The U.S. Geological Survey, in cooperation with the Southwest Florida Water Management District, initiated a study to quantify the inflows and outflows in the Floral City, Inverness, and Hernando pools of the Tsala Apopka Lake Basin in Citrus County, Florida. This study assesses hydrologic changes in pool stages, groundwater levels, spring flows, and streamflows caused by the diversion of streamflow from the Withlacoochee River to the Tsala Apopka Lake Basin through water-control structures. A surface-water/groundwater flow model was developed using hydraulic parameters for lakes, streams, the unsaturated zone, and the underlying surficial and Upper Floridan aquifers estimated using an inverse modeling calibration technique. After calibration, the model was used to assess the relation between inflows and outflows in the Tsala Apopka Lake Basin and changes in pool stages.

Simulation results using the calibrated surface-water/groundwater flow model showed that leakage rates from the pools to the Upper Floridan aquifer were largest at the deep lake cells and that these leakage rates to the Upper Floridan aquifer were the highest in the model area. Downward leakage to the Upper Floridan aquifer occurred beneath most of the extent of the Floral City, Inverness, and Hernando pools. These leakage rates depended on the lakebed leakance and the difference between lake stages and heads in the Upper Floridan aquifer. Leakage rates were higher for the Floral City pool than for the Inverness pool, and higher for the Inverness pool than for the Hernando pool. Lakebed leakance was higher for the Floral City pool than for the Hernando pool, and higher for the Hernando pool than for the Inverness pool.

Simulation results showed that the average recharge rate to the surficial aquifer was 10.3 inches per year for the 2004 to 2012 simulation period. Areas that recharge the surficial aquifer covered about 86 percent of the model area. Simulations identified areas along segments of the Withlacoochee River and within land-surface depressions that receive water from the surficial aquifer. Recharge rates were largest in physiographic regions having a deep water table. Simulated heads in the Upper Floridan aquifer indicated the general flow directions in the active flow model area were from the northeast toward the southwest and then westward toward the coast, and from the southeast toward the northwest and then westward toward the coast, consistent with flow directions inferred from the estimated potentiometric surface map for May 2010. The largest inflow in the water budget of the Upper Floridan aquifer was downward leakage from the overlying hydrogeologic unit. The largest outflow in the water budget of the Upper Floridan aquifer was spring flow.

The calibrated surface-water and groundwater flow model was used to simulate hydrologic scenarios that included changes in rainfall rates, projected increases in groundwater pumping rates for 2025 and 2035, no flow for the 2004–12 period through the eight water-control structures in the Tsala Apopka Lake Basin, and the removal of the Inglis Dam and the Inglis Bypass Spillway on Lake Rousseau. Scenario simulation results were compared to annual average calibrated water levels and flows from 2004 to 2012. Simulated declines in the Tsala Apopka Lake pool stages under the 10-percent lower rainfall scenario were about 0.8, 0.3, and 1.3 feet (ft) for the Floral City, Inverness, and Hernando pools, respectively. Simulated groundwater levels under the same scenario declined up to 5.4 ft in the surficial aquifer and up to 2.9 ft in the Upper Floridan aquifer. Under the projected increases in groundwater pumping rates for 2035 that represented an increase of 36 percent from average 2004 to 2012 pumping rates, the simulated declines in the Floral City, Inverness, and Hernando pool stages were, in downstream order, 0.02, 0.06, and 0.04 ft. The largest drawdown under the projected increases in groundwater pumping rates for 2035 was 2.1 ft in the surficial aquifer and about 1.8 ft in the Upper Floridan aquifer. A scenario of decreased rainfall by 10 percent caused greater declines in water levels and pool stages than projected increases in groundwater pumping rates. The simulation with no flow through the eight Tsala Apopka Lake water-control structures resulted in simulated declines in average pool stage of 1.8, 1.9, and 0.5 ft in the Floral City, Inverness, and Hernando pools, respectively. The simulated removal of the two water-control structures in Lake Rousseau caused flow to increase at Rainbow Springs by 28 cubic feet per second, an increase of 4.7 percent from the average calibrated flow for 2004 to 2012.