10.6	Policies and Procedures

All COS exposures, whether target-acquisition, spectroscopic, or imaging, must be checked for bright objects. Any COS targets or fields that cannot be demonstrated to be safe to a reasonable level of certainty in the judgment of the contact scientist (CS) will not be observed. In that case, it is possible that equivalent alternative targets may be approved upon request. Any observations that trigger the onboard safety mechanisms will not be rescheduled.

Before the Phase II deadline, a COS GO must send to his/her CS the results of ETC calculations for each discrete target (the “ETC run #” is now a required field within APT), and reports on any unsafe or unknown stars from APT/BOT for each field, either showing that the observations are safe or documenting any unresolved issues. (An exception is moving-target fields, which cannot be cleared until the scheduling windows have been established.) It is not expected that all such issues will be resolved by the Phase II deadline, but they should at least be identified and have planned resolutions by then.

Light from a bright nearby source could scatter into the PSA. For example, a target that is safe for the BOA may scatter enough light into the PSA to violate our screening limits. The region of concern is an annulus extending 5 to 15 arcsec from the center of the PSA. Any field object falling in this annulus may not produce a global count rate in excess of 1 × 105 counts/s per segment for the FUV channel or 2 × 105 counts/s for the NUV, or a local count rate over 3.3 counts/s/pixel in the FUV or 250 counts/s/pixel in the NUV. At present, the APT/BOT does not search for such objects, so they must be checked by hand. In such cases, count rates must be estimated using the ETC as though the source were at the center of the PSA.

In worst cases, new ground-based data or HST CCD UV exposures may be required to clear the fields for BOP; in general, the latter must be covered by the existing Phase I time allocation.

For unsafe targets, one solution is to change to a less-sensitive instrument configuration: one could use the BOA, MIRRORB, or both (though the BOA reduces image quality; see Figure 8.2), a higher-resolution grating, or a less-sensitive wavelength setting. Note that the medium-resolution gratings actually have higher throughput than G140L when the data are rebinned, but are subject to brighter limits.

For unsafe field objects that threaten to fall into the non-target aperture, an orientation restriction (ORIENT) may be used to constrain the spacecraft roll angle and thus the position of the non-target aperture, but such constraints will limit the scheduling of the observation.

GOs planning COS observations of unpredictably variable targets, such as cataclysmic variables, should be aware of the special BOP procedures in effect for such cases. These include quiescence verification immediately preceding the COS observations, as detailed in ACS ISR 2006-04, which applies to all HST detectors subject to BOP. Observations of flare stars are allowed with COS (and STIS) only if the Contact Scientist is convinced that the target would not violate BOP limits even in its brightest state. STScI reserves the right to limit the number of visits requiring quiescence observations within 20 days or less of an HST observation to no more than 12 such visits per Cycle.

A pointing or configuration change after the targets and fields have been cleared by the STScI BOP review must be approved by the COS Team on the basis of a specific scientific justification and a new BOP review by the GO, which may be submitted via the CS.