The hydrodynamic model is forced by measured water levels on each side of the inlet and wind within the study area. Model accuracy is evaluated by comparing predictions with measurements from a current profiler located within the channel, mid-way between the Gulf of Mexico and the Laguna Madre. Results are compared for varying grid resolutions, 45m, 15m, and 5m. For each grid resolution option three vertical layer schemes are compared each with an increasing number of sigma layers and hence computational load. Resolution of the grid is a factor due to small unusually deep and shallow areas within the channel that can average to non-representative value in a coarse grid. Performance is quantified using RMSE, Bias, Central Frequency of 0.26m/s, maximum and minimum different model/measurement difference. Comparisons between model results and current velocity measurements are conducted for typical weather conditions and for conditions resulting in respectively strong flood and ebb flows. The most effective model is selected by balancing model accuracy and computational efficiency.
The selected calibrated model is then applied to predict current velocity in a hypothetical engineered channel that will provide a parallel path between the entrance of the Packery Channel and the Upper Laguna Madre. The parallel path will go through an existing system of residential canals and engineered lagoon, (Lake Padre) which is directed connected to Packery Channel. Three different channel widths are investigated. Proper dimensioning of the connecting channel geometry, and channel cross-section in particular, are important to predict current velocity and potential locations of higher erosion and scour. The impact of the width of the additional channel connecting the two water bodies is discussed in terms of current velocity during typical wind and tidal forcing as well as conditions resulting in strong flood and ebb flows.