Interpolation and smoothing

Smoothing is omnipresent in astronomy, because almost always measurements performed at discrete positions in the sky need to be interpolated into a smooth map for subsequent analysis. Still, the statistical properties of different interpolation techniques are very poorly known. In this paper, we consider the general problem of interpolating discrete data whose location measurements are distributed on the sky according to a known density distribution (with or without clustering). We derive expressions for the expectation value and for the covariance of the smoothed map for many interpolation techniques, and obtain a general method that can be used to obtain these quantities for any linear smoothing. Moreover, we show that few basic properties of smoothing procedures have important consequences on the statistical properties of the smoothed map. Our analysis allows one to obtain the statistical properties of an arbitrary interpolation procedure, and thus to optimally choose the technique that is most suitable for one's needs