Bridging Gaps in Satellite Observations of River and Delta Landscapes Using Image Warping

Satellite images are increasingly used to monitor changes in fluvial landscapes such as channel migration, bar development, and avulsion. Yet, spatial and temporal gaps in image data - due to intervals between observations, cloud cover, or sensor malfunctions - limit their applicability. This study tests whether image warping, a well-established technique that generates smooth transitions between images, can bridge these gaps in observations of fluvial landscapes. The approach applies several steps to satellite images: (a) creating channel masks, (b) delineating channel topology, (c) correlating topologic components and defining control points, (d) calculating transformations between image pairs using control points, and (e) warping channel masks based on these transformations. This approach produces intermediate channel configurations from pairs of input images. We apply this technique to two case studies: the rapidly migrating Ucayali River in Peru and the actively prograding Wax Lake Delta in Louisiana. The reconstructions show channel migration and/or progradation that are consistent with actual observations, indicating reasonable estimates to fill gaps between satellite observations. Importantly, the accuracy of the reconstructions depends on the careful selection of the input images to avoid abrupt geomorphic changes that could compromise the warping process. By combining the surface reconstruction for the Ucayali River with a simple model for channel and floodplain aggradation, we construct a stratigraphic model directly informed by the satellite observations. These case studies suggest that applying image warping to Earth-surface observations has the potential to overcome gaps in satellite data availability and translate observations directly to models for fluvio-deltaic stratigraphy.