Skip to main content
U.S. flag

An official website of the United States government

Scientists from the USGS, University of Connecticut, and U.S. Fish and Wildlife Service developed an automatic algorithm for DEtection and Characterization of cOastal tiDal wEtlands change (DECODE) using Landsat time series.

Coastal tidal wetlands are among the most productive and valuable ecosystems in the world—providing critical benefits and services to society such as flood control and storm surge mitigation, water quality maintenance, and carbon storage. 

Tidal Marshland in the Plum Island Estuary, Massachusetts
Tidal marsh of Plum Island Estuary, Cape Cod, Massachusetts.

However, frequent land-use change and sea-level rise jeopardize these important ecosystems and the services they provide by interfering with their hydrologic and ecologic functions, thus significantly reducing their climate resiliency. Knowing where, when, and what kind of changes have occurred is critical to both regional and national efforts to identify, plan for, and restore coastal tidal wetlands. Unfortunately, large-scale mapping of these ecosystems is extremely difficult because they are constantly changing.

To fill this gap, scientists from the USGS, University of Connecticut, and U.S. Fish and Wildlife Service developed a new automated algorithm, called DECODE, to track coastal tidal wetland changes using dense Landsat time series. It is designed to specifically capture key characteristics of coastal tidal wetlands, including condition change, while neutralizing the effects of routine tidal fluctuation. 

DECODE consists of three elements—spectral break detection, land cover classification, and change characterization—and can provide accurate land cover and land change maps for coastal tidal wetland areas fully automated at a high spatial resolution (30 meters) for large areas. 

The scientific team used DECODE to create annual cover and change maps in the northeastern United States from 1986 to 2020. The overall accuracy of land cover classification and change detection is about 95.8% and 99.8%, respectively. They found that approximately 3,283 square kilometers of coastal landscape, 12 % of the study area, changed at least once, and condition changes were the primary change type at 84.3%. Additionally, vegetated coastal tidal wetland decreased consistently—about 2.6 square kilometers per year—in the past 35 years, largely due to conversion to open water caused by sea-level rise. 

Looking to dive deeper? Read the recently published article in the journal Remote Sensing of Environment, titled “Detection and characterization of coastal tidal wetland change in the northeastern US using Landsat time series,” where this collaborative scientific team provides a comprehensive description of DECODE, details how it works, and discusses the results of using DECODE to track the status of coastal tidal wetlands in the northeastern United States from 1986 to 2020. Authors include Dr. Xiucheng Yang (University of Connecticut; UCONN), Dr. Zhe Zhu (UCONN), Dr. Kevin Kroeger (U.S. Geological Survey, Woods Hole Coastal and Marine Science Center; USGS WHCMSC), Dr. Zhiliang Zhu (USGS), and Scott Covington (U.S. Fish and Wildlife Service).