Data provide a lens through which we see the world, but the interpretation of what we see depends crucially on how data are collected, and on modeling assumptions and other decisions we make analyzing them. The climate system itself can be conceptualized in terms of hierarchies of interacting processes acting on different scales in space and time. If analyses of climate data are to help improve understanding of these multiscale physical processes, then data should also be viewed within a commensurate framework.
Data hierarchies are not separate from equation, model, or simulation hierarchies. Rather, they provide a mechanism for examining experimental or observational evidence in order to evaluate and improve them. In this workshop, we will examine 1) basic paradigms for modeling hierarchical relationships, both from a statistical viewpoint and within the dynamical systems approach, 2) the application of these paradigms to facilitate the formulation of hierarchies for understanding climate processes, 3) their application specifically to equation, model, and simulation hierarchies given a priori, 4) quantification and propagation of data-based modeling errors and uncertainties through the hierarchies, and 5) interdisciplinary issues arising from the massiveness of data collected or generated in modern climate science.