Modeling the hydrodynamics and salinity of the Pontchartrain Basin

The area of New Orleans was hit by Hurricane Katrina in August 2005. A large part of the city got flooded due to bad design, construction and maintenance of the levee system. In order to increase the level of protection of the city, the levees are heightened and strengthened in the framework of the Hurricane and Storm Damage and Risk Reduction System (HSDRRS). On the long term the restoration of coastal wetlands is also part of the program. Since the 1930’s wetland erosion in coastal Louisiana has been recorded. The main cause of the erosion is the canalization of the Mississippi River, land subsidence and sea level rise, and salt water intrusion by alteration of drainage patterns. One way of initiating wetland restoration is the construction of diversions. The Violet Diversion is the largest diversion planned in the Pontchartrain Basin. Water from the Mississippi River is diverted into Lake Borgne and the Biloxi marsh in order to decrease salinities in those target areas. In order to get more insight in the impact of the diversion on salinity (gradients), hydrodynamic and salinity modeling of the Pontchartrain Basin is desired. Due to lack of data and time, model calibration on salinity was not accomplished. The goal of this study is to model a dynamic equilibrium of yearly averaged salinity in the Pontchartrain Basin. The lessons learned from this study can be a start for subsequent modeling efforts of the Violet Diversion. In Delft3D-FLOW a grid was set-up to model tidal propagation in Lake Borgne. The grid consists of a little less than 53,000 nodes. The initial bathymetry and roughness are taken from the ADCIRC SL15 model. The model is forced with the amplitudes and phases of the ten most important tidal constituents. In order to calibrate the model, the tidal channels are enlarged and the bottom friction is decreased. The necessity of these changes was already proven by the application of the harmonic method on the Pontchartrain Basin, as well as the moderate results of previous model studies. The model is calibrated on tidal amplitudes (accuracy within 10%) and fluxes through the tidal passes (accuracy within 1%). Phases were considered less important. Salinity was implemented by simulating initial salinities and river discharges on top of the tide. Comparing 2D with 3D simulations, gravitational circulation occurs in the 3D modeling. This causes an increased salt water intrusion from the Gulf of Mexico towards Lake Borgne and the Biloxi Marsh. However, the salinities in this target area are too low in the dynamic equilibrium situation. This is explained by the underestimation of transport by tides and Mississippi River discharge towards Lake Borgne. Previous model studies proved that circulation around the continental shelf cannot be neglected for tidal transport. Also, the Mississippi River discharge can flow around the Birdfoot. Due to the choice of the model domain, that flow cannot occur in this study. Using the tide-calibrated model for salinity studies, it is recommended to model in 3D to simulate the gravitational circulation. Nontidal water level elevations and currents should be included in the boundary conditions. This can be achieved by enhancing the model domain to capture a larger part of the Gulf of Mexico and the Mississippi Birdfoot. Then the flow around the Birdfoot can also be simulated. Wind should also be added to the hydrodynamic simulations. The measured salinities and the target salinities show seasonal variation. Therefore future modeling should strive for real-time simulation by forcing the model with time-series. The diversion flow can then be varied per month or season.