It Came from the Lake: Hydrilla Mapping in Southeast US Reservoirs

Floating (red) and submerged (green) Hydrilla distribution in Lake Seminole, September 2014. Image Credit: Southeast Ecological Forecasting II Team

Floating (red) and submerged (green) Hydrilla distribution in Lake Seminole, September 2014. Image Credit: Southeast Ecological Forecasting II Team

This is a part of the 2015 Fall VPS. For more VPS articles, click here

Category: Mapping Water Quality
Project Team
: Southeast Ecological Forecasting II
Team Location: University of Georgia – Athens, Georgia

Authors:
Benjamin Page
Brandon Hays
Pradeep Kumar Ragu Chanthar
Linli Zhu 

Mentors/Advisors:
Dr. Deepak Mishra (Department of Geography, University of Georgia)
Dr. Susan Wilde (Warnell School of Forestry and Natural Resources, University of Georgia)

Abstract:
Hydrilla verticillata is an invasive aquatic plant which has rapidly spread through many inland water bodies across the Southeastern United States (SEUS) by outcompeting native aquatic plants and displacing fish populations. Consumption of water for drinking and power generation as well as recreational use of lakes has been threatened by the spread of Hydrilla. In recent years, Hydrilla has served as a vehicle for the spread of a toxic cyanobacteria (Aetokthonos hydrillicola) responsible for the neurodegenerative disease Avian Vacuolar Myelinopathy, which causes massive fish kills and bald eagle deaths throughout Georgia. Using Landsat 8 Operational Land Imager (OLI) data, a rapid assessment tool was developed to map the extent of Hydrilla and predict future spread throughout the SEUS by quantifying seasonal biomass through a time-series analysis. A normalized difference vegetation index was performed and overlaid on the green band to differentiate between floating and submerged vegetation. Depth measurements were taken at Lake Herrick using Secchi disks to calculate light attenuation and identify the lower boundary of photosynthetically active regions. These data informed an unsupervised classification model, which was further trained using aerial imagery acquired by an unmanned aerial system. This model was validated with in situ biomass measurements and local knowledge. Next, data collected from 2013 to 2015 were used to create a time series of true color composite images, which were analyzed to map regions of expanding Hydrilla presence. The time series, in conjunction with a quantitative model of Hydrilla growth, were used to predict future Hydrilla hot spots.

Previous Story / Next Story

Topic: