SAMPLE VARIANCE IN LARGE-SCALE COSMIC MICROWAVE BACKGROUND ANISOTROPY MEASUREMENTS

Abstract

The effects of limited sky coverage in large-angle cosmic microwave background anisotropy experiments are investigated by computing the variance of the angular two-point correlation function with an incomplete sphere. We find that, assuming a power spectrum of density fluctuations with spectral index n = 1, the Galactic cut of half-width 20° (40°) about the Equator made by the COBE DMR experiment would induce a sample variance on the rms temperature fluctuation [(ΔT/T) rms ]2 (or equivalently, the correlation function at zero lag), which is 12% (38%) greater than the cosmic variance with a whole sky coverage. This result is about two times smaller than the naive expectation that the cosmic variance is enhanced by a factor of [Formula: see text], where A is the solid angle sampled by the experiment. We also find that the sample variance of the correlation function at nonzero lag can approach the cosmic variance limit. Our approach provides an analytic way of finding a theoretical error to the theoretical prediction for a particular experiment (either large- or small-scale), without having recourse to computationally intensive Monte Carlo or maximum likelihood methods.