ESTIMATION OF KARST AQUIFER TRANSMISSIVITY USING MODFLOW TO SIMULATE MONITORING WELL RESPONSE TO STORM EVENTS

Responses of monitoring wells to changes in stream stage or conduit hydraulic head during high discharge events provide a means to evaluate formation transmissivity on a large scale. This method can be particularly valuable in karst aquifers, where high transmissivities make it difficult to conduct traditional aquifer tests; however, diffuse recharge may be large during the storm events that produce conduit head changes. This study used data from the karstic Floridan Aquifer in northern Florida to evaluate the effect of neglecting diffuse recharge on transmissivity estimates. At the Santa Fe River Sink/Rise system, the Santa Fe River enters the Floridan aquifer at the River Sink, and re-emerges downstream at the River Rise. Between these locations, cave divers have mapped over 5 km of conduits. Hydraulic heads are monitored through time at locations within the conduit system and in wells monitoring the surrounding aquifer. Four major storm events have been recorded by the monitoring network. To incorporate recharge effects, aquifer response to storm events was simulated using a one-dimensional model implemented in MODFLOW. The time-varying hydraulic head at the conduit was simulated as a specified head boundary condition. Diffuse recharge was estimated as rainfall minus potential evapotranspiration calculated using the Penman-Monteith method. The transit time of recharge through the unsaturated zone is not known; given the relatively thin unsaturated zone and its expected high hydraulic conductivity, the simplifying assumption was made that recharge reaches the aquifer on the same day as rainfall occurs. Calibration of the model for individual monitoring well hydrographs yields estimated transmissivity for the aquifer between the monitoring well and conduit. Results suggest that the importance of including diffuse recharge in the analysis of transmissivity varies with the nature of the storm event (i.e., the relative magnitudes of diffuse recharge and conduit head change), but can be significant. Effort needed to construct a one-dimensional MODFLOW model is similar to application of the analytical solution.