EURAC RESEARCH (IT)
ClouDInSAR addresses one of the main remaining barriers to the wider adoption of interferometric SAR (InSAR) techniques in Earth Observation analysis chains: the computational complexity and infrastructure demands required to process Sentinel‑1 Single Look Complex (SLC) data at scale. While Sentinel‑1 has been providing a continuous and global SAR data stream since 2014, the exploitation of its full interferometric potential—particularly for multi‑temporal techniques such as Persistent Scatterer Interferometry (PSI) and Small Baseline Subset (SBAS)—remains limited to users with access to substantial computing resources and specialized processing environments. In contrast to optical EO data, which are more easily integrated into common analysis frameworks, SLC‑based InSAR workflows involve complex processing steps, large data volumes, and non‑trivial data management challenges that prevent many users from exploiting these powerful techniques.
The ClouDInSAR project was developed to close this gap by designing and implementing a fully cloud‑native, open‑source InSAR processing framework integrated into openEO within the Copernicus Data Space Ecosystem (CDSE). The project foresees the execution of complete Sentinel‑1 interferometric workflows directly in the cloud, enabling users to perform advanced InSAR analyses without downloading data or managing complex processing chains locally. The solution emphasizes openness, reproducibility, and interoperability, ensuring that workflows are transparent, portable, and adaptable to a wide range of scientific and operational use cases.
Two representative use cases were defined to evaluate the suitability, robustness, and performance of the ClouDInSAR framework. The first use case focuses on land surface characterization through multi‑temporal coherence analysis. Long‑term coherence time series were generated for all glaciers in South Tyrol, processing 25 Sentinel‑1 bursts across five acquisition geometries. Persistent coherence patterns associated with stable, moving debris surfaces allowed the automated detection of debris‑covered glacier areas. The resulting masks were validated against existing glacier inventories derived from high‑resolution orthophotos and LiDAR data, demonstrating that cloud‑native coherence time series are reliable and suitable for operational cryosphere monitoring.The second use case targets surface deformation monitoring using interferogram time series. The ClouDInSAR workflow was applied to generate consistent interferometric stacks, with particular attention to ensuring technical compatibility with MintPy, a widely used tool for multi‑temporal InSAR analysis. The interferograms produced in the openEO environment were successfully ingested into MintPy, confirming that the cloud‑generated outputs meet the requirements for downstream PSI and SBAS processing. This demonstrates the ability of ClouDInSAR to support hybrid cloud‑desktop analysis workflows and to interoperate with established InSAR software ecosystems.
ClouDInSAR is defined to satisfy two closely linked objectives. The first is to significantly lower the entry barrier for advanced InSAR analyses by providing users with easy access to scalable, cloud‑native processing capabilities directly where Sentinel‑1 data are hosted. By abstracting infrastructure complexity and offering standardized building blocks, the project promotes the integration of InSAR into everyday EO analysis chains for both research and operational applications. The second objective is to ensure computational efficiency by balancing storage, processing cost, and performance. Rather than pre‑computing and storing exhaustive interferometric products, ClouDInSAR emphasizes on‑demand, user‑driven processing, ensuring flexibility while keeping resource usage under control. Overall, ClouDInSAR demonstrates that fully cloud‑native InSAR processing within openEO and CDSE is not only feasible but also highly effective. By extending openEO into a complete interferometric processing environment, the project establishes a robust and scalable foundation for next‑generation SAR applications, supporting reliable deformation monitoring, surface characterization, and large‑scale, reproducible InSAR analyses.
