Driven by fiscal constraints, increasing pressure exists to employ simulation for driving exploration into new and more effective methods for next generation modeling activities. The shorter timescales required as well as the need to rapidly incorporate operational elements demanded by these applications bring new challenges. In particular, cost reduction in Synthetic Natural Environment (SNE) development using parameterized scenario generation with compositional modeling is one of the core problems that need to be addressed. In particular, component-based development of complex SNE components requires domain-specific repository organization with intelligent component retrieval mechanisms along with reasoning methods for conflict identification before component assembly. Furthermore, integrating disparate SNE components needs sophisticated metadata interoperability analysis, diagnosis, and merge tools. In this paper, a brief overview of an ontology-oriented and web-based SNE retrieval and fusion framework is discussed. The proposed approach aims to support component-based SNE development lifecycle. Specifically, DARPA Agent Markup Language (DAML), the ontology language of Semantic Web, is proposed to uniformly specify and facilitate reasoning about the service profiles of environmental models as well as SNE data components. The framework builds on DAML and constitutes: (1) an ontology engineering subsystem, (2) an agent-based distributed metadata (ontology) broker infrastructure, built on Trident's Interchange^SE - Object Oriented Data Repository, for SNE publishing, discovery, and retrieval, (3) diagnosis methods for SNE component certification before publishing, and (4) compositional consistency analysis and verification methods for integrated SNE component assemblies.