Reconstructing cosmological fields using tessellation methods

Abstract

Astronomical observations, physical experiments as well as computer simulations often involve discrete data sets supposed to represent a fair sample of an underlying smooth and continuous field. Reconstructing the underlying fields from a set of irregularly sampled data is therefore a recurring key issue in operations on astronomical data sets. Conventional methods involve artificial filtering through a grid or a smoothing kernel and fail to achieve an optimal result. Here we describe a fully self-adaptive geometric method which does not make use of artificial filtering, and which makes optimal use of the available information.