Master: Sharing models? Feedback from the community.

• Supervisor: Fred Fonseca
• Contact Person (IFGI): Patrick Maué
• Contact Person (INPE): ?
• Responsible Student:

• Objectives:
  • Get feedback from the model community.
  • Develop, distribute, and analyze a questionnaire.
• Expected Results:
  • Requirements for the Model Web architecture
  • Input for GEOSS community
• Mobility Measure: IFGI → INPE

Sharing data needs information about the location of repositories and archival formats. Sharing models needs understanding about the applicability of each algorithm to the species being modelled; it also requires a good documentation explicit and implicit assumptions behind the model. Sharing results needs communication of the species distribution maps as well as producing reports and adding comments. [1]

If we assume that there is an interest to share models in the community for environmental modeling, we need to find out what their idea of a “model web” is. According to GEOSS, the model web is a “distributed, multidisciplinary network of independent, interoperating models called the Model Web. The goal is to create a framework under a set of broad goals and data exchange standards or guidelines that would grow organically, without central control, to serve the needs of many disciplines. Models contributed to this Model Web would be developed and operated by different researchers from a variety of disciplines including ecology, hydrology, air quality, climatology, the ocean sciences, etc.”

The idea is to develop a questionnaire addressing the modeling community. The feedback will be analyzed, and gives us hopefully an idea what requirements we should have for the model web, and which open research challenges we should focus on first.

[1] Sharing executable models through an Open Architecture based on Geospatial Web Services: a Case Study in Biodiversity Modelling (Karla Donato Fook, Silvana Amaral, Antônio Miguel Vieira Monteiro, Gilberto Câmara1, Marco A. Casanova)