4 The Computational Model 
The hydrodynamic model used was a modified version of the CH3D model 
developed originally by Sheng (1986), with many changes having been made 
by WES. The model can simulate time-varying 3D hydrodynamics due to 
tides, wind, river inflow, and density currents induced by salinity and 
temperature gradients. In the horizontal plane, computations can be made on 
Cartesian or boundary-fitted grids. In the vertical, the model employs sigma 
stretching, which permits the same number of layers in shallow and deep 
portions of the water body. As depths increase, the vertical extent of each 
layer increases proportionately. Johnson et al. (1989) give additional details. 
Numerical Grid 
In the horizontal, a variable, rectilinear grid, which has sufficient 
resolution and which was successfully used in previous WES studies of the 
harbors (Seabergh 1985), was used (Figure 12). The grid had a total of 
12,032 horizontal cells (128 cells in the east-west direction and 94 cells in the 
north-south direction) and was aligned to coincide with the Inner Harbor 
entrance channels. Minimum cell width was 235 ft. The grid extended 
4.2 miles seaward of the middle breakwater and covered an area of 
approximately 146 square miles. In the vertical, after some sensitivity testing, 
three sigma-stretched layers were used. 
Calibration and Verification 
After a careful review of the field data collected in 1987 (McGehee et al. 
1989), the periods of 7-11 August and 19-23 August were selected for calibration 
and verification of the model. The earlier period represented a large spring 
tide condition, while the later period was near a mean tide. Measured surface 
elevations at an offshore tide gauge were used for the ocean boundary 
condition and already contained the effects of winds. Wind data (velocity, 
magnitude, and direction) measured north of the Los Angeles main channel 
entrance were used for model calibration and verification. On the basis of 
Sensitivity tests, the winds in the model were blocked off for inner harbor 
channel cells to account for protection due to structures in the surrounding 
area, which is highly industrialized. Winds over the rest of the study area 
were assumed to be spatially uniform but varying in time. On the basis of 
Chapter 4 The Computational Model 
