Keywords
open modelling interface, linking, model interoperability, integrated modelling
Start Date
1-7-2010 12:00 AM
Abstract
The Open Modelling Interface (OpenMI) was launched end of 2005 with the aim to become a global standard for linking models and tools in the environmental domain with focus on water. Over the past few years, the user and development community has grown substantially and various well known models have become compliant. Some of the uses did not adopt the OpenMI.Standard interfaces completely, but used a slight deviation to achieve their goal in a similar style. Improvements would be necessary to become a true global interface standard instead of a style for developing new model codes. Starting in 2007, a core group of six institutes has worked on an upgrade of the OpenMI towards version 2.0. A long list of deficiencies was composed, having a few use cases as general guidance for improvement. The proposed redesign, based on similar leading concepts and a similar data model, required however a non-backward compatible upgrade of the interface standard to remove the weak points from the first version. This decision allowed the OpenMI to become more suitable for a larger range of applications, from non-time dependent Geographical Information Systems (GIS) towards e.g. master-slave controlled modelling frameworks. The OpenMI 2.0 standard has been open for review early 2010. Once completed and processed, the release of OpenMI 2.0, in both C# and Java is expected late 2010. This paper will discuss the reasons for change in more detail and highlights how the proposed solution meets the needs in a better way.
From OpenMI 1.4 to 2.0
The Open Modelling Interface (OpenMI) was launched end of 2005 with the aim to become a global standard for linking models and tools in the environmental domain with focus on water. Over the past few years, the user and development community has grown substantially and various well known models have become compliant. Some of the uses did not adopt the OpenMI.Standard interfaces completely, but used a slight deviation to achieve their goal in a similar style. Improvements would be necessary to become a true global interface standard instead of a style for developing new model codes. Starting in 2007, a core group of six institutes has worked on an upgrade of the OpenMI towards version 2.0. A long list of deficiencies was composed, having a few use cases as general guidance for improvement. The proposed redesign, based on similar leading concepts and a similar data model, required however a non-backward compatible upgrade of the interface standard to remove the weak points from the first version. This decision allowed the OpenMI to become more suitable for a larger range of applications, from non-time dependent Geographical Information Systems (GIS) towards e.g. master-slave controlled modelling frameworks. The OpenMI 2.0 standard has been open for review early 2010. Once completed and processed, the release of OpenMI 2.0, in both C# and Java is expected late 2010. This paper will discuss the reasons for change in more detail and highlights how the proposed solution meets the needs in a better way.
