The U.S. Geological Survey (USGS), in cooperation with the Allegheny County Sanitary Authority, developed a validated two-dimensional Resource Management Associates2 (RMA2) hydrodynamic model of parts of the Allegheny, Monongahela, and Ohio Rivers (Three Rivers) to help assess the effects of combined sewer overflows (CSOs) and sanitary sewer overflows (SSOs) on the rivers. The hydrodynamic model was used to drive a water-quality model of the study area that was capable of simulating the transport and fate of fecal-indicator bacteria and chemical constituents under open-water conditions.

The study area includes 14 tributary streams and parts of the Three Rivers where they enter and exit Allegheny County, an area of approximately 730 square miles (mi²). The city of Pittsburgh is near the center of the county, where the Allegheny and Monongahela Rivers join to form the headwaters of the Ohio River. The Three Rivers are regulated by a series of fixed-crest dams, gated dams, and radial (tainter) gates and serve as the receiving waters for tributary streams, CSOs, and SSOs.

The RMA2 model was separated into four individual segments on the basis of the U.S. Army Corps of Engineers navigational pools in the study area (Dashields; Emsworth; Allegheny River, Pool 2; and Braddock), which were calibrated individually using measured water-surface slope, velocity, and discharge during high- and low-flow conditions. The model calibration process included the comparison of water-surface elevations at five locations and velocity profiles at more than 80 cross sections in the study area. On the basis of the calibration and validation results that included water-surface elevations and velocities, the model is a representative simulation of the Three Rivers flow patterns for discharges ranging from 4,050 to 47,400 cubic feet per second (ft³/s) on the Allegheny River, 2,550 to 40,000 ft³/s on the Monongahela River, and 10,900 to 99,000 ft³/s on the Ohio River. The Monongahela River was characterized by unsteady conditions during low and high flows, which affected the calibration range.

The simulated low-flow water-surface elevations typically were within 0.2 feet (ft) of measured values, whereas the simulated high-flow water-surface elevations were typically within 0.3 ft of the measured values. The mean error between simulated and measured velocities was less than 0.07 ft/s for low-flow conditions and less than 0.17 ft/s for high-flow conditions.