calendar Add meeting dates to your calendar.

 

Paper No. 8
Presentation Time: 3:45 PM

SIMULATING KARST FEATURES IN THE UNCONFINED UPPER FLORIDAN AQUIFER IN FLORIDA, GEORGIA, AND ALABAMA


CRANDALL, Christy A., KATZ, Brian G. and BERNDT, Marian P., U.S. Geological Survey, 2639 North Monroe St, Tallahassee, FL 32303, crandall@usgs.gov

The Upper Floridan aquifer in Alabama, Florida, and Georgia is unconfined, containing numerous springs, sinkholes, and internal drainages associated with karst topography. Sinkholes provide direct recharge to the Upper Floridan aquifer; Springs discharge directly to streams along relatively rapid (<25years) groundwater flowpaths. A groundwater flow (MODFLOW) and particle tracking (MODPATH) model was used to predict travel times to selected wells and springs. To achieve a reasonable match between the median simulated age of particles and the average measured apparent age of water from age tracers (3H, SF6, CFC), karst features were explicitly incorporated into the model. Because few of the numerous sinkholes have been mapped a geographical information system (GIS) method was developed to incorporate sinkhole features into the model. Sinkhole distribution was determined using the USGS National Elevation Dataset and closed-basin depression routines in GIS to produce a polygon feature data set containing closed-basin depressions. Each polygon feature contains a centerpoint and an associated drainage area. Centerpoints (presumably the most likely drainage point) with associated areas were intersected with the model-grid and accumulated by model-grid cell to produce the total number of closed-basin depressions and total closed-basin drainage area per model-grid cell. The results were used to create a sinkhole multiplier array, defined as the total number of sinkholes multiplied by the total drainage area for each model-grid cell. The sinkhole multiplier array was adjusted, scaled and then applied to recharge, hydraulic conductivity, and porosity values. Preferential flow directions (fractures) were incorporated into the model using horizontal anisotropy; Springs were incorporated into the model using the drain package. The resulting simulated median age of water compared favorably with the average measured apparent age to within 2.6 to 10.4 years.
Meeting Home page GSA Home Page