Great Lakes Water Resources II

  1. Back to Archive
NDVI, MNDWI, and DSWE calculated using Landsat 8 OLI imagery (May-September 2019), blended with red, green, and blue color bands, respectively. The northern shore of Lake Superior is displayed, including Minnesota and Ontario. Areas of blue correspond to standing water, while dark purple areas correspond to dry uplands. Light purple areas represent likely wetland extent, where decision-makers can target conservation and monitoring efforts. Keywords: Landsat 8 OLI, Lake Superior, DSWE, NDVI, MNDWI, Minnesota, Ontario, wetlands, wetland extent

A Google Earth Engine Tool to Automate Wetland Extent Mapping Using Optical and Radar Satellite Sensors in the Great Lakes Basin for Wetland Management and Monitoring

The Great Lakes Basin is one of the world’s largest freshwater ecosystems. The Basin harbors over 200,000 acres of wetlands that provide a variety of environmental, ecological, and recreational functions to over 30 million people in the region. Some of these functions include improving water quality, mitigating flood impacts, providing wildlife habitat, and housing recreational activities. However, due to anthropogenic activities, habitat conversion and degradation threaten to disrupt or destroy remaining wetland ecosystems. Maps of wetland distribution based on ground surveys are costly and labor-intensive, prohibiting timely evaluations of wetland loss and gain. The Great Lakes Water Resources II team at the NASA Jet Propulsion Laboratory developed the Wetlands Extent Tool 2.0 (WET 2.0) in Google Earth Engine to automate mapping of wetland distribution in the Great Lakes Basin. The team partnered with the US Fish and Wildlife Service, Environmental Protection Agency, Minnesota Department of Natural Resources, the National Oceanic and Atmospheric Administration, and Ducks Unlimited. WET 2.0 incorporates Landsat 8 Operational Land Imager (OLI), Sentinel-1 C-band Synthetic Aperture Radar (C-SAR), and Sentinel-2 Multispectral Instrument (MSI) data. Utilizing a Random Forest classifier, WET 2.0 is capable of automatically mapping wetland extent in the entire Great Lakes Basin, achieving a mean overall accuracy of 80.12% when tested in Michigan. Findings and maps produced by WET 2.0 will enable our partners to identify areas of ecosystem degradation and wetland destruction in order to enact environmental practices and policy initiatives to maintain environmental and economic health in the area.

Project Video:

Wetlands Extent Tool 2.0 (WET2.0)

Location
California - JPL
Term
Spring 2020
Partner(s)
US Fish and Wildlife Service, National Wetlands Inventory
Minnesota Department of Natural Resources
US Environmental Protection Agency, Office of Research and Development
Ducks Unlimited
University of Minnesota
National Oceanic and Atmospheric Administration, Office for Coastal Management
Michigan Technological University
Natural Resources Canada, Canada Centre for Mapping and Earth Observation, Canada Centre for Remote Sensing
NASA Earth Observations
Landsat 8 OLI
Sentinel-1 C-SAR
Sentinel-2 MSI
SRTM
Team
Vanessa Valenti (Project Lead)
Erica Carcelen
Katie Lange
Nick Russo
Advisor(s)
Bruce Chapman (NASA Jet Propulsion Laboratory, California Institute of Technology)
Benjamin Holt (NASA Jet Propulsion Laboratory, California Institute of Technology)