Figure 1. The TIME model domain. [larger image]

Surface-water levels, ground-water heads, flow velocities, salt concentrations, rainfall, and meteorological data have been collected to support the model development. A plot of water-level data from three monitoring stations (SR, P35, and SH1) in Figure 2 illustrates the diminishing effect of tidal propagation through the mangrove marsh ecotone. Seasonal water depths in the wetlands typically range from 0 to 1 m and the semi-diurnal tide range along the Gulf coast is on the order of 1 meter.

Figure 2. Water-level data recorded in Shark River (SR), Shark River Slough (P35 and SH1), and Trout Creek, a tributary to Florida Bay. [larger image]

TIME Model Grids

The TIME model grid of 194 (E/W) by 174 (N/S) 500-m-square cells representing mean land-surface elevations is shown in Figure 3. The grid was generated from topographic data collected by helicopter GPS in the wetlands and bathymetric surveys of the coastal areas. The TIME model grid size was chosen to coincide with the resolution and simulation requirements of ecological models. A companion grid of vegetation types is shown in Figure 4. Twenty land-cover classifications derived from 1997 Landsat Thematic Mapper imagery were aggregated into seven vegetation types and one open water class. Vegetation classes are used to evaluate evapotranspiration processes, frictional-resistance effects, and wind-stress conditions. Frictional-resistance and wind-sheltering coefficients based on the vegetation map are shown in Figures 5 and 6, respectively. Resistance coefficients determined from field data are pro-rated to the 500-m grid from the 30.5-m Landsat imagery. Wind sheltering coefficients are based on the extent of emergent vegetation associated with each plant type. Evapotranspiration rates are evaluated from modified Priestly-Taylor equations correlating ET to solar radiation and water depth.

Figure 3. Land-surface elevations. [larger image] Figure 4. Vegetation types. [larger image] Figure 5. Frictional-resistance coefficients. [larger image] Figure 6. Wind-sheltering coefficients. [larger image]

Figure 7. Locations of road culverts, flow stations, and hydraulic control structures in the TIME model domain. [larger image]

• flow discharges, water levels, and salinities
• time sequences of grids of rainfall data
• meteorological parameters and wind conditions

Wetland, canal, culvert, and hydraulic structure water levels and discharges define inflow conditions at the freshwater model boundaries identified in Figure 7. Tide levels and salt concentrations at coastal stations are used to define flow and transport conditions along the southwest Gulf coast and Florida Bay boundaries.

Internal and boundary inflows from hydraulic structures and road culverts used to conduct TIME model simulations are shown in Figure 8. Inflows from the L-31W and C-111 canals, the sum of the S-12A-D releases and inflows through 19 culverts to the east and 49 culverts to the west are plotted in Figure 8. Tide levels at Lostmans, Broad, and Shark Rivers and the mean of water levels at McCormick Creek, Taylor River, and Trout Creek are illustrated in Figure 9. The significantly greater range of the semi-diurnal tides along the Gulf coast versus the wind-driven tides of Florida Bay is evident in Figure 9. An example input rainfall grid for the TIME model is shown in figure 10. Hourly precipitation grids are interpolated from 47 rain gages, circles in Figure 10, and 23 sets of NEXRAD rainfall estimates at un-gaged sites, triangles in Figure 10. (NEXRAD rainfall estimates were obtained from the University of Miami.)

Figure 8. Inflows specified as open boundary conditions. [larger image]

Figure 9. Water-surface elevations specified as open boundary conditions. [larger image]

Figure 10. Rainfall on June 16, 1999, between 1500 and 1600 hours in 500-m-square TIME grid resolution. [larger image]

Figure 11. Water levels simulated (light colors) and measured (dark colors) at P33, NP206, P36, P34 and P37. [larger image]

• time series of flow quantities at cell locations
• time series of mass and constituent fluxes through transects
• grids of flow quantities at specified times

Simulated and measured water levels are shown in Figure 11 as light and dark colors, respectively. Mean differences (measured minus simulated) for the two-month simulation are 0.026, -0.044, -0.047, 0.027, and 0.103 m at monitoring sites P33, NP206, P36, P34, and P37 respectively (gage locations shown in Figure 1). (This model simulation setup does not include groundwater exchanges.)

Figure 12. Flow speeds and directions simulated (red) and measured (blue) at SH1. [larger image]

Figure 13. Preliminary simulated flow depths and velocities at midnight on August 31, 1999. [larger image]