Subsurface Remote Sensing of Kelp Forests

Macrocystis pyrifera, or giant kelp, provides structure and support for many marine species, and its forests rank as one of the most ecologically productive systems in the world. Traditional, in situ measurements of kelp biomass and productivity are episodic, costly, and provide limited spatial coverage across the often wide swaths of kelp ecosystems. While satellite methods have been developed to estimate kelp biomass and productivity, satellite observations are also limited, as standard practices for measuring terrestrial vegetation cannot be applied with the same confidence to marine vegetation. Here, data gathered from flights with the MASTER sensor over the Santa Barbara Channel allowed the development of two algorithms to assess the surface and subsurface areal extent of kelp in multispectral imagery. The first, a marine vegetation index (MVI), was developed from imagery to capture both surface and sub-surface vegetation pixels. The second algorithm is based on a spectral library for kelp radiance collected from field samples and modeled using the radiative transfer equations with the HydroLight software package. The endmember collection from this library was used in the Spectral Angle Mapping tool in ENVI to identify kelp at various depths. Outputs from each of these algorithms were then compared to the Normalized Difference Vegetation Index (NDVI). Analyzing spectral properties of sub-surface features will facilitate the use of satellites in measuring extent and productivity of marine ecosystems. Furthermore, these tools allow researchers to directly quantify the depth and extent of subsurface vegetation, greatly enhancing existing methods.