Figure 1: NDVI image of part of Dare County, showing distinct difference of wetlands from surrounding units. Gold outline indicates location of NCDCM wetlands units, corresponding with differences in vegetation.
Student Investigators:
Peter Johnson, New Mexico Institute of Mining and Technology (project lead)
Brendan McAndrew, Christopher Newport University
Bryan Glover, Christopher Newport University
Project Overview
Hyde, Tyrrell, and Dare counties in eastern North Carolina occupy the coastlines of the Albemarle and Pamlico Sound estuaries, especially the continental side ÛÒ the ÛÏInner Banks,Û counterpart to the more tourism-based Outer Banks. Coastal erosion, rising sea levels, and increasing storm intensities pose a well-documented threat to economic interests on the Outer Banks. Of less prominence, but of no less importance, is the potential of these same influences to impact the estuarine ecosystems and communities of the Inner Banks. Habitat loss resulting from inundation of coastal marshes, combined with rising sea levels and increased storm intensity may lead to increased damage to properties from flooding, as well as impacts to fisheries and wildlife resources. The Pamlico Sound, located in North Carolina, is the largest estuarine lagoon along the East coast of the United States. The North Carolina Division of Coastal Management (NCDCM) holds responsibility for managing coastal resources and mitigating coastal hazards. To assist them in their work, a team of students from the DEVELOP National Program, a NASA Science Mission Directorate Applied Sciences-sponsored workforce development internship program, partnered with the NCDCM to assist with the assessment of coastline changes and habitat loss in coastal wetlands utilizing satellite remote sensing.
Figure 2: Normalized Digitized Vegetation Index (NDVI) for portion of Tyrrell County. Green color is least vegetated; grey is greatest. Red colors indicate intermediate levels of chlorophyll. Gold-colored outlines show NCDCM locations of wetlands.
By studying the ecosystem functions and the trends in shoreline movements, the DEVELOP team helped create an analysis of shoreline change and how coastal erosion has impacted the estuary systems. This project’s goal is to detect changes along the shoreline and estuaries, and measure how they are changing in timing, magnitude, and the types of events that can cause changes. A number of factors could play into this; the estuaries face daily erosion which depletes the land surrounding many homes and structures, while tropical cyclones and nor’easters can cause large-scale erosion over short periods of time. Erosion of the Outer Banks has been well-documented, but within the Inner Banks the magnitude of the problem is not well-established. If the shorelines of the two large coastal estuaries undergo rapid, major change, the danger to structures, historic areas, and wildlife habitats is undeniable.
With an accurate method of evaluating coastlines based on satellite images, the NCDCM could examine past satellite images and evaluate the magnitude and timing of coastline changes. Although the NCDCM conducts pentennial high-resolution evaluations of the coastline using aerial photography, this method cannot evaluate the differences caused by singular events such as tropical cyclones due to the expense of aerial overflights. Thus, a medium-resolution but higher-frequency dataset could assist the NCDCM in making policy decisions concerning coastal management by giving them a viable method of evaluating the events that can cause major coastline change Knowledge of which o these events cause major changes and where those changes are located can help identify property and habitats that are at risk; this, in turn, can help better evaluate how to manage the coastline to ensure a minimum of property loss, ecological harm, and economic damage from events such as hurricanes and nor’easters (an American contraction of Northeastern hurricanes). To achieve all of this, the coastline must first be located accurately.
Figure 3: Landsat band 5 (b5) image of southern Tyrell County. Note color distinctions along wetland boundaries.
This project utilized six Landsat satellite images from the USGS GloVis website (Figures 1 and 2). The six images were from different time periods within the same year to limit inter-annual variability. Each image was classified into two categories, land or water, where land pixels were assigned a value of 0 and water pixels were assigned a value of 1. By stacking the six images and adding the sum total for each pixel, three results were possible. If the pixel sum total was 0, then the pixel was always land and remained unchanged. If the pixel was always water the sum was 6. Any sum between these two extremes was variable ÛÒ that is, it contained water sometimes and land at other times. Tidal factors, storm events, mass wasting, and other phenomena could all contribute to a change in pixels from land to water thereby indicating the presence of a dynamic coastline. This variability indicated that the pixel was part of the shoreline, being near enough to water to be sometimes submerged, while also being exposed. Images of the shoreline were produced and overlaid on top of a high-resolution shoreline shapefile provided by NCDCM. Comparison of the two images indicated acceptable accuracy of this shoreline method, indicating that it could be useful to the NCDCM for general evaluation of shoreline location and changes.
The DEVELOP team assisted the NCDCM by exploring methods by which their estuarine wetlands could be examined using satellite imagery. By comparing a shapefile of wetlands provided by NCDCM to a variety of bands and band manipulations of the Landsat imagery, the DEVELOP team determined that these particular wetlands tended to show decreased chlorophyll (obtained using a standard Normalized Digitized Vegetation Index) relative to surrounding land units. Additionally, a distinction from bare soil (e.g. tilled fields) could be accomplished by further examination of the mid-infrared band (b5) of Landsat (Figure 3). The location of wetlands could be determined by combining NDVI and b5, allowing the NCDCM to examine older images and ascertain how wetlands have changed since 1983.
Figure 4: Alligator in estuarine waters in eastern North Carolina. Photo courtesy of the North Carolina Division of Coastal Management
Coastal wetlands contribute a number of useful functions. As a unique habitat, they can be home to endangered species and serve as fish habitats, avian breeding grounds, and home to other flora and fauna (Figure 4). They also provide the livelihood of thousands in the fishing and seafood industry. Furthermore, they provide a buffer against storms, absorbing some of the impact of storm surges and high winds that might otherwise damage homes and communities near the shore. Consequently, the loss of these habitats must be monitored carefully and enhanced management and decision making support is needed for coastal North Carolina.
Project Mentors:
Dr. Kenton Ross ÛÒ Science Systems & Applications, Inc., NOAA National Environmental Satellite, Data, and Information Service (NESDIS)
Dr. Jeffrey Warren – North Carolina Division of Coastal Management
Scott Geiss – North Carolina Division of Coastal Management
For more information about the DEVELOP National Program and other student projects, please visit http://develop.larc.nasa.gov.