By Jacob A. Napieralski, Earth and Atmospheric Science Department, Purdue University, Indiana

Varying 3D views of the Buffalo River plume using vertical and horizontal slices (approximately 6 meters below water surface). The results show a decrease in turbidity levels after 24 hours and that the sediment either settled near the first breakwall (~2,000 feet from the mouth) or moved north, toward the head of the canal. Vertical exaggeration is x50.

Determining the fate of suspended sediments in the Buffalo River and Black Rock Canal, located in western New York, was an integral component of an environmental study conducted by the Great Lakes Center at Buffalo State College. The purpose of this research was to monitor and model transient events in large water bodies, such as the Buffalo and Niagara Rivers, during rainfall and base flow events.

The lower reaches of the Buffalo River have been designated a Great Lakes Area of Concern in part because of the resuspension of contaminated sediments and the contribution of combined sewer overflows (CSO) to the river. Sediments are an ideal sink for a variety of contaminants and can be a direct source of contamination to a water body (if resuspended), so it is critical to determine the path these sediments take during phases of transportation and deposition. Sediment plumes cause habitat degradation by increasing water temperature while decreasing dissolved oxygen levels, and the survival of aquatic life may be dependent on reducing the influx of these sediments. As a result, the impact of storm-induced sediment resuspension and CSOs in the Buffalo River is recognized as a negative impact on urban water quality and was the focus of this study.

To address this question, two research vessels equipped with Seabird Conductivity/Temperature/Depth (CTD) oceanographic profilers were used to collect water quality parameters, including temperature, dissolved oxygen, and turbidity (suspended sediment). These parameters were collected at 85 sampling sites during two rainfall events and two base flow events. At each sampling site, data loggers were deployed from the research vessels and data was collected every 0.5 meter as they were lowered to the river bottom and subsequently raised to the vessel. The x,y locations of the sampling stations were established with Magellan GPS units, and the z locations were established by pressure transducers on the profilers. As a result, data was collected every 0.5 meter of depth at the 85 sampling sites during nine sampling runs.

ArcGIS (ArcView) was required to compile, analyze, and view an extensive amount of water quality data. As a first step, however, 2,500 depth points were collected from U.S. National Oceanic and Atmospheric Administration navigation maps and imported into ArcView to construct a bathymetric model of the northeastern portion of Lake Erie, the Niagara and Buffalo Rivers, and the Black Rock Canal. This was necessary for the modeling process, as it provided boundary constraints (e.g., shorelines, break walls) to the sampling area. The points were kriged, which is a statistical process by which known data values are used to estimate neighboring areas where data values were not collected or are unknown, using ArcGIS Spatial Analyst and the end result was a bathymetric model of the study area that could be viewed three-dimensionally.

The water quality data (x,y,z data with multiple values) was kriged to create thematic maps of each water quality parameter at varying depths. The kriged data was clipped to the bathymetric model, and the resulting models were used to assess the quality of data (finding and deleting sampling errors) and to view a preliminary outcome of kriging runs before ever running the model.

Environmental Visualization System, one of several modeling software packages designed by Esri Business Partner C-Tech Development Corporation (Huntington Beach, California), was chosen for the modeling because of the sound geostatistical analysis and attractive visualization tools, including animations and flyovers. Kriging the Buffalo River data in equal horizontal and vertical directions would have yielded erroneous results because the modeling had to consider any direction and magnitude of fluid flow. EVS-Pro software's anisotropy ratio number provided a horizontal weighting factor that was altered according to the flow regime of the river. As a result, the data was three-dimensionally kriged while considering general fluid flow characteristics.

"The software programs worked interchangeably with each other," says Gordon S. Fraser, director, Great Lakes Center, Buffalo State College, New York, "allowing us to simultaneously view thematic maps while still adjusting modeling parameters. As a result, we were able to transform an enormous amount of data into state-of-the-art visualizations that could be rotated and translated in real time, producing a variety of viewing perspectives."

Several of the modules, such as those allowing slices, were used to transform model output to three-dimensional shapefiles for easy input into ArcView. For this study, vertical and horizontal slices were used to examine spatial variations within the sediment plume during a single sampling event or changes in plume characteristics between sampling events. "There were several approaches we could use to view the data," says Fraser. "As an example, cross-sectional profiles of the Buffalo River sediment plume were viewed during various stages of the rainstorm to determine the spatial extent of the plume as well as any temporal changes in plume characteristics."

Results show that the levels of turbidity in the Buffalo River were relatively high during the beginning of storm events and gradually decreased as the storm dissipated. In addition, the majority of suspended sediment was restricted to the lower half of the water column, and much of it settled outside the mouth of the river near the head of the Niagara River. However, when turbidity levels were relatively high in the Buffalo River (not necessarily restricted to rain events), the sediment plume could be seen entering the stagnant waters of the Black Rock Canal.

Prior studies in this area have yielded spatially or temporally limited views of the structure of the Buffalo River, but combining ArcGIS and EVS-Pro provided a new viewing perspective of the water mass during varying weather conditions. Future work can use this approach to assess the individual impact of CSOs or alter the sampling protocol to increase the spatial or temporal resolution of the sediment plume model.

For more information, contact Jacob Napieralski, graduate research fellow, Purdue University (tel.: 765-494-0276), or Gordon Fraser, director of the Great Lakes Center, Buffalo State College (e-mail: frasergs@buffalostate.edu), or visit C-Tech at www.ctech.com.