Keywords
Peer-to-peer ontology; federated model sets; integrated environmental modelling; model integration.
Location
Session B1: Research Infrastructures for Integrated Environmental Modeling
Start Date
16-6-2014 10:40 AM
End Date
16-6-2014 12:00 PM
Abstract
Integrated modelling is a critical environmental modelling tool that brings together sets of models and data to describe the impacts of various management practices and choices within the environment. Although each model may (or may not) run on separate hardware, typically these model sets are organized together as a large integrated, centralized model. These integrated models share a single conceptual schema and semantic framework. Additionally, information flow and convergence of boundary conditions between the models are generally built on a fixed framework. Although this is a logical approach that has proven to be effective, it has several limitations. The models need to be managed and coordinated by a centralized administrator, and adding new models to the model set or revising the existing models can be time consuming and challenging. Moreover, as larger, more complex systems are modelled across various scales, the management and coordination challenges increase geometrically. This paper describes a proposed decentralized approach to the coordinated sharing and exchange of information between autonomous models. The proposed approach consists of independent models and model systems loosely federated together to represent various environmental choices. As envisioned here, a federated model consists of self-describing information entities (e.g., models, data, and user input); a library that contains the set of available models and manages the model development community; a federation schema that defines the purpose of the federation, describes the models used in the federation, and defines the topology of the federation; and a federation management system that coordinates information transport, convergence of boundary conditions, and error.
Included in
Civil Engineering Commons, Data Storage Systems Commons, Environmental Engineering Commons, Hydraulic Engineering Commons, Other Civil and Environmental Engineering Commons
A Proposed Approach to the Development of Federated Model Sets
Session B1: Research Infrastructures for Integrated Environmental Modeling
Integrated modelling is a critical environmental modelling tool that brings together sets of models and data to describe the impacts of various management practices and choices within the environment. Although each model may (or may not) run on separate hardware, typically these model sets are organized together as a large integrated, centralized model. These integrated models share a single conceptual schema and semantic framework. Additionally, information flow and convergence of boundary conditions between the models are generally built on a fixed framework. Although this is a logical approach that has proven to be effective, it has several limitations. The models need to be managed and coordinated by a centralized administrator, and adding new models to the model set or revising the existing models can be time consuming and challenging. Moreover, as larger, more complex systems are modelled across various scales, the management and coordination challenges increase geometrically. This paper describes a proposed decentralized approach to the coordinated sharing and exchange of information between autonomous models. The proposed approach consists of independent models and model systems loosely federated together to represent various environmental choices. As envisioned here, a federated model consists of self-describing information entities (e.g., models, data, and user input); a library that contains the set of available models and manages the model development community; a federation schema that defines the purpose of the federation, describes the models used in the federation, and defines the topology of the federation; and a federation management system that coordinates information transport, convergence of boundary conditions, and error.
