A Preliminary Assessment of a Newly-Defined Multispectral Hue Space for Retrieving River Depth with Optical Imagery and In Situ Calibration Data

Bathymetry is a key element in the modeling of river systems for flood mapping, geomorphology, or stream habitat characterization. Standard practices rely on interpolation situ depth measurements obtained with differential GPS total station surveys, while more advanced techniques involve bathymetric LiDAR acoustic soundings. However, these high-resolution active are not so easily applied over large areas. Alternative methods using passive optical imagery present an interesting trade-off: they fact that wavelengths composing solar radiation attenuated at same rates water. Under certain assumptions, logarithm ratio radiances two spectral bands linearly correlated depth. In this study, we go beyond defining multispectral hue retains all information. Given n coregistered bands, invariant lies (n−2)-sphere embedded Rn−1, denoted Sn−2 and tagged ‘hue hypersphere’. It can be seen as generalization RGB ‘color wheel’ (S1) higher dimensions. We use mapping to identify hue-depth relation 35 km reach Garonne River, high resolution (0.50 m) airborne four data from 120 surveyed cross-sections. The distribution pixels modeled mixture components: one component represents substrate hue, other ‘deep water’ hue; parameters fitted such membership probability ‘deep’ correlates