Download Now Free registration required
Cloud computing has emerged as one of the preferred platforms for on-demand, event-driven computing. However, literature on event-driven scientific modeling on the cloud is sparse. This paper presents the authors' vision and architecture to provide a "Scientific modeling as a service" in an on-demand fashion. They choose one of the most widely used groundwater models, MODFLOW2000, from USGS to develop a ModflowOnAzure service on the Windows Azure platform as the first implementation. Some specific issues on handling scientific models in Azure are presented including file transfer and synchronization through the dropbox API and provenance tracking in the cloud using the open provenance model.
- Format: PDF
- Size: 342.7 KB