The practical integration of climate and weather models in the curriculum of undergraduate and graduate courses at universities is the focus of the VAST system. It provides a working environment for model experiments which can be used for a wide range of applications beginning with the first steps in the field of climate modelling until the production phase of scientific projects. The environment combines the compute resources with an easy-to-use webbased configuration of the models.

A unified interface facilitates the handling of the models. The computeserver hosts the model installations and is linked in a network with the webbased user interface and the database server, the hearts of the virtual laboratory VAST. In the database different setups, called instances, can be managed. Each instance defines a role-workflow-matrix, linked the models to the courses and users.

In a typical configuration the following roles are available: The administrator builds the network of compute nodes, registers the models and assigns roles. The modeller configures scientific questions and the default parameter setup of the models. The docent combines the questions in courses and enroles the students into the courses. The tutor supervises the courses, e.g. unblock the model questions for students. The student executes the experiments and analyses the results.

The VAST system is fully modularized. The model installations can be used in standalone mode, mainly for testing, and inside the VAST network, controlled with the web-frontend. The concept of the role-worklow-matrix can be adjusted to different use cases. Since September 2015 the system is in production phase and was used for graduate schools and master courses in meteorology. The list of available models ranges from simple models: an energy balance model and the PUMA; models of intermediate complexity: PLASIM and CLIMBER; regional models: COSMO-CLM; to complex models: ECHAM5 and COSMOS1.

 

Slides to this talk