An inter-agency collaborative computing framework for fast flood mapping using distributed remote sensing data

During flood disasters, obtaining a complete view of the entire flood area is an urgent issue. Satellite remote sensing can provide Earth observation (EO) images for the flood-influenced region. Multi-source EO data can enhance the temporal resolution and provide cross-validation. But EO data are commonly acquired and archived by different space agencies. The current flood disaster mitigation procedure includes several steps of EO data preparation and processing that requiring much manual effort. In this paper a near real-time flood mapping system for automatic and rapid flood mapping based on online EO data is presented. This paper provides the solutions for distributed EO data access and integration, automatic and fast EO data processing, fast flood visualization with common visual format. The collaborative computing framework provides an interoperable architecture to integrate heterogeneous data and computing resources. The system can automatically process satellite imagery to detect the extent of flooding in a near-real-time manner based on workflow engine. Inter-agency computing resources and algorithms were harnessed in the collaborative environment. Open-source GIS toolkits were used to implement and develop the components of the system. The experiment proves that the framework is quite efficient for flood mapping based on EO data. A flood event in Southeast Asia was selected to test the validate and evaluate the system.