GROUNDWATER TRACING AS A MEANS TO COLLECT DATA FOR GROUNDWATER MODEL DESIGN AND CALIBRATION

The Woodville Karst Plain of northwest Florida contains an extensive saturated conduit system with more than 60 km of mapped tunnels that connect a series of sinkholes to “karst windows” and large magnitude springs. The contribution zone for this system extends from north of Tallahassee to the Spring Creek springs on the Gulf Coast. This type of aquifer is best characterized as a dual or triple porosity system consisting of matrix, conduits, and possibly fractures. In such systems, groundwater velocities can span more than 30 orders of magnitude; more than 97% of the flow is typically through the conduits; and more than 95% of the storage is in the matrix. Groundwater protection efforts in north Florida are focusing on groundwater modeling as a means of delineating vulnerability zones however these efforts are met with significant challenges because of the dual porosity nature of the aquifer.

To help address this problem, artificial tracers were used to calculate groundwater velocities in the conduits over distances of approximately 300 and 2600 m. Both traces revealed groundwater velocities of between 0.02 and 0.03 m/s, Reynolds Numbers of 105, and nearly identical standardized breakthrough curves. When compared to natural tracer data such as from CFCs, tritium, or tritium/helium-3, which have revealed groundwater ages on the order of decades, these data indicate that the groundwater ages must be a composite of very old water from the aquifer matrix and very young water from the dissolved conduits. Effective groundwater modeling efforts in karst aquifers must, therefore, not rely on composite age dates for travel-time calibration/validation because they do not accurately represent conduit residence times. It is imperative that model designs honor the dual/triple permeability nature of the aquifer using travel-time data collected from the conduits and matrix and/or fracture permeability data collected from other testing methods.