PUBLICATIONS

• Publications by topic
  • Groundwater
  • Surface Water
  • Water Quality
  • Water Use

• Publications by date
  • Recent publications

Annual Water Data Report

USGS Publications Warehouse

USGS IN YOUR STATE

USGS Water Science Centers are located in each state.

Development of a water-quality modeling framework for the Neuse River estuary

Conference Proceedings
By Jerad D. Bales and Jon Mangles

Abstract

Watershed management of nutrient inputs to coastal systems is a complex task. In many basins, nonpoint-source inputs exceed point- source loadings, but nonpoint sources are difficult to quantify. Management of nonpoint sources is even more challenging because of the effects of interannual weather variations and the cumulative effects of decisions made by numerous landowners and resource users. Compounding the challenge is the fact that nutrients released near the headwaters of a coastal basin likely do not have the same effect on coastal waters as nutrients released directly to the estuary. Despite the complexities of the natural system and the inadequacy of information, resource managers must make decisions about allowable nutrient inputs to coastal waters. Recent court decisions and pending lawsuits are accelerating the need to determine and establish total maximum daily loads (TMDLs) for receiving waters.

Decisions on acceptable nutrient inputs can be based on an experimental approach, on collection and interpretation of data, or on results from water-quality models to project the effects of proposed changes in nutrient inputs on water-quality conditions. The experimental approach --institute a change in loadings and measure the results--is iterative, and the effects of changes in loadings and hydrologic variations are difficult to distinguish. However, this approach can be implemented quickly, meeting public demands for action, and can be a useful first step in improving water quality if it is recognized that further changes in nutrient loadings may be warranted as new information becomes available.

Data collection and interpretation is fundamental to documenting changes in estuarine water quality in response to natural variations and management actions, understanding important processes, and developing more rigorous water-quality models. Data-collection efforts need to be carefully designed and periodically reviewed, and should include significant resources for data archival, interpretation, integration of results from all data-collection efforts, and reporting. Extrapolation of findings derived from one set of conditions (climatic, hydrologic, and water-quality management) to possible future scenarios, however, is difficult without a formal set of algorithms, or a model.

Water-quality models, whether statistical, empirical, or mechanistic, have been successfully used to set discharge limits and establish TMDLs in rivers. Decision-making applications of estuarine water- quality models, however, are less routine than riverine applications. Applications of riverine water-quality models typically are made for steady low-flow conditions, but estuarine water-quality models must account for the dynamic nature of meteorological, riverine, and coastal ocean conditions on estuarine processes. Moreover, estuarine water- quality models generally do not fully represent the complexities of long-term biogeochemical cycling, nor the effects of nutrient inputs on productivity and phytoplankton succession. Finally, the public often perceives the primary water-quality issue as the appearance of diseased or dead fish and shellfish, rather than elevated nutrient levels, which are more readily regulated and modeled. Reliable, usable, and linked hydraulic-water chemistry-fisheries models have yet to be developed.

A dynamic water-quality modeling framework is being developed by the U.S. Geological Survey, in cooperation with the North Carolina Division of Water Quality, for the Neuse River estuary. The primary application of the model is to evaluate the effects of a proposed 30-percent reduction in nitrogen loadings to the estuary on selected response variables--primary chlorophyll a and dissolved oxygen concentration. The model domain extends from Oriental upstream about 63 kilometers, and includes seven embayments, or tidal creeks, along the estuary. The mainstem of the estuary is divided into 35 segments which range in length from 500 to 7,100 meters, and in surface width from 100 to 6,700 meters. Each segment is subdivided into as many as 10 layers, one-meter thick, depending on water depth. Hydrodynamic and water- quality conditions are assumed to vary longitudinally from segment to segment, and vertically from layer to layer. Data collected by the North Carolina Division of Water Quality, the University of North Carolina Institute of Marine Sciences, and the U.S. Geological Survey during March-October 1991 are being used to construct and test the model. More recent complete data sets are not available.

Application of this model to the Neuse River estuary exemplifies some of the difficulties associated with using results from estuarine water-quality models to make management decisions. All physical, chemical, and biological processes in the estuary need not be modeled. For example, phytoplankton respond to a variety of phenomena (for example, physical transport, stratification, sediment resuspension and light attenuation, and benthic and pelagic grazing), and phytoplankton growth results in a number of changes to the estuary (for example, depletion of inorganic nutrients, dissolved oxygen supersaturation, increased benthic oxygen demand, and changes in the form and toxicity of selected trace metals). But, modeling of estuarine response to changes in nutrient loadings requires that the key processes be identified, understood, and reasonably modeled, and that processes which are peripheral to the decision-making issues be minimized. Water-quality simulations also are typically performed by using historic meteorologic and hydrologic data, to answer such questions as, "What would dissolved-oxygen concentrations have been in the Neuse estuary in 1994 if nutrient loadings had been 30 percent lower than observed?" However, historic conditions will not be exactly repeated in the future, so it cannot be known with certainty whether future estuarine response will be the same as in the past. Further uncertainty is introduced into model predictions because the Neuse River estuary water-quality model includes more than 50 parameters, many of which cannot be directly measured, and by variability associated with data and model algorithms. However, if these, and other, limitations associated with the Neuse estuary water-quality model are recognized, the modeling framework can be an important tool in the wise management of coastal waters.