You are here:
Hyperspectral Remote Sensing of New England Coastal Waters to Predict Seagrass Distribution
Citation:
Keith, D., G. Thursby, AND S. Rego. Hyperspectral Remote Sensing of New England Coastal Waters to Predict Seagrass Distribution. American Geophysical Union (AGU) Ocean Sciences Meeting, New Orleans, LA, February 21 - 26, 2016.
Impact/Purpose:
This study proves that spectral data acquired from the Hyperspectral Imager for the Coastal Ocean (HICO) on the International Space Station (ISS) can be transformed in water quality information in support of nutrient criteria development in coastal and estuarine systems. During his study, data products were input into a regionally calibrated seagrass bio-optical model to predict and map spatial and temporal patterns in seagrass distributoin based on light attenuation and water clarity for Narragansett Bay. The application of space-based information represents another example of how remotely sensed data can assist state and federal clients in sustainable management practices and inform the general public about water quality conditions.
Description:
The U.S. Environmental Protection Agency is working to improve its ability to quantify and predict aquatic (freshwater, estuarine, marine) ecosystem response and recovery to changing nutrient loads. The objective of this research is to quantify the relationship of nutrients with management response and recovery trajectories in aquatic ecosystems.In this study, we use light attenuation as a measure to determine the clarity of New England estuarine waters in response to nutrient loading. The clarity of these waters is important to the sustainability of healthy seagrass habitats which support the stability, nursery function, biochemical cycling and trophic dynamics of coastal ecosystems. We suggest that numeric nutrient thresholds and recovery trajectories can be derived for seagrass management based on light attenuation (Kd) retrieved from remotely sensed data.The Hyperspectral Imager for the Coastal Ocean (HICO) instrument onboard the International Space Station offered EPA the opportunity to model the quality of light in two southern New England estuaries. Using atmospherically corrected images, apparent optical properties (e.g., remote sensing reflectances (Rrs)) were retrieved and used to derive optical models which estimated chl a concentrations, CDOM absorption coefficients and TSS concentrations. These estimates were then input into a regionally calibrated seagrass bio-optical model which predicted the attenuation of surface light in the water column. Studies have indicated that the seagrass Zostera marina optimally grows at depths where approximately 20 per cent of surface light is available for photosynthesis. Using the surface attenuation values, the water depth at which the 20 per cent threshold occurs can be calculated. Combining the threshold depth information with bathymetric data, the locations for optimal conditions for Z. marina colonization can be identified and mapped on regional or local spatial scales. In this study, we present maps of areas suitable for colonization of Zostera marina for the Narragansett (Rhode Island) and Buzzards Bays (Massachusetts) estuarine systems during the 2013 and 2014 growing seasons.