Skip Links

USGS Home Contact USGS Search USGS

Prepared in cooperation with International Upper Great Lakes Study Board
and U.S. Army Corps of Engineers

Detection of Conveyance Changes in St. Clair River Using Historical Water-Level and Flow Data with Inverse One-Dimensional Hydrodynamic Modeling

By David J. Holtschlag and C.J. Hoard

ABSTRACT St. Clair River is a connecting channel that transports water from Lake Huron to the St. Clair River Delta and Lake St. Clair. A negative trend has been detected in differences between water levels on Lake Huron and Lake St. Clair. This trend may indicate a combination of flow and conveyance changes within St. Clair River. To identify where conveyance change may be taking place, eight water-level gaging stations along St. Clair River were selected to delimit seven reaches. Positive trends in water-level fall were detected in two reaches, and negative trends were detected in two other reaches. The presence of both positive and negative trends in water-level fall indicates that changes in conveyance are likely occurring among some reaches because all reaches transmit essentially the same flow. Annual water-level fall in reaches and reach lengths was used to compute conveyance ratios for all pairs of reaches by use of water-level data from 1962 to 2007. Positive and negative trends in conveyance ratios indicate that relative conveyance is changing among some reaches. Inverse one-dimensional (1-D) hydrodynamic modeling was used to estimate a partial annual series of effective channel-roughness parameters in reaches forming the St. Clair River for 21 years when flow measurements were sufficient to support parameter estimation. Monotonic, persistent but non-monotonic, and irregular changes in estimated effective channel roughness with time were interpreted as systematic changes in conveyances in five reaches. Time-varying parameter estimates were used to simulate flow throughout the St. Clair River and compute changes in conveyance with time. Based on the partial annual series of parameters, conveyance in the St. Clair River increased about 10 percent from 1962 to 2002. Conveyance decreased, however, about 4.1 percent from 2003 to 2007, so that conveyance was about 5.9 percent higher in 2007 than in 1962.

Report PDF (1,93 MB) For additional information contact: Director U.S. Geological Survey USGS Michigan Water Science Center 6520 Mercantile Way, Suite 5 Lansing, Michigan 48911-5991 (517)887-8903 http://mi.water.usgs.gov Part or all of this report is presented in Portable Document Format (PDF); the latest version of Adobe Reader or similar software is required to view it. Download the latest version of Adobe Reader, free of charge.

Suggested citation:

Holtschlag, D.J., and Hoard, C.J., 2009, Detection of conveyance changes in St. Clair River using historical water-level and flow data with inverse one-dimensional hydrodynamic modeling: U.S. Geological Survey Scientific Investigations Report 2009–5080, 39 p.

Contents

Abstract

Introduction

Study Area

Problem

Purpose and Scope

Data and Models Provided by Other Agencies

Flow Data

Conventional Current-Meter Measurements

Acoustic Doppler Current Profiler Measurements

Water-Level Data

The Standard 1-D Hydrodynamic Model of the St. Clair River

Model Schematization

Model Geometry and Parameterization

Approach

Water Levels at Gaging Stations

Water-Level Fall between Gages

Reach Conveyance

Conveyance Ratios

Inverse 1-D Hydrodynamic Modeling of the St. Clair River

1-D Hydrodynamic Model of the St. Clair River

Inverse Modeling

Nonlinear Regression

Parameter Convergence

Parameter Magnitudes and Uncertainties

Parameterization Strategy

Simulation of Flow under Annual Parameterizations Derived from Inverse Modeling

Results of Conveyance Analyses

Inferring Reach-Conveyance Changes from Water-Level Measurements

Water-Surface Fall

Conveyance Ratios

Inferring Changes in Reach Conveyance from 1-D Hydrodynamic-Model Parameters

Time-Varying Parameter Estimation

Effective Channel Roughness Estimated from Conventional Measurements

Effective Channel Roughness Estimated from Acoustic Doppler Current
Profiler Measurements

Fixed-Parameter Estimation

Scenario 1962 to 1968

Scenario 1981 to 1985

Scenario 1996 to 2007

Integration of Time-Varying and Fixed-Parameter Estimation Results

Effective Channel Roughness Ratios and Conveyance Ratios

Inferring Changes in River Conveyance from 1-D Hydrodynamic Simulations

Summary and Conclusions

Acknowledgments

References Cited