Vertical Hydraulic Gradients

The spatial distribution of recharge and discharge at ENR is shown by a contour map of vertical hydraulic gradients for a representative day in the wet season (fig. 11). Recharge occurs over all of the western two-thirds of ENR, and, based on the direction and high magnitude of the gradients, recharge appears to be the dominant interaction with ground water at ENR. Concluding that recharge is the dominant interaction with ground water from vertical gradients is consistent with the results of Choi and Harvey (2000), who found that recharge was an order of magnitude greater than discharge in ENR on the basis of a water and chloride budget.

Figure 11. Distribution of vertical hydraulic gradients in Everglades Nutrient Removal Area (ENR), north-central Everglades, south Florida [larger image]

Both the steep land-surface slope across the ENR and the abrupt drop in water elevation across the western boundary levee contribute to the substantial driving force for recharge. Water level in the eastern part of ENR ranges between 14 and 15 ft and declines to between 11 and 12.5 ft on the western side of ENR. Water level then declines another 3.75 ft across the western boundary levee to the seepage canal located just outside the ENR (fig. 2). The water level in the seepage canal nearly always was held constant at just below 8 ft (relative to 1929 NGVD).

Figure 12. Water levels (top, A) and vertical hydraulic gradients (bottom, B) at Everglades Nutrient Removal (ENR) project site M102, north-central Everglades, south Florida. Vertical gradients are calculated relative to each well and to surface water. A positive gradient indicates the potential for downward flow, also known as recharge. [larger image]

The magnitude of vertical hydraulic gradients correlated positively with surface-water level in ENR (Choi and Harvey, 2000). Choi and Harvey (2000) showed that recharge in ENR was controlled by the rate of surface-water pumping into ENR, because that source of water strongly affects surface-water level, which in turn affects the driving force for recharge. Surface-water pumping generally is greatest in ENR during the wet season, which causes increases in ground-water recharge by as much as a factor of 3 during that season (Choi and Harvey, 2000).

Ground-water discharge occurred along a relatively narrow band on the eastern side of ENR (fig. 11). The driving force for discharge was the water level across the L-7 levee in WCA-1, which was 2.4 ft higher on average (between June 1996 and October 1998) compared with ENR. Discharge was correlated positively with water-level differences across the L-7 levee, with highest water levels in WCA-1 and greatest water-level differences usually occurring in the wet season (Choi and Harvey, 2000). As pointed out by Choi and Harvey (2000), ground-water discharge in ENR was small (approximately one-tenth) compared to recharge. Although the largest vertical hydraulic gradients were near levees, vertical hydraulic gradients also were discernable in interior areas of ENR. An example from site MP102, a site that consistently showed downward hydraulic gradients with a magnitude of approximately 0.005, is shown in figure 12. Most of the interior areas of the western two-thirds of ENR had similarly small downward hydraulic gradients, which, in total, accounted for approximately 6 percent of all recharge in ENR (Harvey and others, 2000).

WCA-2A

Highest vertical hydraulic gradients in WCA-2A were observed near the levee at site S10-C (Harvey and others, 2000). Similar to ENR, vertical hydraulic gradients at S10-C varied positively with the water-level difference across the levee. Normally, the water-level difference across the Hillsboro levee is about 2.8 ft, which is similar to the condition on the east side of ENR. Unlike the L-7 levee, the Hillsboro levee contains spillways, called the S10 structures, that are opened periodically to release water from WCA-1 into WCA-2A. During the wet season, it is not uncommon for the S10 spillways to be opened, allowing drainage from WCA-1 into WCA-2A. When spillways open, water levels in the S10C headwater canal drop rapidly while water levels increase rapidly in the tailwater canal. Vertical hydraulic gradients at S10C temporarily decline by approximately a factor of 3 when the spillway is open (Harvey and others, 2000, p. 222).

Figure 13. Water levels (A) and vertical hydraulic gradients (B) at Water Conservation Area 2A (WCA-2A) site F1, north-central Everglades, south Florida. Vertical gradients are calculated relative to each well and to surface water. A positive gradient indicates the potential for downward flow, also known as recharge. [larger image]

Sites in the interior of WCA-2A had lower vertical hydraulic gradients compared with the S10C site (Harvey and others, 2000). Representative water levels and vertical gradients with time at site F1 are shown in figure 13. Surface water-level measurements and ground-water head measurements at interior sites typically tracked each other very closely, often differing by only hundreds of a foot. Many of the vertical gradient calculations are below 0.002, corresponding to approximately a 0.03-ft head difference. Three-hundredths of an inch is near the limit of maximum achievable accuracy in measuring vertical head differences in the Everglades, because of the fundamental limits of the equipment used. Because of this limitation, determining with confidence the direction of the vertical hydraulic gradient in the interior of WCA-2A sometimes was challenging.

Despite the limitations on estimates of vertical gradients, patterns were apparent in vertical hydraulic gradients in interior areas of WCA-2A. The following interpretations are based on gradients between the shallowest instrumented layer in the aquifer (layer 2 instrumented with the GW-4 well) and surface water at all interior study sites. During the dry season, upward flow (discharge) was indicated on the western transect (F1 -F4 -U3) and neutral or possibly downward flow (recharge) was indicated on the E transect (E1 - E4 -U1) (fig. 3). During the wet season, there was greater variability in vertical hydraulic gradients. Surface-water and ground-water interactions in interior wetlands of WCA-2A appear to be affected most by occasional large water releases from WCA-1 after heavy rains near the end of the dry season. If those water releases are large enough to bring the water level in the tailwater canal above 13.5 ft (relative to 1929 NGVD), surface water from the tailwater canal will overflow a berm and flow to the south through the WCA-2A wetland. Flow over the berm usually occurs at least once and often occurs two or more times per year (Harvey and others, 2000). Over a period of weeks to months following a water release, a wave of water initially1 to 4 ft high propagates southward through the wetland. As the wave peak passes a site on the transect, the direction of vertical ground-water flux reverses, from discharge to recharge (fig. 13).

Summary of Vertical Hydraulic Gradients

Vertical hydraulic gradients from both ENR and WCA-2A are greatest near levees within approximately 0.5 km of levees. The direction of vertical gradients near levees generally was constant with time, although magnitudes varied by a factor of 2 or 3 depending on the surface-water level difference across the levee. Vertical hydraulic gradients near levees generally were smaller in WCA-2A than ENR (figs. 12 and 13). Because the water-level differences that drive flow beneath levees are similar at both sites, smaller vertical gradients probably are associated with the higher hydraulic conductivity (K) of the Surficial aquifer at WCA-2A (Harvey and others, 2000).

Vertical gradients were approximately 50 to 100 times smaller in interior areas of the wetlands compared to sites near levees. Nevertheless, those fluxes are important to the basin-wide water balance because of the much larger areas of wetland involved. Recharge and discharge in interior areas of the wetlands are not primarily the result of ground-water flow beneath levees. In ENR, the greater than normal topographic and water-level slope toward the northwest (because of land subsidence and human engineering of the ENR wetland), exerts a major control over recharge in interior areas of ENR. The release of water from WCA-1 into WCA-2A, a necessary part of water management in WCAs, also appeared to have an important effect on the direction of vertical flux in WCA-2A, causing relatively rapid reversals between recharge and discharge at research sites.