Advanced Imagery Brings New Angle to Wetlands Research

Article Author: 
- by Amelia Jaycen, Publications Intern, Office of Research and Economic Development, with feature editing by Julie West

Water flowing through its natural environment follows loosely predictable paths according to the landscape structure around it, forming a hydrologic cycle that affects the entire planet.  In wetland ecosystems, plants play an important role in purifying water and maintaining a healthy watershed, and human alterations to the landscape can diminish the natural system’s ability to filter water to its full potential. 

The University of North Texas has a legacy of research in water salient issues, with information and benefits flowing richly throughout north Texas into systems and organizations whose business is the study, regulation, and mitigation of water issues, including municipalities managing waterways within their jurisdiction.

Dr. Sam Atkinson, Director, UNT Institute of Applied Science.
- UNT Faculty Photograph

Regents Professor of Biology Dr. Sam Atkinson and his students are using advanced imaging technology to augment the study of wetlands and understand how plants contribute to their functioning.  A combination of Global Positioning Satellites (GPS), governmental and private earth imaging satellites, and a variety of sophisticated Geographic Information System (GIS) resources from UNT’s Center for Remote Sensing offer great potential for gathering data, and tracking and evaluating water resources, croplands, rangelands, urbanization, and a spectrum of conservation projects.

“Images from one season to the next can be precisely overlaid using geo-referencing points, making it easy to study correlations between vegetation, location, weather, and water quality,” Atkinson said.  “We can virtually watch plant growth on a screen, using a slider bar to fade back and forth between images.”

Atkinson is director of the UNT Institute of Applied Science, a thriving interdisciplinary research unit composed of faculty, students, and affiliate researchers from disciplines including environmental science, geography, biology, chemistry, computer science and engineering, and philosophy and religion studies.  The IAS mission is to “foster, facilitate, and conduct science-based interdisciplinary environmental research that seeks to understand how human actions impact the environment, and to use that knowledge to suggest scientific, engineering, policy and/or educational solutions to environmental problems.” To this end, the Center for Remote Sensing (CRS) is a valuable resource not only for IAS and other UNT research projects but for organizations working with UNT. 

A ring of hibiscus flourishing at the Grand Prairie wetland site, despite the area being nearly completely dry on the surface.  Researchers hope advanced imagery can eventually allow automatic species identification by leaf pattern and color.  

Protection of U.S. waterways, including wetlands, is the subject of federal policy under the Clean Water Act, and municipalities and industries are required to comply with water protection and watershed mitigation measures.  Some simply meet the requirements, while others go beyond the basics with an ecosystems approach that considers the needs of both human-built and natural systems.

Environmental sustainability is a focus for the City of Denton and the City of Grand Prairie, and both organizations have turned to UNT through the years to help create lasting benefits for Texas water resources and the communities that thrive on them. UNT is partnering with these municipalities in two research projects that will advance understanding of wetland plants and how they help filter water throughout a watershed system.    

“In the Grand Prairie project we are trying to understand how well we’re doing at off-setting adverse impacts on the wetland and surrounding watershed, and in Denton we are looking at how water quality changes through a wetland,” Atkinson said.  “The link between them is the use of imagery to quantify an important aspect: the wetland vegetation in the landscape.”

Remotely sensed wetlands

Plant studies depend extensively on visual data gathered up-close by researchers in real-time as they work; observation, photographs, maps, and measurements are central to work in environmental sciences, historical and contemporary alike.

Researchers are often expected to present reports that are laden with photos and visuals yet are also comprehensive and data-rich with geographic and timeline components. GIS tools fill this need and are used to create advanced graphics and visual data systems that can store, manipulate, and display several levels of information.

Marty O'Neill ('09), computer scientist, UNT Institue of Applied Science staff, member of Center for Computational Epidemiology and Response Analysis (CeCERA).  - Image by Amelia Jaycen. 

Atkinson and his team use custom designed visual data systems that incorporate GPS, mapping, geo-referencing, 3-D modeling software, and data collected from GIS databases like Google Earth to analyze ecosystems.  Images collected over time can be viewed as a multi-dimensional set or individually by year to determine plant species, their number, location, growth rates, and other variables.  The cumulative result is a valuable complex of data. 

“The technology allows us to build and examine three-dimensional models of the study area that can help us understand real-world systems,” computer science doctoral candidate Marty O’Neill said.  O’Neill (’09) works in the Institute of Applied Science and the Center for Computational Epidemiology and Response Analysis (CeCERA), and he assists Atkinson with programming the remotely sensed imagery system and processing wetland images into composites.

“As a computer scientist, I’ve never put on boots to go study a landfill or wastewater treatment plant before; this is my first real experience with environmental field work,”  O’Neill said. “That is the great thing about the Institute and UNT as a whole—the potential for interdisciplinary collaboration is around every corner.” 

A healthy balance

 Picture of the Grand Prairie wetland site with ring of hibiscus (left) and the Trinity River oxbow (right) obtained from the Google Earth satellite imagery database.

Situated near an oxbow, or U-shaped bend, in the west fork of the Trinity River, Grand Prairie’s landfill is a prime spot for the study of wetlands.  In 2001, when the city sought the necessary permits to install the landfill and a dam to control water flow into the oxbow, city, state, and federal water regulation agencies struck a balance that allowed construction of the facility in return for deed restriction of 77 acres of the surrounding area.  Grand Prairie leaders embarked on a green initiative to maintain the health of the local watershed and protect the oxbow.

The city first looked to UNT’s Institute of Applied Science for help in 2009. They contracted a research study, “Biotic and Hydrologic Characterization of the Wetlands,” in order to gain background information necessary to gauge the success of proposed mitigation efforts.  Principal investigators Drs. Kevin Stevens, James Kennedy, and Paul Hudak were enlisted to conduct a vegetation and seed bank study, macroinvertebrate sampling and identification, and a hydrology report, respectively. 

“We’re working toward having a fully functioning wetland, and we can define what we mean by fully-functioning based on how it was to begin with, using assessments taken in the first iteration of our project with UNT,” environmental science alumna (’04) and City of Grand Prairie Solid Waste and Recycling Director Dr. Patricia Redfearn said. 

UNT research scientist Joe Snow shows a piece of alligator weed, an invasive species that can reproduce from every part that breaks off a plant.  - Image by Amelia Jaycen. 

Since the original study, four more contracts between the city of Grand Prairie and UNT researchers have continued the wetland work, with the latest two completed by Atkinson and UNT research scientist Joe Snow.  The city has constructed a six-acre wetland and plans to create two smaller, adjacent wetland sites. 

In the current project, “Production and management of a Diverse Native Plant Community for the Wetlands on the City of Grand Prairie’s Landfill,” the goal is to achieve enough vegetation and diversity of species to facilitate a healthy wetland ecosystem. 

The process of populating the area with plants that can survive weather conditions in north Texas involves careful species selection and sourcing of native plants. Snow has planted at least 10,000  plants at the Grand Prairie site, sourcing seeds from the oxbow, other sites in the Trinity River watershed, the Clear Creek Natural Heritage Center, and the North Texas Greenbelt.

UNT research scientist Joe Snow installs plants at the Grand Prairie wetland site, assisted by Amanda Lindbergh ('10 TAMS).
- Image by Amelia Jaycen.

“Historically, hydrology reports show that the Grand Prairie site has a fluctuating water level system, but the really unusual weather we’ve had the last two years has been challenging,” Snow said.  However, even in hot, dry weather conditions the area still maintains enough soil moisture: close proximity to the water table provides the underlying hydrology necessary to support wetland function and allow the plants to thrive.

“I really feel like had we not involved researchers who knew exactly what they were doing, we would have wound up with an alligator weed monoculture on both sides of the wetland, with absolutely no proper wetland function and zero biodiversity,” Redfearn said.

The invasive alligator weed poses a threat to the area’s native seed bank if left unmanaged.  Researchers have used a combination of chemical and biological controls to restore the indigenous plant community, including targeted sprays and aggressive re-planting of native species.

This spring, researchers began the fourth planting season at the Grand Prairie wetland, and it has been valuable to track and study the success of their plantings through several seasons. Planting progress is tracked in the field using Global Positioning System (GPS) equipment. 

UNT offers Certification in Global Information Systems (GIS) including GPS training through the Department of Geography. GIS offers essential planning and management tools for city and county governments, utility companies, and various private sector firms that use geographic and spatial information, and many students in the natural sciences as well as public administration, engineering, and other fields use the skills extensively in their work.  Amanda Lindbergh is a Texas Academy of Math and Science (TAMS) alumna (’10) who begins graduate studies in the geography department this fall. 

After completing her Bachelor of Science degree in biology at Texas State University, Lindbergh applied for a position as research assistant to help with the Grand Prairie wetland mitigation project.  In addition to seeding, tending, and transplanting scores of plants, she and other assistants are trained to use advanced tools to compile data on the planting process.  

Research assistant Amanda Lindbergh ('10 TAMS) uses GPS equipment to mark the location, species, and number of plants installed.  - Image by Amelia Jaycen.

Walking around the perimeter of each cluster of plantings, Lindbergh enters data into a GPS unit that saves the species and number of plants with the exact latitude and longitude locations. Back at the lab, the data is super-imposed onto high-resolution imagery of the wetland using global positioning points to make a virtual map showing where plants have been installed on the site. 

“We are using GIS software and tools that can provide measurements equivalent to actually taking a tape measure out there,” Snow said.  “An overhead view of our planting compared with our GPS maps can yield a good correlation of what you actually see versus what we have recorded – something that would be very hard to compare outside of careful tracking methodology.”

With advanced imaging tools and the help of O’Neill’s expertise in programming and data processing, Atkinson’s remote sensing system adds a new vantage to the study of wetland plant populations that significantly augments research obtained from the ground.

Data gathered on the wetland’s plant communities allows Snow to compare plant growth and survival.  Remotely sensed imagery is also used to compare wet and dry times at the wetland, and images are referenced to document the fluctuation in seasons as it correlates to the presence of vegetation.

 “Using these technologies, we can quantify various aspects of the landscape; in this case it is how much vegetation is surviving at both sites and which species,” Atkinson said. “Instead of simply looking at one variable at a time, we can begin to understand the emergent properties that exist when multiple environmental systems are working together.” 

Denton wetlands – a local legacy

In the City of Denton, water collected from 82 different sub-basins affecting Denton’s three major watersheds is analyzed either in the city’s laboratory or in Atkinson’s UNT laboratory, allowing interns to gain valuable experience where industry meets the classroom.  “We have a wonderful relationship with the City of Denton that goes back a long time,” Dr. Sam Atkinson said. 

City of Denton wastewater treatment facility, labeled with collaborative UNT research projects. - Image provided by David Hunter.

In 1991 the City of Denton constructed a wetland between the Pecan Creek Water Reclamation Plant and the stream bottoms feeding Pecan Creek that lie effluent of the plant, and Atkinson and his students began studying it right away.  In 2001, the city’s Watershed Protection Program was launched as part of a plan to reduce pollutants in Denton surface waters, incorporating infrastructure established by an Environmental Protection Agency (EPA) EMPACT grant awarded to the City of Denton and the University of North Texas in 1998. 

The Watershed Protection Program employs interns and student volunteers from UNT’s departments of geography, chemistry, biology, environmental science, engineering, electrical engineering, and computer science, many of whom complete theses and dissertations at the wetlands and other research sites nearby.

As an intern with the watershed program, UNT environmental science graduate student Amesha Morris has become familiar not only with the course local waters take through systems in nature and municipal systems, but she has gained personal experience with the rich history of collaborative research surrounding the sites where she works. 

Amesha Morris, UNT graduate student and intern for the City of Denton Watershed Protection Program, pictured at the Denton wetland site where she completed her thesis research this spring. - Image by Amelia Jaycen.

“Lake Lewisville receives effluent from 22 different wastewater treatment plants over a span of about 15 miles.  My concern is whether pollutants in the water degrade enough through natural processes to eliminate potential problems,” Morris said. 

Manager of Watershed Protection and Industrial Pretreatment for the City of Denton, David Hunter (’99, ’02 MPA) said the engineering paradigm is changing, and engineers are starting to think more seriously about where water goes downstream, how to fit nature into the equation, and what affects water quality from an ecosystems standpoint. “Pharmaceuticals and pesticides are a concern for everyone with an interest in clean water,” Hunter said.

Wastewater treatment is a process that uses a series of filtering mechanisms—including grit removal, clarifying, biological cleansing by specialized microbes, sand filters, charcoal, and UV light treatment—in order to purify large volumes of water as quickly as possible before it is released.  Water flowing out of a treatment facility on its way back to nature is additionally filtered by the microorganisms, plant community, and soil in the wetlands system.

Satellite imagery from Google Earth (top) compared to remotely sensed imagery (bottom) gathered by Atkinson's team.  

“Understanding how extensively a wetland or other natural system can break substances down and what the characteristics are of the physical, chemical, and biological pathways through which water moves in the natural purification process is information we can apply to our human-made system,” Hunter said.

For her thesis research in spring 2013, Morris compared estrogen effects on fish in water at the sand filter (one of the final phases of processing in Denton’s treatment facility) with fish in water additionally processed by the constructed wetland.  One obvious difference in the water at the end of the wetland, Morris said, is that vegetation has absorbed and metabolized some substances out of it. 

“Knowing the effects of these environmental estrogens on model species like those Amesha studies gives us an understanding of what happens when other organisms are exposed to it, and we are also learning what plants can do to help the filtering process.” Atkinson said.

Counting plants and gauging the density and variety of species in a wetland is not easy fieldwork, and Morris and fellow researchers must carefully wade through the water so as not to damage or disturb the plants under study. Remotely sensed imagery, with its wide-angle vantage from above, can aid this process immensely and yield more accurate data.

From the field to the screen, and beyond

While satellite imagery from Google Earth and other sources is useful, monitoring plants in a small area of a few acres still presents a problem.  Both satellite imagery and aerial photography are expensive and hard to coordinate, and researchers often experience long delays before obtaining images of the areas they are studying. Even with images in hand, identification of plants based on subtle differences such as leaf shape and color is difficult without more advanced tools.

A computer generated 3-D model of the Denton wetland site, with tiles
marking the vantage points of remotely sensed images taken to create
a mosaicked image.

The challenges inherent to the work at UNT, the City of Denton, and the City of Grand Prairie become opportunities for advanced study through collaboration using Atkinson’s custom-designed technology and aerial mosaic techniques. Hundreds of pictures of wetlands are taken in a pre-determined pattern from various vantage points, and spatial coordinates are automatically recorded along with each photograph. Multi-view 3D reconstruction software is used to process these pictures into formats useful for environmental science research. 

“This process yields images in which each pixel represents less than a square centimeter on the ground – an extremely high resolution. Comparing images of the same location taken at different times enables researchers to observe changes in vegetation using imagery viewed in the lab versus observations in the field,” computer scientist O’Neill said. “This brings another aspect to the research that would not be possible otherwise.”

Atkinson said the potential for computer-based assistance in plant and water research are endless. Researchers hope to achieve a level of precision that could allow studies of water color, non-terrestrial plants, and computer recognition of leaf patterns for automatic species identification; experimental directions include infrared imaging that could reveal insight into plant health and type. 

“These kinds of studies are not far from possible,” Atkinson said.  “New imaging technologies can open all kinds of doors for advanced water research at UNT.” 

A point cloud of the Denton wetland shows points that overlap in all remotely sensed images.  Points are used to combine pieces of each photograph into one composite, high-resolution image for study.  
Modeling images provided by Marty O'Neill, satellite imagery provided by O'Neill, Joe Snow, and Google Earth.
Features Editor: Julie K. West.