To date, the systems approach to science, which emphasizes the connections among phenomena studied at different scales and by different disciplines, is causing dramatic changes in how scientific results are communicated. These changes drive a shift on how to propagate data with certain properties so that it can be used intelligently by others. In this work we elaborate on three major factors governing the propagation module of data provenance. The proposed propagation module provides an efficient solution for many critical problems in the management and provenance of scientific data. Unlike previous work, our work aims at realizing data provenance from models to storage. The natural representation of data as objects and its utility for capturing provenance has led us to consider a new storage architecture based on the object-based storage (OSD) technology. An outline of this framework is discussed.