Publication
An Empirically Derived Trimodal Parameterization of Underwater Light in Complex Coastal Waters – A Case Study in the North Sea
Jochen Wollschläger; Beke Tietjen; Daniela Voß; Oliver Zielinski
In: Frontiers in Marine Science (FMarS), Vol. 7, No. 512, Pages 1-16, Frontiers Media SA, Lausanne, 6/2020.
Abstract
As an essential parameter for all kinds of aquatic life, light influences life cycles and the behavior of various marine organisms. However, its primary role is that of a driver for photosynthesis and thus primary production, forming the basis of the marine food web. As a simplification when dealing with light, a common measure (e.g., used in biogeochemical models) is the photosynthetically active radiation (PAR), which integrates the spectral distribution of photon flux between 400 and 700 nm into a single value. While passing through the water column, light is attenuated by the water itself and its optically active substances (OAS) [e.g., phytoplankton, chromophoric dissolved organic matter (CDOM), and non-algal particles], summarized in the diffuse attenuation coefficient of downwelling radiation (Kd). Existing parameterizations for light attenuation in models often consider only phytoplankton as parameter, which is not sufficient for coastal areas where the contributions of CDOM and suspended mineral particles can be substantial. Furthermore, they mostly ignore the spectral variability of Kd by attenuating PAR with only a single coefficient. For this reason, this study proposes a parameterization of Kd that involves all relevant OAS and that attenuates PAR in three bands (trimodal approach). For this, the hyperspectral underwater light field was examined on three expeditions in different areas of the North Sea and along the British and Irish coasts. The derived Kd spectra were stepwise decomposed in the contributions of the different OAS and used in combination with direct OAS measurements to derive substance specific attenuation coefficients for the three bands. For comparison, also a monomodal and a spectral parameterization were developed. Evaluation showed that the trimodal approach was almost as accurate as the full spectral approach, while requiring only marginally more computational performance as the classical monomodal approach. Being therefore an excellent compromise between these factors, it can act as a valuable, yet computational affordable addition to biogeochemical models in order to improve their performance in coastal waters.