Research papersClimate control of decadal-scale increases in apparent ages of eogenetic karst spring water

- Apparent age of Florida spring discharge increased with time from 1997 to 2013.
- Apparent age increase unaffected by pumping or recharge from Tropical Storm Debby.
- Increase in apparent age caused by excess precipitation from ∼1960 to 1980.
- Excess precipitation results from El Niño years during warm phase AMO.
- Groundwater management should consider effects of global climate on recharge.

Water quantity and quality in karst aquifers may depend on decadal-scale variations in recharge or withdrawal, which we hypothesize could be assessed through time-series measurements of apparent ages of spring water. We tested this hypothesis with analyses of various age tracers (3H/3He, SF6, CFC-11, CFC-12, CFC-113) and selected solute concentrations [dissolved oxygen (DO), NO3, Mg, and SO4] from 6 springs in a single spring complex (Ichetucknee springs) in northern Florida over a 16-yr period. These springs fall into two groups that reflect shallow short (Group 1) and deep long (Group 2) flow paths. Some tracer concentrations are altered, with CFC-12 and CFC-113 concentrations yielding the most robust apparent ages. These tracers show a 10-20-yr monotonic increase in apparent age from 1997 to 2013, including the flood recession that followed Tropical Storm Debby in mid-2012. This increase in age indicates most water discharged during the study period recharged the aquifer within a few years of 1973 for Group 2 springs and 1980 for Group 1 springs. Inverse correlations between apparent age and DO and NO3 concentrations reflect reduced redox state in older water. Positive correlations between apparent age and Mg and SO4 concentrations reflect increased water-rock reactions. Concentrated recharge in the decade around 1975 resulted from nearly 2 m of rain in excess of the monthly average that fell between 1960 and 2014, followed by a nearly 4 m deficit to 2014. This excess rain coincided with two major El Niño events during the maximum cool phase in the Atlantic Multidecadal Oscillation. Although regional water withdrawal increased nearly 5-fold between 1980 and 2005, withdrawals represent only 2-5% of Ichetucknee River flow and are less important than decadal-long variations in precipitation. These results suggest that groundwater management should consider climate cycles as predictive tools for future water resources.