Unsupervised Optical Classification of the Seabed Color in Shallow Oligotrophic Waters from Sentinel-2 Images: A Case Study in the Voh-Koné-Pouembout Lagoon (New Caledonia)

Research demonstrates that unsupervised optical classification using Sentinel-2 images can effectively distinguish seabed colors in shallow, oligotrophic waters, even in complex bathymetric settings. The study applied Lyzenga correction and spectral standardization to separate water column reflectance from seabed signals, revealing five distinct seafloor clusters when standardized data were used—an improvement over the three clusters identified without standardization. This method achieved over 75% explanatory power for each wavelength band and improved classification accuracy, highlighting its potential for remote sensing of biogeochemical indicators like chlorophyll-a concentration in fragile coastal ecosystems such as New Caledonia's lagoons.

Wattelez, G., Dupouy, C., & Juillot, F. (2022). Unsupervised Optical Classification of the Seabed Color in Shallow Oligotrophic Waters from Sentinel-2 Images: A Case Study in the Voh-Koné-Pouembout Lagoon (New Caledonia). Remote Sensing, 14(4), 836. https://doi.org/10.3390/rs14040836