Stereo 3D cloud motion from tandem satellites

In this study, we introduce a software framework developed during the SWI 2023 Study Group for accurate geolocation and 3D reconstruction of cloud observations. Jointly proposed by the European Space Agency (ESA) and the Royal Netherlands Meteorological Institute (KNMI), this project extracts and transforms cloud observation data into actionable insights. Implemented primarily in Python3 with libraries such as numpy, matplotlib, and scipy, the framework provides a robust pipeline for processing netCDF data, including cloud observations and locations. It handles three main tasks: reconstruction, motion extraction, and shape analysis. Key procedures include 3D reconstruction, bulk motion extraction, shape analysis, cloud selection and smoothing, and optical flow. The focus is on accurately extracting cloud latitude and longitude from observed positions, considering the true height of the cloud surface.