The background at J and H seen by HST is a few hundred times smaller than at ground-based observatories. The background at K is only marginally better on HST, due to the telescope’s thermal emission. However, observations in the thermal regime (longward of 1.8 μm) may be more advantageous with NICMOS if high angular resolution is a requirement for the science goal. Also, with NICMOS, one gains in stability of the thermal background.
• Observations of extended sources in the thermal regime (longward of 1.8 μm) may need to obtain background observations as well (chopping off the target). Given the stability of the thermal background, however, it will not be necessary to get background measurements more frequently than once per orbit. For point sources or extended sources which do not fill the camera field-of-view, images of the thermal background can be obtained with dithering (see recommendations below).
• For the purpose of removing cosmic rays, photon and cosmic-ray persistence, detector artifacts, and for averaging out flat-field sensitivity variations, observers are strongly advised to dither an observation as much as possible. This implies dividing single-orbit observations into at least three exposures and multi-orbit observations into two exposures per orbit. The general advice of dithering is generally not applicable to observations of faint sources around/near bright ones. If the bright source saturates the detector, the saturated pixels will be affected by persistence; in this case, the observers have two options: 1. Not dithering, to avoid placing the faint target on the saturated pixels; 2. Dithering by large amounts (by roughly one full detector quadrant) to move away from the persistence-affected region.
• The dithering requirement poses a practical upper limit of ~1,500 seconds to the longest integration time for a single exposure. This is roughly equivalent to having 2 exposures per orbit. Observers wishing to detect faint targets should work out their S/N requirements by determining (e.g., with the ETC) the S/N achieved in a single exposure and then co-adding n exposures according to , until the desired S/N is achieved. The latter is an essential step given the large read-out noise of the NICMOS detectors.
• In the case of crowded fields, observers are advised to dither their observations with sub-pixel sampling. In post-processing, the images should be combined with the MultiDrizzle software (Fruchter, A. and Sosey, M. et al. 2009, "Multidrizzle Handbook", Version 3.0, (Baltimore, STScI)), otherwise the limiting sensitivity of NICMOS will not be reached due to PSF overlap and confusion.
• Proposers who want to use the NIC3 grisms should be aware that the spectral resolution quoted in the Handbook (R~200) is per pixel; the actual resolution, calculated over 2 pixels, is R~100.
• Observers proposing to use the NICMOS coronagraphic hole may want to consider observing the same object twice in the same orbit, or back-to-back in adjacent orbits, with an in-between roll of the spacecraft for optimal PSF subtraction.
• Observers proposing to use the NICMOS coronagraphic hole may want to consider adding contemporary flat-field observations.
• For fields containing faint targets only, the linear MULTIACCUM sequences (SPARS...) should be preferred. They are best suited for removing instrumental effects from the astronomical data. However, for fields containing both bright and faint sources, logarithmic sequences should be preferred (STEP...), as they offer the largest dynamic range and allow the calibration software to recover saturated targets.
• When designing dithering patterns, observers should take into account that the sensitivity across each detector changes by as much as a factor ~2.5 at short wavelengths and by a factor ~2 at long wavelengths. The sensitivities used in the ETC and in this Handbook are average values across each detector. The sensitivity variations will mostly affect observers interested in the full field-of-view.
• The NIC3 PSF is undersampled and intrapixel sensitivity variations are large in this camera (see Xu, NICMOS ISR-2003-009). Photometry on point sources can vary by >0.2 mag in J and up to 0.2 mag in H depending on the placement within a pixel. Observers are encouraged to consider sub-pixel dithering in their NIC3 observations; 4–6 dithering positions minimum are recommended.
• The bottom 10–15 rows of the three cameras’ fields of view are somewhat vignetted and interesting targets should not be placed in this region.
• For coronagraphic observations: coronagraphic hole movements and HST focus changes (breathing) will result in residual noise during PSF subtraction. PSF stars should be observed close in time to the primary target.
• HST absolute pointing is only good to about 1 arcsecond (); dithering patterns should be designed to place science targets away from the edges of the cameras by at least this amount.