|Annu. Rev. Astron. Astrophys. 1991. 29:
Copyright © 1991 by Annual Reviews Inc. All rights reserved
An ideal redshift survey would sample with a well understood selection function a fair volume of the universe. In practice, the approach to a fair volume may yet be unsatisfied, and the selection function is adversely affected by problems of catalog incompleteness, limited accuracy of catalogued parameters, instrumental biases, and loosely defined survey criteria. Those practical concerns often affect the definition of survey strategies by as much as the definition depends on the particular scientific objectives and technology used.
In discussing survey strategies, one should remember that the sky surface density of galaxies to mpg ~ 15.5 is nearly one per square degree, which implies a sky density of ~ 25 to mJ 18, ~ 85 to mJ 19, and so on. Galaxies are strongly clustered on relatively small scales; the galaxy-galaxy correlation function can be described by a power law (r) = (r / r0), with a slope 1.8 and a scale-length r0 ~ 4-8h-1 Mpc. To obtain estimates of the fair properties of the galaxian distribution, a survey needs to overcome both the noise that arises from small-number statistics and that from small-scale fluctuations in the distribution itself. The trade-offs are different, depending on the geometry chosen for the surveyed volume and the specific scientific goals of the survey. In some cases, when the goal is the determination of the large-scale behavior of (r), achieving completeness to a flux limit within the sampled volume might not be necessary, as discussed by Kaiser (1986).