Recharge and Discharge Estimates

Seepage-Meter Results from ENR

Seepage-meter measurements were made at ENR between August 1996 and April 1998. The absolute values of fluxes ranged from less than 0.04 cm/d (0.001 ft/d) to as much as 20 cm/d (0.66 ft/d). The typical uncertainty of a flux measurement was ±50 percent, based on replicate measurements from side-by-side seepage meters (Harvey and others, 2000). The lower end of that measurement range is the estimated detection limit of seepage meters at the study sites. The higher end of the measurement range is two orders of magnitude higher than average daily precipitation and evapotranspiration in the Everglades (Harvey and others, 2000).

Seepage-meter measurements indicated that recharge occurred in most areas of ENR, which is consistent with the distribution of vertical hydraulic gradients (fig. 11). Time-averaged discharge and recharge fluxes on a transect across ENR are shown in figure 16. Fluxes were above detection at eight out of nine sites, and, for comparison, exceeded average annual precipitation (0.4 cm/d) at four out of nine sites.

Figure 16. Time-averaged discharge and recharge fluxes across transect A-A’ in Everglades Nutrient Removal (ENR) project, north-central Everglades, south Florida. [larger image]

All of the seepage-meter data at ENR over approximately a 2-year period are summarized and displayed graphically in table 14 and figure 17, respectively. The direction of vertical exchange at a site usually was constant over time, with recharge occurring at sites on the western side and in the interior and discharge occurring on the eastern side. The greatest contrast in fluxes was between sites near levees and sites in the interior of ENR. Fluxes near levees typically were two orders of magnitude higher than fluxes in the wetland interior. The observed pattern of decreasing fluxes with distance from levees is consistent with results from the hydrogeologic model used in this study, suggesting an exponential decrease in flux with distance from the levee (fig. 18).

Seepage-meter measurements and hydrogeologic modeling agree that proximity to levees was the most important factor affecting the magnitude of recharge and discharge at ENR (fig. 18). Temporal variability in vertical fluxes generally was much smaller than spatial variability at ENR. Vertical fluxes at a location typically varied over time by less than a factor of 3, whereas fluxes varied spatially across ENR by over an order of magnitude.

Figure 17. Cumulative distributions of vertical fluxes measured with seepage meters at sites in the wetland interior in ENR (top, A), and at sites near levees in ENR and WCA-2A (bottom, B). Note scale differences on x-axis (vertical flux). [larger image]

Darcy-flux calculations from WCA-2A

At WCA-2A, recharge and discharge fluxes were computed using Darcy’s law. Daily calculations were made for all sites in WCA-2A for the period January 1997-December 1998.

Discharge occurred on the tailwater side of the S10C levee site at WCA-2A. The median flux was approximately 2.5 cm/d, which is similar to the sites nearest levees in the ENR (table 15). However, fluxes were more variable at the S10C site compared with ENR. Greater variability in fluxes at the S10C site compared with levee sites at ENR is consistent with the greater variability in surface-water levels and vertical hydraulic gradients discussed earlier. The important difference with ENR is that, at WCA-2A, the major water supply is through a spillway. When opened, the S10C spillway lowers the local water level in WCA-1 and raises water levels in WCA-2A in a matter of hours. In contrast, water is supplied to ENR by rapidly transporting water through a supply canal. Consequently, changes in the supply rate have less effect on the water-level difference between WCA-1 and ENR.

Recharge in the interior areas of WCA-2A was greater by approximately a factor of 4 compared with ENR (0.2 and 0.05 cm/d, respectively). Larger vertical fluxes in the interior of WCA-2A are caused, at least in part, by the larger hydraulic conductivity of the peat at WCA-2A. On average, hydraulic conductivity of peat is approximately a factor of 6 greater in WCA-2A compared with ENR. Also important in explaining larger vertical fluxes at WCA-2A are the larger surface-water level fluctuations compared with ENR. Larger fluctuations in surface-water level increase the driving forces for recharge and discharge. Although greater than ENR, recharge and discharge in the interior of WCA-2A still are relatively small, even when compared with average annual precipitation or evapotranspiration (approximately 0.40 and 0.35 cm/d).

Figure 18. Observed discharge and recharge and simulated fluxes in Everglades Nutrient Removal (ENR) project, north-central Everglades, south Florida. Both mean and median fluxes are plotted for each study site. [larger image]

Temporal variation in the interior fluxes at WCA-2A generally exceeded spatial variation, which is opposite of the result in ENR. Discharge occurred approximately 75 percent of the time at sites F1, F4, U3, and E4. Recharge occurred at site E1 approximately 75 percent of the time (table 15). Discharge and recharge each occurred approximately half the time at site U1. Fluctuations in recharge and discharge included reversals in flow direction at all sites except S10C (figs. 18 and 19). It was important, therefore, to characterize temporal variability of vertical fluxes at interior sites in WCA-2A. Flux percentiles for each site are shown in figure 19 and table 15. Over the period of the study, the vertical fluxes at most sites in the WCA-2A interior changed direction at least once. In general, recharge was more common at sites in the WCA-2A interior at the beginning of the 2-year period, and discharge was more common at the end (fig. 20). For sites E1, F1, and U3, the flux direction changed only once during the 2-year period. Flux direction changed various times at sites E4 and F4. The flux direction changed frequently at site U1 (every 20 to 40 days) on cycles that appeared to correlate with surface-water levels (fig. 20).

Figure 19. (left) Cumulative distributions of vertical fluxes calculated by Darcy’s law on Eastern transect (top, A), and Western transect (bottom, B), in Water Conservation Area 2A (WCA-2A), north-central Everglades, south Florida. [larger image]

Figure 20. (right) Vertical fluxes observed over a 2-year period on the Eastern transect (top, A), and Western transect (bottom, B), in WCA-2A, north-central Everglades, south Florida. [larger image]

There is no single method that is advisable for averaging vertical fluxes to account for spatial and temporal variability. For the purpose of a water balance (discussed in the next section), the approach used here was to average the median fluxes for "discharge" sites to characterize average discharge, and to do the same for "recharge" sites. The quotations mean that the definition of a discharge or recharge site depended on an ability to determine the dominant flux for a site, that is, where either discharge or recharge occurred at least 75 percent of the time. Neither discharge nor recharge dominated at site U1. That site was, therefore, not considered in the average calculations. The result of averaging across remaining sites was a discharge flux of 0.11 cm/d and a recharge flux of 0.18 cm/d. These results must be considered preliminary because data only were available for a relatively small number of sites. It is worth noting that other possible approaches to averaging the data set gave similar results, which provides some confidence in these preliminary estimates of discharge and recharge in WCA-2A.

Because water levels are always higher in WCA-1 compared to WCA-2A, it is clear that reversals between recharge and discharge are not consistent with the hypothesis that ground-water flow beneath levees is the main driving force. Instead, the results support the previous conclusions from hydrogeologic modeling, that ground-water flow beneath levees is effective mainly at distances less than 0.5 km from levees, and that vertical fluxes in the wetland interior are controlled by other processes. For example, regional gradients in water-table slope and surface-water level fluctuations are factors that could affect recharge and discharge in the wetland interior. The combination of those factors drives reversals between recharge and discharge at various timescales. For example, interannual rainfall variability, and its effects on regional water tables and water-resources management, drive multi-year fluctuations that explain the longest term trends in recharge and discharge. With only 2 years of data, however, those longer term trends are difficult to characterize except to indicate that long-term net recharge is correlated with multi-year periods of relatively wet conditions, whereas long-term discharge is correlated with multi-year periods of relatively dry conditions.

The present study showed that shorter timescale water-level fluctuations of weeks to months can control variations in recharge and discharge, including reversals in direction of fluxes. Water releases through spillways are important because those disturbances propagate as waves into the basin interior. Wave movement through the wetlands creates areas of high and low pressure over relatively short distances (hundreds of meters to kilometers). As discussed previously, recharge tends to occur beneath wave peaks whereas discharge tends to occur away from wave peaks. Another factor involved that was not previously discussed is the movement of pressure waves horizontally through the Surficial aquifer, in addition to wave propagation in surface water. Pressure waves travel at different speeds through different layers of the aquifer, causing simultaneous upward discharge and downward recharge from the sand layer beneath the peat. That process was most clear at the E4 site, where vertical hydraulic gradients indicate that the sand layer acts as a source of water that simultaneously discharges to surface water and recharges to the underlying limestone aquifer at times of low water in WCA-2A (Harvey and others, 2000, p. 225).

Seepage-meter Results from WCA-2B and WCA-3A

Seepage meters at the one study site in WCA-2B indicated persistent recharge over time. At site WCA-3A15, neutral conditions or slight recharge was indicated. Only seepage-meter data were available from those sites; therefore, interpretations are made without the benefit of vertical head gradients to compare for consistency at other sites. Seepage-meter measurements were spaced far apart in space and time in those areas because of the remoteness of these study sites. Consequently, evidence for short-term reversals in vertical fluxes at those sites could not be thoroughly evaluated. Nevertheless, the results from those two sites are valuable because they suggest a time-averaged trend of recharge at a site in WCA-2B referred to as 2BS, and fluxes that are below detection or slightly on the side of recharge at a site in WCA-3A referred to as 3A-15 (fig. 1).