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.

Dynamic water-quality modeling of Catawba River reservoirs

Project Abstract
By Jerad D. Bales

Abstract

The 7,860 square kilometer Catawba River basin in North Carolina extends from the mountains of the Blue Ridge physiographic province to Lake Wylie, which lies along the North Carolina-South Carolina state line in the low rolling terrain of the Piedmont Province. Essentially all of the Catawba River is regulated, and the only extended free-flowing reach of other river in North Carolina is between Bridgewater Dam at Lake James and the headwaters of Lake Rhodhiss. Seven mainstem reservoirs are located on the Catawba River in North Carolina, and six of the seven reservoirs were constructed before 1930.

The Catawba River basin is the most densely populated river basin in the State, and waters from the basin are in high demand for a variety of uses. The ability to meet demand depends in large measure on maintaining good water quality. Despite the importance of the water resources of the Catawba River, little is know about hydraulic circulation, as well as constituent transport and transformation in these reservoirs. Consequently, the U.S. Geological Survey, in cooperation with the Western Piedmont Council of Governments and the Charlotte-Mecklenburg Utility District, has been conducting investigations of water quality in selected reservoirs of the Catawba River. To date, these reservoirs include Rhodhiss Lake, Lake Hickory, and Mountain Island Lake.

An objective of these investigations was to develop a dynamic water-quality model for each of the reservoirs, in order to (1) simulate circulation, mixing, and residence times for a range of hydrologic and thermal conditions and to (2) predict reservoir water-quality response to possible changes in nutrient inputs to the reservoir. Similar approaches were used for all three reservoirs. Information for Lake Hickory is summarized herein.

A two-dimensional, laterally averaged hydrodynamic and water-quality model was constructed for Lake Hickory. The model was calibrated using data collected during April 1993-March 1994. The model has the capability to simulate water level, currents, heat transport and temperature distribution, conservative material transport, and the transport and transformation of 11 chemical constituents. The model was applied to Lake Hickory to characterize travel and residence times during stratified and unstratified conditions, and to evaluate the effects of changes in ortho-phosphorus loadings on dissolved-oxygen and algal concentrations in the reservoir.

The model extends for a distance of 22 kilometers along the longitudinal axis of the reservoir, and encompasses five embayments. The mainstem of the reservoir is divided longitudinally into 43 computational segments, and each segment is further subdivided into one-meter thick layers. Measurements of inflows to and outflows from the reservoir, inflow water temperature, meteorological conditions, and inflow constituent concentrations are required for model operation.

Measured and simulated water levels were in good agreement, with a root mean square difference of 0.046 m between measured and simulated values. Simulated water temperatures reproduced seasonal variations in near-surface and near-bottom water temperature in Lake Hickory from the headwaters to the forebay. Simulated values were generally within 1 degree of corresponding measured values. The vertical temperature distribution in Lake Hickory varied from near isothermal conditions, to strongly stratified, and back to very small top-to-bottom differences in water temperature. Simulated vertical distributions agreed quite closely with measured distributions, even for complex vertical distributions of temperature.

Simulated dissolved oxygen concentrations for Lake Hickory exhibited the same general patterns and magnitudes as measured values, and measured and simulated values seldom differed by more than 2 milligrams per liter. The onset of low-dissolved oxygen conditions, and the recovery to higher levels was very well-predicted throughout the reservoir. It appears, however, that the hypolimnetic volume of the reservoir may have been underestimated, leading to an under-prediction of dissolved-oxygen concentrations in September, when the reservoir became vertically mixed. Simulated algal concentrations were generally within 25 percent of measured values, and the timing of algal blooms was predicted accurately.