(A) WFPC2: High Latitude parallels
Filters/Exposures: Parallels in are to be taken F814W, F606W, F450W, and F300W, in the following manner:
1) One orbit cases: Do two exposures in F814W only, CR-split evenly
2) Two orbit cases: Add two exposures in F606W, CR-split evenly
3) Three orbit cases: One orbit in F814W and two in F300W, CR-split.
4) Four orbit cases: Add one orbit of F450W, CR-split evenly.
5) Five orbit cases: Add one orbit of F606W, CR-split evenly.
6) 5 orbit cases (usually consisting of two or more 5-orbit pointings). Take first 5 orbits as above, then take the rest in F300W.
Science Goals and General Usability: In every available parallel opportunity, a single orbit in F814W will be taken. This data will be useful for the following science goals, amongst others:
a) Gravitational shear imaging programs, assuming that a good fraction of the primary targets will get dithered enough that the F814W images themselves can be sub-stepped. The best possible sampling and the highest possible resolution is needed for gravitational shear studies. A systematic study of gravitational shear has been carried out using WFPC2 parallels (Casertano et al 2000), with about 350 fields and just under 200 galaxies per field; their results are limited primarily by statistics. Another study will be made through the 1200 orbit STIS imaging parallel GO program of Schneider et al., which will make a major assessment of the problem. A solid gravitational shear study requires good sky coverage with hundreds of fields, each of which should have at least 50-60 well imaged galaxies so that the ellipticities and their deviations from random position angles can be determined with sufficient statistics. This will be the case for the STIS GO program. The additional WFPC2 F814W images will only help in this regard, since WFPC2 has a much bigger field. A single orbit in F814w will provide about 200 galaxies to I=25 mag, of which about half will be exposed well enough to determine ellipticity and PA. In addition, two-orbit cases will provide some color information which can be to improve the quality of the measurement by separating galaxies in redshift space. This data set is complementary to the STIS GO program, since it will go less deep, but cover more area, and so add to it in a different part of parameter space.
b) These F814W images will also be used to search for strong gravitational lenses. The HST Medium Deep survey has been used rather successfully to search for strong gravitational lenses. At least three convincing cases of Einstein crosses were found in several hundred parallel images. So with a discovery rate of one strong lens per hundred parallel fields, the 1000-2000 single-orbit F814W parallel fields per year will yield a few dozen new strong gravitational lenses per year. This is a significant increase over existing numbers. When available, color is very useful to strengthen the selection of lens candidates.
c) The three+ orbit parallels will add deeper exposures in F300W. This is essential, since ACS will have no wide field imaging capabilities below 3800 Ang, and the STIS MAMA's cannot be used for UV imaging in parallel. Hence, WFPC2+F300W is the only near-UV wide field imaging capability that HST has until WFC3 comes along. While the WFPC2 F300W parallel program in Cycle 4-8 has been slow going, it has not been useless. John MacKenty reports that the 100-300 sec ``sponge parallels'' in F300W are useless, but that the orbital exposures, of which he now has several hundreds, yielded in total more than 5000 galaxies to U=23-24 mag. This is a UNIQUE data set, that CANNOT ever be obtained from the ground, and that can be used for comparison to the redshifted rest-frame near-UV morphology of distant galaxies seen with HST in the I-band or beyond. At the same time, it is only 5000 objects from several hundreds of parallel exposures. Most of these objects are very poorly exposed and it is only the brightest 10% that are well enough exposed for high redshift comparision. Hence, continuing this program with deeper F300W parallels in Cycle 9 and 10 seems a reasonable thing to do. Realistically, one can expect around a hundred 3+ orbit WFPC2 parallel fields in each of Cycle 9 and 10. With two orbits in F300W, each Cycle would yield several hundreds of galaxies to U=24-25 mag, of which several dozen would be exposed well enough for meaningful high redshift comparisons. Altogether, this effort would result in several hundreds of usable galaxies in F300W, a major step forward in this area.
d) The 4 and 5-orbit opportunities should be used to take F450W and then F606W. Together with the F814W and F300W, these will be useful for crude photometric redshift estimates, which are needed in gravitational shear studies. They will also be useful for Lyman-dropout searches in the redshift range z=3--5. While the individual 4--5 orbit WFPC2 parallels won't reach very deep, the expected several dozen of such fields per cycle will allow one to find a significant number of dropouts at z3--5, sampling the top of the OLF at those redshifts.
Another main goal of the F450W images is to add significantly
to the current F450W parallel data base, which is no more than about 50
fields. Realistically, many dozens of
such parallels can be expected
per year in F450W. Each of these will add 100-200 galaxies, of
which 50-100 will be bright enough that they can be classified.
Currently, one of the major uncertainties in interpreting the
galaxy counts and evolution of the galaxy luminosity function
is the very uncertain normalization of the galaxy luminosity
function as a function of type. This hinges upon our ability to
classify galaxies in the range 17