Simulation of root water uptake with heterogeneous vegetation cover and influences of urbanization and topographic attributes on hydrology of shallow water table environments

Field data collected from the proposed Tampa Bay Regional Reservoir site in west central Florida is utilized to calibrate and validate the HYDRUS-2D, a two dimensional finite element variable saturation model. The model is modified to compute Root Water Uptake (RWU) in two sub regions, the riparian zone and the pasture land in the same transport domain by normalizing the root zone over each part of the domain separately. Actual RWU during wet season is 40 percent higher than that during the dry season. According to modified model results RWU from riparian zone is 38 percent and 56 percent higher than pasture land during the dry and wet seasons respectively. The model is then coupled with a software package for digital terrain and river network analysis, RIVER TOOLS, to quantify the influences on local hydrology due to urbanization and ground slope for the same landscape. To visualize the complexity of the landscape and topography at the Long Flat Creek, a tributary of Alafia River in west central Florida, a Digital Elevation Model (DEM) is developed from one-foot contour, and used to extract the topographic settings of the landscape in the source area using the DEM. The impact of urbanization is simulated as increased fractions of impervious areas upstream, keeping the riparian corridor as a buffer zone. Multiple simulations are conducted with different fractions of surface sealing corresponding to different degrees of urbanization. A 50 percent increase in urbanization reduces the cumulative infiltration, base flow and transpiration by 48.4, 47.3 and 41.4 percent respectively. The corresponding figures for 75 percent increase are 23.6, 13.6, and 44.5 percent respectively. In addition, different simulations are conducted for different slopes to isolate the impacts of slope on local hydrology. The effect of surface slope on infiltration is found to be negligible due to the presence of high porosity sandy soil in the study area. The milder slopes generate lower base flow than steeper slopes. Comparison of depth to water table computed from HYDRUS-2D and TOPMODEL concepts revealed that depth to water table increases with increasing urbanization and computed water depths from both methods provided a good match with observed water levels.