Representation of conditional random distributions as a problem of "spatial" interpolation

Abstract The problem of specifying random distributions conditional upon external "independent" factors may be seen as a spatial interpolation problem of conditional moments in a generalized phase space. Different techniques solving this interpolation problem are presented, and the different requirements for applications for simulatio and forecast purposes are discussed. The design of universal empirical coordinates is outlined and the concept of data assimilation by means of forecast schemes is sketched. The Interpolation Proble