Keywords
environmental modelling, web services, modeling services, environmental modeling languages
Start Date
1-7-2012 12:00 AM
Abstract
Environmental modeling projects consider increasingly complex systems and thus require integration of skills of several people and information from several sources. To overcome problems that modelers face in such projects calls for new approaches. We define “distributed environmental modeling” as the organized use of Internet technology for environmental modeling. This paper examines the hypothesis that distributed environmental modeling has two main forms: collaborative modeling and use of shared resources. We examine this hypothesis in the light of experiences obtained from case studies. We present a case study where a prototype was developed for shared development of Bayesian networks. We present and discuss other cases where we have explored services for modeling and ways to link such services together for modeling. A general conclusion is that it is a benefit to be able to work with text files and/or formal structured descriptions of the objects and tasks. This kind of data lends itself well to be used with existing tools like databases and version control systems. Objectorientation allows the division of the overall task into parts, which can then be integrated in the distributed system level. Another conclusion is that standardization efforts regarding high-level environmental modeling languages would be useful from the viewpoint of distributed modeling. However, another side of distributed modeling is the change in modeling paradigm, which may require changes in mindsets.
Distributed Environmental Modeling
Environmental modeling projects consider increasingly complex systems and thus require integration of skills of several people and information from several sources. To overcome problems that modelers face in such projects calls for new approaches. We define “distributed environmental modeling” as the organized use of Internet technology for environmental modeling. This paper examines the hypothesis that distributed environmental modeling has two main forms: collaborative modeling and use of shared resources. We examine this hypothesis in the light of experiences obtained from case studies. We present a case study where a prototype was developed for shared development of Bayesian networks. We present and discuss other cases where we have explored services for modeling and ways to link such services together for modeling. A general conclusion is that it is a benefit to be able to work with text files and/or formal structured descriptions of the objects and tasks. This kind of data lends itself well to be used with existing tools like databases and version control systems. Objectorientation allows the division of the overall task into parts, which can then be integrated in the distributed system level. Another conclusion is that standardization efforts regarding high-level environmental modeling languages would be useful from the viewpoint of distributed modeling. However, another side of distributed modeling is the change in modeling paradigm, which may require changes in mindsets.
