Ocean color remote sensing: choosing the correct depth weighting function

Values of reflectance and remote sensing reflectance are proportional to the ratio of sea water backscattering to absorption. However, in vertically non-homogeneous waters, this fraction needs to be depth weighted. The usual practice uses normalized vertical transmittance profiles as the weighting function. Recently, it was shown that the correct approach is to use, instead of transmittance, its first derivative. We used both approaches to calculate spectral reflectance and remote sensing reflectance over a submerged bubble cloud and chlorophyll rich layer and compared the results with a radiative transfer Monte Carlo code. We also compared several methods of approximating diffuse attenuation (not measured directly) to estimate the effect on calculating reflectance. Our results show that the traditional method of IOP weighting is inadequate in the presence of bubble clouds and/or chlorophyll rich layers. This is relevant for both "ground truth" studies and inverse methods of remote sensing (including lidar ones) for vertically inhomogeneous ocean sea waters.