Monitoring and Benchmarking Geospatial Workflows for Efficient and Reproducible Earth Observation

The rapid growth of Earth Observation data from satellites, drones, and other sensors has led to increasingly complex geospatial workflows. These workflows often involve multiple software components, heterogeneous computing environments, and large data volumes. As a result, performance bottlenecks and inefficient use of computing resources can significantly impact research productivity and reproducibility. This lecture introduces the fundamental concepts needed to understand and evaluate the performance of geospatial workflows. The session begins with an overview of typical EO computing environments, from laptops and workstations to high-performance computing clusters and cloud platforms, and discusses why simply scaling resources is often insufficient without proper analysis of workflow behavior. Participants learns how to interpret key system metrics, such as CPU usage, memory consumption, and disk I/O, and how operating system tools can be used to observe these metrics during workflow execution. The lecture then introduce the concept of benchmarking as a systematic method for measuring and comparing workflow performance. Through real examples and discussion, participants learns how benchmarking can support different research contributions, from methodological comparisons to state-of-the-art algorithm development.