THE EFFECT OF FLOOD MAGNITUDE ON CARBONATE DISSOLUTION RATES DURING SPRING REVERSALS

Carbonate dissolution rates are an important control on denudation rates in carbonate landscapes. Although kinetics of carbonate dissolution has long been studied through laboratory experiments, it remains largely unstudied in natural settings characterized by extensive groundwater – surface water mixing, such as karst aquifers, in which flooded rivers can have stages elevated above groundwater heads, directing flow into the aquifer. Flood waters are commonly enriched in organic matter (OM) and oxygen relative to groundwater, and when the OM is mineralized to CO₂, pH decreases, subsequently increasing calculated carbonate dissolution rates. To assess the effects of flood magnitude on dissolution rates and total dissolution, four spring reversals were sampled in north-central Florida, two at Madison Blue Springs (MBS), which discharges to the Withlacoochee River, and two at Peacock Springs (PS), which discharges intermittently to the Suwannee River. Peacock Spring was sampled during flooding in 2009 and 2010 when the river rose 2.8 m in 14 days and 8.5 m in 15 days, respectively. Madison Blue Spring was sampled during flooding in 2011 and 2012, when the river rose 2.3 m in 10 days and 0.7 m in 14 days, respectively. Saturation states with respect to calcite of the water were estimated using PHREEQc based on time series measurements of solute concentrations through the events, and these saturation states were used in kinetic equations to estimate dissolution rates throughout the flood. Peak dissolution rates correspond to flood magnitudes: 0.017 mmol/cm²day for the 2009 PS flood, 0.013 mmol/cm²day for the 2010 PS flood, 0.015 mmol/cm²day for the 2012 MBS flood, and 1.35 x 10^-11 mmol/cm²day for the 2011 MBS flood. At PS, dissolution rates and a cave surface area estimated from average observed diameter were used to calculate a 0.16 mm increase in cave diameter over the two month long flood in 2009. This value is substantially higher than the 0.03 mm/yr previously calculated for average regional denudation through rainfall and surface water geochemistry. These dissolution rates show that spring reversals are capable of inducing greater localized dissolution over two months than occurs regionally over an annual average, reflecting the highly heterogeneous nature, in both time and space, of dissolution in karst aquifers.