Project Leads: Eli Holmes, Northwest Fisheries Science Center. Troy Anthony Russo, UW Statistics & Mathematics. Kaira Nair, UW Computer Science
Data Scientists: Valentina Staneva and Scott Henderson
Ocean color (OC) remote sensing is critical for understanding ocean productivity and marine ecosystems, and the new hyperspectral satellite missions such as NASA PACE greatly expand its applications. However, OC sensors cannot penetrate clouds, leaving large data gaps across the oceans. We developed and validated a deep-learning gap-filling approach using the data themselves to create training data. Our project scales this method to and an operational workflow from a regional prototype to a global product.

