Linking Data, Models and Tools: An Overview

Complex questions trigger researchers around the world to link data, numerical models and tools of different origins together for integrated modeling of the environment and related socio-economic fields. Different researchers have, however, chosen different ways to link such resources. As a result a seemingly wide range of interfaces and frameworks have been defined: some have used low level interfaces others more abstract and object oriented ones; some systems may require little or no code changes whereas others promote to fundamental rewriting of your code. So, if we all try to link data, models and tools, why are we then using such different environments? The basic idea is that the various approaches address supposedly conflicting demands like generality, flexibility, ease of use, accuracy and performance. What are the benefits of the various approaches? This paper addresses these questions by looking at a.o. common component architecture (CCA), earth system modeling framework (ESMF), FRAMES, object modeling system (OMS) and OpenMI. Are they really conflicting or do they to a large degree complement each other?