A purely spaceborne approach for nearshore bathymetry monitoring

Nearshore bathymetric monitoring is critical for understanding and quantifying the blue economy, yet in-situ surveys are costly and time-consuming. While remote sensing has been able to overcome some of these limitatons, a purely spaceborne approach has so far not been achievable. Here, we provide a method for combining spaceborne optical imagery and ICESat-2 spaceborne lidar to generate an end-to-end method of nearshore bathymetric modeling. Discrete benthic depths are extracted from ICESat-2 data using custom open source software (C-SHELPh) and combined with optical imagery using Machine Learning algorithmsto create spatially continuous bathymetric models for a range of study sites distributed globally. We demonstarte an ability to generate national scale bathymetric models at high spatial resolution (2 - 30 m) and with low error (6%), outperforming existing interpolated multi-year surveys and providing new information in data scarce areas. In addition, our methods are readily repeatable, enabling time-series change detection and monitoring. Our open source approach provides both the research and applications communities with methods and products for modeling and monitoring nearshore coastal sub-aquatic structure, providing new transformative data to understudied portions of our oceans.