Geologic, vegetative and climatic controls on coupled hydrologic processes in a complex river basin: Lessons learned from a fully integrated hydrologic model

This dissertation documents the first implementation of an integrated transient 3D surface water-groundwater-land surface process model, ParFlow.CLM, to evaluate the interacting geologic, climatic and vegetative controls on water budget components and streamflow generation processes over the Santa Fe River Basin in North Central Florida. Model predictions indicate that evapotranspiration (ET) is the most important water balance component in the basin comprising 77% of rainfall. Geologic conditions and vegetative properties were found to exert primary control on the spatial variability of streamflow generation processes in the basin through their influence on the balance between rainfall, ET, runoff and infiltration processes. Climatic variability was found to provide primary control on the temporal variability of streamflow generation processes.;Model predictions indicate that in the upper basin more than 95% of streamflow is generated by recent near-stream rainfall. In contrast, in the lower basin the majority of streamflow is contributed by the Upper Floridan Aquifer, with the fraction of subsurface flow averaging approximately 77% at the outlet of basin. A global sensitivity analysis of the model revealed that the permeability of the Intermediate Aquifer System is the most influential factor driving hydrologic response throughout the SFRB. Particle tracking experiments predicted that the median age of streamflow in the upper basin ranges from approximately 1 day at the peak of storm hydrographs to approximately 7 days at the end of stormflow recession, with travel time distributions that vary over time but are generally well-fit with log-normal distributions at the peak of the storm hydrograph. The median age of subsurface contributions to streamflow in the lower portion of the basin was predicted to be approximately 17 years, and the travel time distribution for the subsurface contribution is well-fit by a gamma distribution showing fractal properties that do not vary significantly over time. The fraction of new stormwater versus old groundwater in the streamflow in the unconfined region, and thus the shape of the total streamflow travel time distribution, varies as a function antecedent conditions, storm magnitude, time during the storm, and assumptions regarding the contrast in hydraulic conductivity between high permeability zones and the porous matrix.