Abstract
The article presents a new hybrid bio-optical transformation (HBT) method for the rapid modelling of bathymetry in coastal areas. The proposed approach exploits free-of-charge multispectral images and their processing by applying limited manpower and resources. The testbed area is a strait between two Greek Islands in the Aegean Sea with many small islets and complex seabed relief. The HBT methodology implements semi-analytical and empirical steps to model sea-water inherent optical properties (IOPs) and apparent optical properties (AOPs) observed by the Sentinel-2A multispectral satellite. The relationships of the calculated IOPs and AOPs are investigated and utilized to classify the study area into sub-regions with similar water optical characteristics, where no environmental observations have previously been collected. The bathymetry model is configured using very few field data (training depths) chosen from existing official nautical charts. The assessment of the HBT indicates the potential for obtaining satellite derived bathymetry with a satisfactory accuracy for depths down to 30 m.
Highlights
Multispectral imagery provides an alternative way to acquire marine environmental data for several operational, as well as scientific, purposes [1,2,3,4], in different time and spatial scales
The range of extracted depths is strongly related to the water clarity, or the so-called ocean color (IOPs, turbidity, etc.) The main issue is that the ocean color depends directly on the optical properties of the water, which in turn vary spatially and temporally [6]
We further investigated the effects of inherent (IOPs) and apparent optical parameters (AOPs) on depth modelling
Summary
Multispectral imagery provides an alternative way to acquire marine environmental data for several operational, as well as scientific, purposes [1,2,3,4], in different time and spatial scales. Satellite imagery can provide users with adequate coverage and often free-of-charge environmental information, like those acquired by the Landsat 8 (OLI) and Sentinel-2 sensors [5]. The acquisition of spaceborne observations is a common practice, the processing of imagery data to extract bathymetry is a more complex procedure, since it requires special treatment for estimating the subsurface irradiance reflectance signal (R(0−)). After pre-processing the imagery data with respect to atmospheric interference (i.e., via models such as 6SV, OPERA, etc.) [7,8], and air–water surface interaction, the remaining signal propagating from the seabed can provide the bathymetric information. Empirical or analytical-based inversion models are applied for obtaining satellite-derived bathymetry (SDB) [9,10]
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call