Enhancement of a canopy reflectance model for understanding the specular and spectral effects of an aquatic background in an inundated tidal marsh

The presence of water produces unique specular and spectral characteristics in an inundated tidal marsh canopy. The aquatic substrate can affect conventional attempts to retrieve canopy characteristics, such as structure information (e.g., canopy height, leaf area index, etc.) or plant species composition. The background reflectance can also influence spectral analysis of plant characteristics based on hyperspectral data. A model to account for the aquatic substrate would be useful to understanding spectral field measurements and remote sensing of this type of land cover. To that end, an existing vegetation canopy reflectance model is combined with an aquatic background model to account for the effects of an aquatic substrate on the top-of-canopy bidirectional reflectance. The aquatic background model attempts to account for the optical effects of an inundated marsh substrate through the inclusion of first-principle models of water reflectance. The enhanced model is applied to multi-angular reflectance measured along transects of a brackish marsh canopy. This allows us to explore whether the enhanced model can be used in retrieving the leaf area index (LAI) using non-destructive, above- canopy measurements. Then the original and the enhanced canopy reflectance models are compared with multi-angular reflectance data to test whether the change is effective in capturing specular effects of an inundated canopy. Furthermore the reflectance data and model are used to identify the influence of the background on the spectral characteristics of the canopy pertaining to vegetation. The spectral signature produced by the aquatic background model is quite different from the spectra of dry or unsaturated soil, which would be associated with terrestrial applications. The aquatic background model signature is used to explain the features seen in a field spectroscopy experiment, where canopy inundation levels were artificially raised. This project demonstrates the utility of developing a vegetation canopy model with an aquatic background and identifies challenges and directions for improved performance