You are here:
Optical and seagrass characteristics of selected southern New England estuaries from satellite remote sensing
Citation:
Keith, D., P. Colarusso, AND S. Rego. Optical and seagrass characteristics of selected southern New England estuaries from satellite remote sensing. Long Island Sound Habitat Restoration and Stewardship Working Group, NA, Virtual, February 09, 2021.
Impact/Purpose:
Local, state and federal government natural resource managers are always in need of seagrass data for decision making but are limited by available funding and expertise. Most states have established programs that use aerial photography or hydroacoustic sensor (i.e., sidescan sonar) data to map seagrass distribution. Resource managers also rely on dive surveys to generate seagrass abundance data. These techniques are expensive cost-wise, temporally and spatially limited in scope, and the frequency of data collection varies greatly by state. The purpose of this project is to determine if high-resolution imagery routinely collected by satellite remote sensing platforms (e.g., Landsat 8) can be used to estimate not only seagrass distribution but also measures of health (abundance, biomass, and primary productivity) from images acquired at weekly to monthly timescales. This approach could provide a consistent, cost effective alternative for resource managers to gather seagrass abundance data and manage this valuable resource.
Description:
Advances in understanding the optics of shallow water environments, submerged vegetation canopies and seagrass physiology, combined with improved spatial resolution of remote sensing platforms, now enable submerged ecosystems to be monitored at a variety of time scales from earth-orbiting platforms such as Landsat 8. This demonstration project developed tools for extracting seagrass distribution, biomass, and forecasting primary production from Landsat 8 spectral data from in situ diver data collected during Summer 2018 and 2019 for five southern New England/Long Island Sound estuaries. This presentation is focused on Niantic River/Jordan Cove, CT, which were sampled in 2018. Within this estuary, diver-collected samples were analyzed for shoot density (shoots/m2), leaf area Index (LAI, m2seagrass/m2seafloor), and biomass (g/m2). Also, in situ hyperspectral data were collected to understand the underwater light field and the optical character of this estuary. Seagrass LAI and biomass were estimated using remote sensing algorithms constructed during this project. Daily seagrass primary production was estimated from satellite-derived LAI estimates using a series of empirical equations. Surface water samples were collected and analyzed for chlorophyll a (µg/L) and total suspended solids (mg/L) concentrations as well as light absorption by colored dissolved organic matter (m-1). This information was entered in a bio-optical model which predicted the attenuation of light (Kd) for the spectrum photosynthetically available radiation (PAR). Seagrass biometric parameters (LAI, biomass, and primary production) were mapped for Niantic River/Jordan Cove estuary based on spectral data from Landsat 8 images. The results from bio-optical modeling allowed us to understand the relationship between KdPAR (m-1), as well as LAI density, and the percentage of surface light reaching the estimated colonization depth for Zostera marina. These approaches will allow scientists and managers to estimate seagrass health and abundance without the expense of conducting aerial surveys.
