Air Quality Forecasting on Campus Grid Environment

Air Quality Forecasting (AQF) is a new discipline that attempts to reliably predict atmospheric pollution. The application has complex workflow and in order to produce timely and reliable forecast results daily, each execution requires multiple computational and storage resources to be simultaneously and collaboratively available. Deploying AQF on grid is one option to satisfy such needs, but requires related grid middleware to support automated application-specific scheduling and execution on grid resources. This paper presents our initial experience of deploying AQF on a campus grid environment and our current efforts of developing a solution of grid-enabling AQF-like applications in Gracce project. Gracce has the goal to provide domain users a grid platform supporting from the management of an application and its dataset, to the automatic execution and viewing of results. In Gracce, application workflow is described using GAMDL, a powerful data-flow language for domain users in describing application logics. The Gracce metascheduler architecture, which includes a workflow-orchestrated metascheduler, an event-driven workflow engine, and an execution runtime system provides the required functionalities of scheduling application workflow in global level and coordinating workflow executions.