Cloud-based Processing and Delivery of Analysis-Ready SAR Datasets

When monitoring or responding to natural hazards or disasters, end users require data that is easy to access and understand. Synthetic Aperture Radar (SAR) sensors can image the earths surface regardless of cloud conditions or light availability, making it a valuable dataset for mapping surface processes during natural disasters, but processing SAR data to analysis-ready products requires specialized, computationally intensive techniques. Even analysis-ready products can be challenging to work with due to large file sizes, and difficult to interpret for users who are new to SAR. The Alaska Satellite Facility (ASF) hosts and serves the entire Sentinel-1 SAR archive from cloud storage. As part of ASFs mission to make SAR data more accessible, the HyP3 platform was developed to leverage cloud computing for generating analysis-ready products, which users can request on-demand through ASFs Vertex data portal. Because the processing occurs adjacent to the storage in the same cloud environment, costs and processing times are minimized. By publishing these analysis-ready HyP3 products to image services directly from cloud storage, end users can access and explore large datasets with extensive spatial and temporal coverage without having to download any data. The service can be packaged with default visualization settings or multiple visualization options, making the imagery easier to interpret. The data remains in the cloud environment from source to service, saving the user time and local storage space, and reducing ASFs egress costs. Implementing a fully cloud-based, automated workflow, starting with the source Sentinel-1 scenes and ending with an image service, ensures that the services are always displaying the most recently available data for end users to access. Users can interact dynamically with the datasets without having to download or manage large files, and interpretation of the information is facilitated by intuitive visualization settings.