Adaptive bathymetry estimation for shallow coastal waters using Planet Dove satellites

Accurate bathymetric mapping of shallow waters (above 15 m) is essential for a wide range of scientific research, government, transport, and industry globally. Satellite-based bathymetry estimation approaches offer an alternative to traditional shipborne measurements, especially given advancements in their spatial and temporal resolution of satellite imagery, including the new Planet Dove constellation of > 150 satellites providing daily coastal coverage. Dove satellites provide abundant cloud-free images even over cloudy tropical coastal environment, offering an opportunity to generate frequent bathymetry maps at high spatial resolution (4 m). We developed a new adaptive bathymetry estimation algorithm for Planet Dove and similar satellites. The algorithm adaptively tunes a depth estimator according to water column attenuation conditions. The algorithm was tested at five diverse reef sites globally (Lighthouse Reef, Belize; Saona Island, Dominican Republic; St. Croix, U.S. Virgin Islands; Heron Island, Australia; Hawaii Island, U.S.) using 31 satellite images from six Dove satellites. Derived water depth was validated (RMSE = 1.22 to 1.86 m) with field-measured sampling points (61,025) ranging in depth from 1 to 15 m. Algorithm performance was best at depths of 4-10 m. This new adaptive algorithm can be effectively applied to derive high spatial resolution bathymetric maps from Planet Dove satellite imagery across a wide range of conditions.