Answers are grounded in this paper and cite their sources.
The seabed mapping can be employed as a tool for the assessment of the impact of boat anchors on seagrass meadows. The hyperspectral sensor DESIS (DLR, Germany), having both a medium pixel size (30 m), a wide spectral range (400–1000 nm) and a spectral sampling interval of 2.55 nm. The goal of this study is to evaluate the performance of DESIS sensor for the retrieval of the bathymetry and the composition of benthic habitats in coastal shallow waters and to analyze the influence of the spectral sampling interval (2.55 or 10 nm) on the quality of the retrievals. The inversion of the DESIS-derived water reflectance is performed using a semi-analytical model. The performance of the retrieval of the bathymetry is validated using a well-proven bathymetric model developed by French navy (SHOM), namely Litto3D. The abundances of benthic species are compared with in-situ data that were acquired using a Go-Pro camera fixed on a Remoted Operated Vehicle. The bathymetry is retrieved from DESIS data with a root mean square error (RMSE) of 1.69 m for a seabed depth up to 10 m for a spectral sampling interval of 10 nm and increases up to 2.19 m for a spectral sampling interval of 2.55 nm as a result of an increased noise in the signal for such a better spectral sampling interval. The RMSE errors obtained when retrieving the proportion of benthic habitat species when the bottom depth is 10 m are 37% and 4%, respectively for a spectral sampling interval of 10 nm. The errors increase up to 46% and 20% for a spectral sampling interval of 2.55 nm. A pixel size of 30 m could be also a limitation to satisfactorily retrieve benthic composition and bathymetry. Overall, this study confirms the high interest and benefit of hyperspectral satellite sensors for improving knowledge of aquatic ecosystems in shallow waters.
No takes yet. Share an insight, caveat, or question.
Minghelli et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: