Measuring marine hydrodynamics from space using planet satellite imagery

Abstract

The inference of coastal ocean dynamics from consecutive remote sensing images plays a central role in a diverse range of domains such as marine conservation, spatial planning, as well as flood risk. We present a methodology for systematically identifying spatially overlapping image pairs from the PlanetScope archive, with order minute scale time lags and the potential for velocity field inference using classical algorithms. This ability is demonstrated through the novel estimation of submesoscale eddies from PlanetScope image pairs in a range of contexts, providing a key novelty in this paper. These include sea ice floes in the Siberian Sea of Okhotsk, a cyanobacterial bloom in the Baltic Sea, and suspended sediment in the Port of Al-Fao located in the Arabian/Persian Gulf. Additionally, comparison of the latter with coinciding velocity fields from a Delft3D Flexible Mesh (FM) numerical model simulation shows good quantitative agreement in regions with high suspended sediment concentration. We successfully develop a workflow pipeline for identifying and processing image pairs from these opportunistic overlaps, unlocking a new large-scale source of coastal ocean surface velocity data to be used alongside modelling frameworks.