A water-depth correction algorithm for submerged vegetation spectra

Abstract

The substantial difference between the reflectance values at the red and the near-infrared (NIR) regions has been used to develop vegetation indices for remote sensing of green plants. These spectral indices, however, may not effectively be used for underwater plants because the overlying water interferes with the upwelling vegetation signals. We empirically separated the energy absorbed by water and scattered from the water column using an indoor water tank with hypothetical surfaces that either reflect or absorb all the incoming light. Using the experimental data, a function was developed to correct reflectance measured from a shallow water body for the water effects. When applied to independently measured reflectance of underwater vegetation, the algorithm significantly enhanced vegetation signals, especially in the NIR region. This experimentally driven algorithm can improve mapping capabilities of seagrass beds and invasive aquatics in shallow water bodies.