Dissolution in a variably confined carbonate platform: effects of allogenic runoff, hydraulic damming of groundwater inputs, and surface–groundwater exchange at the basin scale

Abstract

In variably confined carbonate platforms, impermeable confining units collect rainfall over large areas and deliver runoff to rivers or conduits in unconfined portions of platforms. Runoff can increase river stage or conduit heads in unconfined portions of platforms faster than local infiltration of rainfall can increase groundwater heads, causing hydraulic gradients between rivers, conduits and the aquifer to reverse. Gradient reversals cause flood waters to flow from rivers and conduits into the aquifer where they can dissolve limestone. Previous work on impacts of gradient reversals on dissolution has primarily emphasized individual caves and little research has been conducted at basin scales. To address this gap in knowledge, we used legacy data to assess how a gradient of aquifer confinement across the Suwannee River Basin, north-central Florida affected locations, magnitudes and processes of dissolution during 2005-2007, a period with extreme ranges of discharge. During intense rain events, runoff from the confining unit increased river stage above groundwater heads in unconfined portions of the platform, hydraulically damming inputs of groundwater along a 200 km reach of river. Hydraulic damming allowed allogenic runoff with SI CAL < À4 to fill the entire river channel and flow into the aquifer via reversing springs. Storage of runoff in the aquifer decreased peak river discharges downstream and contributed to dissolution within the aquifer. Temporary storage of allogenic runoff in karst aquifers represents hyporheic exchange at a scale that is larger than found in streams flowing over non-karst aquifers because conduits in karst aquifers extend the area available for exchange beyond river beds deep into aquifers. Post-depositional porosity in variably confined carbonate platforms should thus be enhanced along rivers that originate on confining units. This distribution should be considered in models of porosity distribution used to manage water and hydrocarbon resources in carbonate rocks.

Vadose CO2 gas drives dissolution at water tables in eogenetic karst aquifers more than mixing dissolution

“…Inputs of CO 2 to lower water tables could have dissolved caves that were flooded as the water table rose with sea level. Later, collapse or river incision processes could have created entrances to these phreatic cave systems, forming Florida's famous springs (Gulley et al, 2011(Gulley et al, , 2013a(Gulley et al, , 2013b. Such a formation mechanism would explain how extensive phreatic cave systems can form in the absence of sinking streams, be recharged entirely by matrix flow and have conduits that extend beneath rivers (Gulley et al, 2013a).…”

Natural variations in calcite dissolution rates in streams: Controls, implications, and open questions

“…Though a few site-specific studies have examined controls on carbonate system dynamics using time series data [e.g., Liu et al, 2004;Groves and Meiman, 2005;Pu et al, 2014], more work is needed to fully understand what determines whether dilution or CO 2 is important in a given setting or under given conditions. Prior work has shown that dilution is an important control on dissolution where runoff from noncarbonate rocks, which is undersaturated, mixes with or displaces groundwater from carbonate aquifers, which is saturated [Gulley et al, 2013[Gulley et al, , 2014.…”

“…Dilute samples with high dissolved organic matter concentrations can produce substantial charge balance errors because dissolved organic carbon interferes with titrations and contributes to charge balance [Hemond, 1990]. Therefore, uncertainty in dissolution rates was assessed by alternately forcing charge balance on calcium and alkalinity [Gulley et al, 2013[Gulley et al, , 2014. Error bars depict dissolution rates calculated using this alternate charge balance forcing.…”

The role of antecedent groundwater heads in controlling transient aquifer storage and flood peak attenuation in karst watersheds

“…Spring reversals occur where phreatic conduits are hydraulically connected to surface streams and when river stage increases above groundwater heads. Consequent reversal of hydraulic gradient between aquifer and river causes river water to flow into the spring vent and conduits can transport floodwater intrusion kilometers away from river channels and allow it to exchange with aquifers 10s to 100 s of meters below the groundwater table (Crandall et al, ; Gulley et al, , ). Spring reversals in northern Florida during a 2009 flooding event were responsible for total transient aquifer storage volumes ranging between 10 4 and 10 5 m 3 at each spring that was documented.…”

“…Transient aquifer storage, defined herein as the temporary loss of river water to aquifers during floods and subsequent return to the river, has been recognized in contributing to flood peak attenuation (Cooper Jr and Rorabaugh, 1963;Zitta and Wiggert, 1971;Pinder and Saurer, 1971;Moench and Barlow, 2000;Chen and Chen, 2003). Transient aquifer storage typically occurs in watersheds that cross hydraulic boundaries, where upstream portions of the watershed are underlain by lowpermeability substrate and downstream portions of the watershed are underlain by substrate with higher permeability (Bonnaci, 1996;Gulley et al, 2013Gulley et al, , 2014. Examples include basins where rivers flow off low permeability crystalline rocks onto adjacent alluvial aquifers (Winter, 1999;Sophocleous, 2005;Payn et al, 2009), glacial terrain (Winter and Pfannkuch, 1976), volcanic terrain (Konrad, 2006) and dryland rivers in semi-arid regions (Costa et al, 1992).…”