||In this talk, I present a general computational method for characterizing the molecular structure of liquid water interfaces as sampled from atomistic simulations. With this method, the interfacial structure is quantified based on the statistical analysis of the orientational configurations of interfacial water molecules. The method can be applied to generate position-dependent maps of the hydration properties of heterogeneous surfaces. I present an application to the characterization of surface hydrophobicity, which is used to analyze simulations of a hydrated protein. I demonstrate that this approach is capable of revealing microscopic details of the collective dynamics of a protein hydration shell.