Data and task management for the on-demand processing of large-scale remote sensed imagery

We describe a prototype system that can provide automated generation of low-level calibrated remote sensed imagery and user configurable on-demand product generation for higher processing levels. The system offers full reproducibility of user orders and processing requirements by employing a relational database and a cluster-based parallel processing system. Such structure allow users to subscribe for products matching specific level and receive them as soon as the data products become available.