3.1 Karst Drainage System
Karst aquifers are one major component of a complex natural drainage system composed of many surface and subsurface hydrologic components (Figure 6) Early on, groundwater professionals recognized that the prevalence and hydrologic significance of recharge from surface runoff draining rapidly into a karst aquifer from point sources (for example, open sinkholes and sinking/disappearing streams) required conceptualizing the aquifer in a non-traditional way. That is, not focusing exclusively on the saturated zone, but also considering the role of the unsaturated zone and its interrelationship with the saturated zone. This led to conceptualizing groundwater flow in the context of the karst drainage system as a whole. Figure 6 illustrates the hydrological zones of a typical karst aquifer, and the major recharge, storage, and flow components within the zones. All of the features contribute to the water budget of a karst aquifer, some being more significant seasonally and/or during and after storm events.
Figure 6 – a) Karst drainage system showing relationship between the karst aquifer (blue-shaded area) and the many different surface and subsurface hydrologic components of recharge, storage, and flow. The dark green box indicates the soil/epikarst area and the dark red box the main karst aquifer subsystems (Modified from Hartmann et. al. 2014). b) Cross-sectional diagram showing distribution of unsaturated (vadose) and saturated (phreatic) zone components. The interface between the two zones is defined by the position (or elevation) of the water table which may rise or fall significantly depending on flow conditions and influx of storm recharge. It is also dependent on the presence of overflow springs and relative vertical distance between overflow and underflow springs. Black dashed line and triangle indicate base-level water table elevation, blue dashed line and triangle indicate raised water table elevation under high- or flood-flow conditions. Vertical scale exaggerated. Modified from Trcek (2007).
Flow paths and water fluxes in karst drainage systems are complex, being controlled by the distribution and interconnection of zones of higher and lower permeability and higher and lower flow velocities; driven by relative differences in hydraulic heads and corresponding hydraulic gradients within and among these zones. In both the unsaturated and saturated zones, flow components exist that may be broadly characterized as “diffuse”. In diffuse flow zones, permeability is widely distributed throughout the rock matrix, micro-to-small-aperture fractures and pores, as well as small conduits and solutional openings, in which laminar, low-velocity flow occurs. In contrast, flow zones described as “conduit” include large solutional voids, pipe-like or channel-like solutional openings, and solution-widened fractures in which the volumetric flow rate is larger than in other types of aquifers. Both laminar and turbulent flow regimes, occur in conduits. In many karst aquifers, conduits are large horizontal dissolution features along bedding planes in soluble formations. These can be less than a meter to multiple meters thick and generally occur where a more soluble formation overlies a less soluble formation.
Flow paths and fluxes vary greatly across spatial and temporal scales, changing seasonally as well as during and after storms. Therefore, it is often necessary to sample and characterize karst aquifer hydrology under designated low-flow, average-flow, and high-flow conditions. In the saturated zone, water levels (potentiometric heads) measured in wells in highly-developed karst aquifers may exhibit significant spatial and temporal differences depending on the spatial variability of permeability zones, antecedent and existing flow conditions (low flow, base flow, or high flow), as well as the timing and variability in recharge that occurs during individual storms.
During seasonal increases in recharge, or after intense and/or prolonged storms, water levels may rise in the aquifer and flow paths may be temporarily re-activated in higher-elevation fractures and conduits that are in the unsaturated zone during low-to-base flow conditions. This mechanism is one cause of the oft-noted phenomenon of changes in groundwater flow directions and shifts in the position and configuration of groundwater divides during higher-flow conditions. Worthington (1991) proposed the term “overflow” routes for storm re-activated flow paths and “overflow springs” to identify the discharge from intermittent spring outlets re-activated during high-flow conditions. Conversely the term “underflow springs” identifies the perennial spring outlets that discharge the base flow of the karst drainage system (Figure 7). Other spring classification systems exist and are frequently used, such as the Meinzer classification (Meinzer, 1927) which is based on discharge and descriptive physical characterizations (for example, artesian versus gravity flow). Other classifications rely on geomorphologic or hydrogeologic descriptors, such as, thermal, bedding plane (or contact), geyser, vent, artesian and seepage, perched and seepage, seep, cave or tubular, offshore, onshore, karst, resurgence (river rise), estavelle (transitions between resurgence (or exsurgence) to a sink depending on ambient hydrologic conditions), instream, headwater, and so on (Copeland, 2003, Mathey, 1989). For karst springs, the advantage of the characterizing springs as underflow or overflow outlets is that it describes each spring’s hydrologic function for groundwater discharge and helps one develop a better conceptual model of the structure and functioning of the entire karst drainage system and its conduit flow.
Figure 7 – Photos showing discharge from overflow and underflow springs in a shallow karst aquifer (limestone) after a storm event: a), multiple small overflow springs discharging from outlets located at slightly higher elevations than the larger, underflow spring in the center of the photo; and b) view from downstream showing additional overflow spring outlets discharging from limestone bedrock in the wall of the ravine. Field photographs provided by Taylor (2021a).
The elevation of underflow springs has much control on the elevation of the water table at the outflow boundary of the karst aquifer, whereas the matrix hydraulic conductivity and the hydraulic capacity of conduits determines the slope of the water table and its fluctuation under varying hydrologic conditions (Ford and Williams, 2007). Underflow springs typically occur at a local or regional groundwater discharge boundary, in the vicinity of the lowest hydraulic heads in the aquifer, which is usually close to the elevation of a nearby base-level stream (White, 1988). Convergent tributary flow through the conduit network to a trunk conduit that discharges through a single large underflow spring is common (White, 1999), but many karst aquifers discharge via a conduit network with multiple underflow spring outlets.
Unsaturated zone hydrology is of critical importance to the investigation and characterization of most karst aquifers. As with other kinds of unconfined aquifers, karst aquifers are recharged by diffuse infiltration of precipitation. However, unique to karst is concentrated recharge of surface runoff through sinkholes and joints derived from surface streams that are diverted underground (as shown in Figure 1c where water flows from the Alapaha River into a river sink at Jennings Bluff-Avoca Tracts, Suwannee River Water Management District Lands, Hamilton County, Florida, USA). These are commonly called disappearing, sinking, or losing stream channels. Sinkholes (also known as dolines) exhibit a variety of shapes and sizes, but are simply closed surface depressions that collect and drain surface runoff into underlying fractures and conduits. Sinkholes may be part of the unsaturated zone, or, where they intersect the water table, part of the saturated zone.
In terms of the surface hydrology, sinkholes may appear to function as isolated or disconnected surface catchments or basins (“zero” order basins). However, it is important to recognize that the runoff collected and drained by a sinkhole is flowing somewhere via underground conduits, either within the basin defined by surface topography or to another topographic basin, spring or stream. When a sink (a losing stream reach or a sinkhole) moves water from one topographically defined basin to another, this is called stream piracy. It is erroneous to consider sinkholes as “non-contributing” drainage areas (Taylor and Doctor, 2017) because during and after storms, sinkholes that are normally dry can act as rapid natural drains that divert surface runoff to underground conduits. It is this type of concentrated recharge that renders karst aquifers more vulnerable to non-point-source pollutants, and other contaminants that spill or leak into the environment. Rates of drainage from sinkholes vary greatly depending on whether the depressions have open “throats” called swallets. The hydraulic capacity of swallets increases with size of the opening and decreases if filled with overlying soil or sediment. In some karst aquifers the rate of sinkhole drainage is impacted by the elevation of the water table or water that is backed-up in karst conduits. Additionally, some sinkholes or openings to the underground network (karst window) reverse the direction of flow depending on the rise or fall of the groundwater table and are called estavelles or a sinkhole flood.
Exercise 2 explains the types of sinkholes and invites the reader to consider how they influence recharge and flow.
Infiltration or recharge to karst aquifers is sometimes identified by the terms autogenic and allogenic. Autogenic recharge originates from infiltration of precipitation that falls on the area directly underlain by the karst aquifer, or from underground diversion of surface runoff that accumulated within the geographic boundaries of the area underlain only by the karstified bedrock (Figure 8). Allogenic recharge is contributed by surface runoff carried into the karst aquifer by sinking or disappearing streams, but which originates through precipitation falling on areas underlain by non-karstic bedrocks. Allogenic recharge contributions and the catchment areas they derive from must be included in the water budget for karst aquifers even though they are geographically and geologically outside the physical boundaries of the karst system. Mixing of allogenic waters with autogenic waters, and the timing and proportions of those recharge fluxes, often profoundly alters karst water chemistry. This hydrochemical signature can provide a useful set of parameters that serve as natural tracers for investigation of the internal conduit structure and hydraulic functions of karst aquifers. Section 5.3 “Water Tracing Tests” presents an example of using natural hydrochemical properties with end member mixing models to understand the groundwater contribution to surface-water streamflow.
Figure 8 – Geological block diagram illustrating the difference between allogenic and autogenic recharge to epikarst and conduits. Modified from Goldscheider and Drew (2007).
Another unique source of recharge, storage, and flow in the unsaturated zone of the karst drainage system is the epikarst (Mangin, 1975; Per Klimchouk, 2015). Epikarst is the uppermost weathered zone of carbonate rocks that possesses substantially enhanced porosity and permeability relative to the deeper parts of the rock mass. Epikarst stores and intermittently distributes infiltrated recharge water to the underlying karst aquifer’s unsaturated zone. Epikarst is an important storage zone that functions as a perched leaky aquifer. Some studies suggest that water storage in the epikarst can be more significant than storage in the saturated zone of the karst aquifer. The enhanced shallow porosity and permeability facilitates considerable lateral flow within epikarst, and, depending on its thickness, decreasing permeability with depth causes flow to converge towards solutional-enhanced, deeply penetrating, vertical fractures, conduits, and sinkhole drains (Figure 9). Recharge percolation from the epikarst to the deeper unsaturated zone occurs as diffuse seepage.
Figure 9 – Illustrations of the epikarst zone and its hydrologic function: a) Cross section showing recharge, storage, and flow characteristics within the epikarst zone (Modified from Doerfliger et. al., 1999). b) Geologic block diagram showing relationship between diffuse recharge from epikarst and concentrated recharge from sinkholes and sinking stream drainage and the underlying conduit network. The epikarst or unsaturated zone can have some perched water tables on top of less permeable sediments, such as a mudstone layer or chert layer within the epikarst-usually composed of highly weathered sediment or loess or sometimes the remaining clay and sand minerals post dissolution of all of the limestone (sometimes called residuum). Water flows through the unsaturated zone via:
(1) diffuse flow through soil or unconsolidated surface materials,
(2) concentrated flow through solution-enlarged sinkhole drains,
(3) diffuse percolation through vertical fractures, and
(4) diffuse percolation through permeable rock matrix.
Subterranean conduits shown as solid black are filled with ground water. Vertical scale exaggerated. Modified by Gunn (1986), from as it appears in Taylor and Greene (2008).