Numerical Simulation of Non-Gaussian Random Fields with Prescribed Marginal Distributions and Cross-Correlation Structure II: Multivariate Random Fields

摘要

We provide theoretical procedures and practical recipes to simulatenon-Gaussian correlated, homogeneous random fields with prescribed marginaldistributions and cross-correlation structure, either in a N-dimensionalCartesian space or on the celestial sphere. We illustrate our methods usingfar-infrared maps obtained with the Infrared Space Observatory. However, themethodology presented here can be used in other astrophysical applications thatrequire modeling correlated features in sky maps, for example, the simulationof multifrequency sky maps where backgrounds, sources and noise are correlatedand can be modeled by random fields.