Background on UDF Parallels Design and Scheduling:

NOTE: This concerns only NICMOS, WFPC2, and STIS observations in programs 9979, 9980, and 9981. ACS/HRC autoparallels were done as part of the primary ACS UDF program, 9978. An auxillary primary program, 10086, was used to fill some hard-to-schedule spots, however no parallels of any kind (not even ACS/HRC) were attached to it.

Unlike that which was done for previous community service deep field campaigns, the UDF parallels, which consist of observations using NICMOS, WFPC2, and STIS, were done in pure parallel mode. The observations were performed in this manner for a couple of reasons. First, given the increased size of the ACS UDF program, there was a desire not to complicate or make anything more difficult in terms of the database loading and processing of the primary ACS observations. Second, due to the very highly subscribed period during which the UDF observations were to be performed, it was very desirable from an overall scheduling point of view (i.e. considering other programs besides the ACS UDF) to allow the ACS UDF observations to be scheduled in more highly-SAA-impacted orbits as frequently as possible in order to leave more of the SAA-free orbits available for other high priority GO programs which needed them such as the NICMOS UDF and other programs utilizing STIS MAMAs and some NICMOS observations. Had the observations been coordinated parallels, this would have negatively impacted the overall HST schedule in terms of using more of the precious SAA-free orbits. As may be understood from these choices, we were willing to lose a certain fraction of potential NICMOS parallels which required SAA-free or lesser-impacted orbits in the cases where ACS primaries could be scheduled in more highly SAA-impacted orbits.

In addition to the issues above, we have used the POMS pure parallel system in some non-standard ways. The calendar builders have done extra manual work to ensure that the ACS UDF primary program 9978 has only observations from the three UDF pure parallel programs, 9979 (NICMOS), 9980 (WFPC2), and 9981 (STIS) attached. The calendar builders have also done extra manual work to ensure that no parallels from these programs have been attached to other GO primary observations which would otherwise take away the opportunities of other GO pure parallel programs. Finally, the calendar builders have also used the POMS system in such a way so as to ensure that the UDF parallels are added in our priority order on a per instrument basis.

Since the POMS pure parallels system works on a priority basis, we designated the NICMOS parallel program as our highest priority, followed by the WFPC2 parallels, and, at lowest priority, the STIS parallels. In practical terms, this means that the NICMOS pure parallels are given a chance to schedule in parallel with every ACS primary visit first, and then, on a subsequent pass, the WFPC2 observations are given a chance to schedule, and finally, after that, the STIS observations are given a chance to schedule with every ACS visit. This proceeds in a cumulative manner, so that POMS is taking into account the fact that parallels from another instrument may already have been added. Within each of these per instrument passes, for any given ACS primary visit, the priority-ranked list of visits is considered by POMS, and the first (i.e. highest priority) one which POMS deems suitable is selected as the candidate to be added to that ACS primary visit during the calendar building process. If POMS determines that none of the choices are acceptable for attachment to a given ACS primary visit, then no parallels for that instrument are crafted for that primary visit. Similarly, if POMS deems that a particular pure parallel template visit is acceptable and a parallel visit for that instrument is crafted for attachment to the primary visit and then the parallel visit does not actually fit when the actual flight calendar is being built, then the parallel visit is dropped and no parallel observations for that instrument are added to that primary visit in the actual flight calendar. In previous tests, this "failure to add" rate for the latter case has been roughly estimated at about 5% of visits.

With the NICMOS pure parallels being our highest priority, we wanted roughly equal exposure time for our two filter choices, F110W and F160W. For WFPC2 and STIS, the issue was less complicated since we only had images in F300W for WFPC2, and slitless spectroscopy with 50CCD + G750L preceded by an image with 50CCD + MIRROR for each STIS visit. Consequently, pure parallel visit templates in our programs were generally in priority order of the longest total exposure times per visit, with additional attempts to have them be the result of the longest exposures, in order to minimize read noise. For NICMOS, we had the additional desire to try and keep a good balance between F110W and F160W. Our highest priority NICMOS template observations for POMS to consider consisted of a 2-orbit mix of F110W and F160W with the longest likely available exposure times possible, and our shortest NICMOS template observations consisted of a number of single-filter visits with similar exposure times and alternating priorities. We created many more template visit options than has been usual for most pure parallel programs due to our desire to be able to control and manipulate priorities on a finer scale than is applied to most pure parallel programs. Please see the Phase II listings themselves to see the detailed listing of priorities and other details, but the basic scheme is outlined in the tables below.

UDF Pure Parallels Design / Strategies

NOTE: Priority 99 = High, Priority 1 = Low

NICMOS - Highest priority UDF pure parallel program

WFPC2 - Intermediate priority UDF pure parallel program

All F300w U-band imaging.

STIS - Lowest priority UDF pure parallel program

All 50CCD + G750L (slitless spectroscopy) with 1 short (100s) 50CCD MIRROR image at the beginning of each visit. Despite diagnostics which call for them, CCD fringe flats are NOT needed, even though the central wavelength is > 7000 Angstroms.