A physics-based approach to estimate water column depth: from different remote sensors in varying coastal environments with incomplete site-specific input data
The ability to accurately estimate water column depth over large coastal areas and/or in remote locations such as reefs is directly relevant to environmental management, exploration, defence and research applications. The potential to use remote sensing for this purpose has been mostly limited to empirical approaches that are not transferable across study areas or data types, are not objective (repeatable), and offer limited accuracy. Here, we present four case studies using an objective, radiative transfer-based approach, SAMBUCA (Semi-Analytical Model for Bathymetry Un-mixing and Concentration Assessment), to produce bathymetry maps using data from four different sensors, in four different locations, and using varying amounts of site-specific data. Water column depth was retrieved from CASI, Quickbird, Hyperion, and MERIS data in environments ranging from coastal, sediment-dominated environments to open ocean coral reef settings. The SAMBUCA approach was the same for all the studies: a semi-analytical model for underwater light fields is parameterised with optical properties of the system and solved using an optimisation-minimisation routine. Notwithstanding, the four studies posed differing sets of challenges, including excessive processing time, insufficient spatial or spectral resolution, and high levels of target-atmosphere-system noise. These challenges were addressed using objective and repeatable methodologies. The results are validated qualitatively and/or quantitatively, depending on the availability of field validation data.