Master of Science in Civil Engineering (MSCE)


Civil and Environmental Engineering

Document Type



Small-scale HEC-RAS models were used to analyze discharges used to test the Louisiana State University and the Louisiana Coastal Protection & Restoration Authority Expanded Small Scale Physical Model (ESSPM) of the Mississippi River. The HEC-RAS models extend from River Mile 228.4 (Baton Rouge) to the Gulf of Mexico that includes the ESSPM reach which begins at River Mile 173.5 (Donaldsonville). The model scales are 1:6000 horizontally and 1:400 vertically. Using the historic river discharges from 2008 through 2015, the small-scale numerical models proved capable of replicating observed stages along eleven sites of the lower Mississippi River within the targets of the statistical performance metrics: RMSE%, Bias and Pearson product-moment correlation, developed for hydraulic modeling of the Mississippi River. Specifically, RMSE% analyses of computed water depth versus the observed depth at each site was less than the 15% target all stations; the Bias metric was consistently less than 10 for all stations; and Pearson product moment coefficient was greater than 0.9 for 80% of the stations for each of the eight years of D15 modeling. Using the small-scale numerical models, the research intended to quantify the difference between a synthetic flow hydrograph used to test the ESSPM and actual flow data. Qualitatively, the stage hydrographs over the eight years indicate the actual discharge data produces six higher peak stages representing roughly 7.5 percent of the 2920 days in the model. Friction slopes for the D15 and Prototype model were compared and found to produce identical characteristics albeit the values of the D15 model friction slope was by its nature was fifteen times that of the prototype. Also, charting of the Froude Number demonstrated the expected equivalency. The HEC-RAS analyses revealed that total shear stress was equal at each of eleven observed data sites for the eight-year modeling period regardless of the inflow hydrograph. Total stream power, however, for the D15 model was roughly 15 to 20 percent higher using the actual river flows. Total stream power for the prototype model did not differ at the various data sites, while stream power at discharges above 575,000 cfs at both D15 and prototype scales were higher for the actual stream flows than for the synthetic hydrograph. The formulae for these parameters are the same except stream power is dependent on discharge where shear stress is dependent on hydraulic radius. Since the channels of both models have a relatively high width-to-depth ratio, the analyses demonstrated the maximum variance of the hydraulic radius to be approximately 12% while the maximum variance of the discharge was roughly 400%. Continued refinement and interpretation of this numerical model is an important element toward the interpretation of the results of the ESSPM and application toward understanding the dynamic hydraulic properties of the lower Mississippi River.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Willson, Clinton