Integrating multidisciplinary models requires linking models: that may operate at different temporal and spatial scales; developed using different methodologies, tools and techniques; different levels of complexity; calibrated for different ranges of inputs and outputs, etc. On the other hand, integration of models requires us to address technical, semantic, and dataset aspects of interoperability. So we need a genuine techniques that enable us to integrate various domain specific models for interdisciplinary study. In this research work, we investigated best practices of System Integration, Enterprise Application Integration, and Integration Design Patterns. We developed an architecture of a multidisciplinary model integration framework that brings these three aspects of
integration together. Service-oriented-based platform independent architecture that enables to establish loosely coupled dependency among various models is presented.