Time-varying volumetric data arise in a variety of application domains, and thus several techniques for dealing with such data have been proposed in the literature. A time-varying dataset is typically modeled either as a collection of discrete snapshots of volumetric data, or as a four-dimensional dataset. This choice influences the operations that can be efficiently performed on such data. Here, we classify the various approaches to modeling time-varying scalar fields, and briefly describe them. Since most models of time-varying data have been abstracted from well-known approaches to volumetric data, we review models of volumetric data as well as schemes to accelerate isosurface extraction and discuss how these approaches have been applied to time-varying datasets. Finally, we discuss multi-resolution approaches which allow interactive processing and visualization of large time varying datasets.