A multi-spectral confocal high-resolution LIDAR for in vivo underwater classification of zooplankton and microplastics

The productivity of zoo and phytoplankton in aquatic ecosystems is vital on local and global scales and is tightly linked to environmental factors. The distribution and abundance of these organisms are shifting due to climate change, impacting food web transfer and food security. In situ monitoring is challenging, with existing techniques limited to small close-range volumes and often requiring vertical dives. We introduce a novel multi-spectral Light Detection and Ranging (LiDAR) system for underwater remote sensing. We demonstrate inelastic detection of sub-millimeter microalgae and free-swimming copepods, classifying them using machine learning. The sensor non-intrusively captures high-resolution volumetric images of microorganisms with voxel-level spectral data, revealing species-specific autofluorescence features used for classification. Furthermore, we accomplish depth-resolved measurements and 3Dreconstruction of microplastic layers in the water column with suppression of out-of-focus contributions from intervening scatterers that would otherwise degrade the performance of conventional imaging technologies. Our approach combines morphology with spectral information, representing a unique and non-intrusive tool for quantitative studies of plankton.