Coupling methodologies for environmental models

Environmental models have been developed to address narrowly defined issues. As these issues become broader and scientific understanding grows, models become more complex. A variety of methodologies couple environmental models to address multifaceted issues, such as cross-media effects, risk analysis, and environmental justice investigations. This study analyzes, from a modeler's perspective, various coupling methodologies according to their characteristics. This analysis is presented as a five-level coupling hierarchy, and includes a representation of a modeling framework for coupled environmental models requiring dynamic feedback during a simulation (e.g., multimedia modeling). The framework addresses the special needs for this type of modeling, including spatial locking for environmental data, autogeneration of simulation metadata, simulation verification, and automated recovery from system problems. (C) 2000 Elsevier Science Ltd. All rights reserved.