May data link: Environmental Hydrodynamic Modeling of Chicago Waterways

by Herbert Morgan, NCSA

"How big, how long, how deep?" asks the FAQ page of the Friends of the Chicago River website. The answer: The widest point of the Chicago River System is 180 feet. The system itself is 156 miles long, and, at its deepest point, 20–21 feet.

Mariano Cantero, Ph.D. candidate in Civil and Environmental Engineering (CEE) at the University of Illinois, has conducted high-resolution computer simulations of gravity current phenomena that are being used to estimate several parameters in the Environmental Fluid Dynamics Code (EFDC). This code is used by the US Environmental Protection Agency (USEPA) to predict flow and water quality in rivers and estuaries.

Marcelo Garcia, CEE Siess Professor at the University of Illinois, describes the EFDC code as a 3-D code that his group wants to use for simulating the flow of the entire river system in Chicago -- all 156 miles. "It's going to have to be parallelized," says Garcia, "to be able to use it efficiently because [the Chicago River] is such a large system, and that's where the collaboration with NCSA is needed."

Working with NCSA's Performance Engineering and Computational Methods (PECM) group, Garcia's team, which also includes S. (Bala) Balachandar, a Theoretical and Applied Mechanics professor at the University of Illinois, was able to resolve some problems. Through discussions with Cantero early in the process, Greg Bauer, a PECM research programmer, provided what he calls "technical-support involvement": memory usage, data file format issues, a work-around for a bug in an SGI library, and additional assistance.

The researchers' current project stems from a discovery, in the winter of 1998, of a bi-directional flow in the Chicago River by scientists from the Illinois Water Science Center, US Geological Survey. On the river's surface, it appeared that the water flowed away from Lake Michigan as one might expect. At riverbed level, however, water flowed toward the lake. This discovery was particularly disturbing because it meant that poor-quality water from the North Branch was potentially flowing into the higher-quality, pristine lake water.

Unlike some scientists who thought the notorious Chicago wind was the factor, Garcia suspected that the problem stemmed from a layer of denser fluid -- a density current -- that flowed below a layer of fluid that was less dense. In 2001, he and his colleagues at the time constructed a three-dimensional hydrodynamic simulation that confirmed his hypothesis. They conducted what Garcia called a "brute force" approach using FLOW-3D to model the whole Chicago River. It took about a week of computations to track a density current front moving all the way from the junction of the North Branch and the Chicago River main stem to Lake Michigan.

Now, the researchers are trying to understand -- in terms of the Chicago River and in anticipation of broader applications -- the factors (i.e., temperature, salinity, suspended sediment, etc.) that cause the density differences. The Chicago River's lock system creates a problem called "lock exchange": one side of the partition (the lock) has a heavy fluid and the other side has a lighter fluid. As in the case with the Chicago River during winter months, heavy fluid can be laden with salt-application runoff from city streets; lighter fluid can be lighter because it is cleaner. When a heavy fluid is introduced into a lighter fluid, the heavy fluid spreads, creating an intrusion.

Salt and temperature are similar in that they move with the fluids. But when density difference is caused by particulate matter (particles such as sand, silt, etc.), they do not necessarily follow the flow of the fluids. They may eventually settle down. So that introduces interesting physics, according to Balachandar. "We are studying whether it is salinity or temperature causing the density difference versus particulate matter...and we have developed a mathematical model in order to describe this problem."

"Some of the things," adds Garcia, "that Mariano has brought to bear are ...mathematical modeling and ... numerical solutions. For that we need supercomputers." As a result, NCSA became vital to the researchers because they needed a lot of computer power. And from Garcia's viewpoint, there is a "good synergy" with NCSA.

Research is funded by Metropolitan Water Reclamation District of Greater Chicago (MWRDGC) and the Coastal Geosciences Program of the US Office of Naval Research (ONR).

FastCounter by LinkExchange