Home Archived October 29, 2018

South Florida Information Access (SOFIA)

publications > paper > modeling decadal timescale interactions between surface water and ground water in the central Everglades > discussion

Modeling decadal timescale interactions between surface water and ground water in the central Everglades, Florida, USA

5. Discussion

Study Area
Figures, Tables & Equations
PDF Version

Short-term estimates of recharge and discharge from interior areas of the Everglades were recently published (e.g. Krest and Harvey, 2003; Harvey et al., 2004; Harvey et al., 2005), and those fluxes show variability over timescales ranging across from weekly, monthly, seasonal, and interannual timescales. Longer-term (decadal) average estimates of recharge and discharge in interior areas of the wetland would be useful. The present paper addresses that need through application and testing of a coupled model of tritium transport and decay in surface and ground water of WCA-2A in the central Everglades. The form of the model not only permits estimates of recharge and discharge, but, through model sensitivity analyses and comparison with independent estimates of ground-water residence time and horizontal flow rates of ground water and surface water, allows some basic assumptions about spatial variability across the wetland, dominance of vertical compared with horizontal flow in the wetland, etc. to be tested.

Tritium was detectable to a depth of approximately 8 m in the 60-m deep Surficial aquifer beneath the central Everglades in WCA-2A. This contrasts with the results of Price et al. (2003) who found detectable tritium to a depth of 30 m in Everglades National Park. Because of our testing and justifications, we can use the simplified approach of estimating recharge and discharge fluxes for the study of Price et al. using Eq. (3) and compare it with our results from the central Everglades. Since the residence time of shallow ground water (based on 3H/3He ratios) was similar in the two studies (approximately 25 years), the greater depth of recharge of young waters in the southern Everglades suggests that recharge and discharge fluxes are probably larger than in the central Everglades, by perhaps a factor of 3 or more. Much of the southern Everglades overlie the highly indurated limestones of the Biscayne aquifer, which is known for its very high hydraulic conductivity (Fish and Stewart, 1991). The north-central Everglades, on the other hand, overlie a sandier unit of the Surficial aquifer, which has a lower hydraulic conductivity (Harvey et al., 2002). Peat thickness, which affects vertical water movement by retarding flow, is generally less in the southern Everglades compared with the north-central Everglades. Greater hydraulic conductivity in the Surficial aquifer and thinner peat support the expectation that recharge and discharge fluxes could be a factor of 3 higher in southern Everglades. However, this preliminary comparison of recharge and discharge fluxes remains a hypothesis at this stage until more estimates of recharge and discharge fluxes in the southern Everglades become available.

An exchange flux of 0.01 cm d-1 is an order of magnitude smaller than independent estimates based on modeling naturally occurring, short-lived isotopes of radium (Krest and Harvey, 2003) and Darcy-flux calculations made for the years 1997-2002 (Harvey et al., 2004). There are several possible explanations for this difference. One is uncertainty in assuming that actual tritium concentrations in surface water are equal to the measured tritium in precipitation. If discharge of deep ground water to the canal bottom at the upstream end of the WCA-2A flow system is substantial, then tritium in Everglades surface water may have been overestimated. The potential effect of this error was investigated by reducing the size of the bomb spike in our simulation (by about 50%), which would decrease the ground-water age estimate, from about 90 to 50 years. The effect on the exchange flux estimate is to increase recharge and discharge flux estimates by about a factor of 2, which is not nearly sufficient to explain an order of magnitude difference between the tritium-based estimate and the other independent estimates.

The order of magnitude differences between estimates of recharge and discharge made using long (ground-water tritium modeling) and short (radium modeling in peat porewater and Darcy flux calculations) is probably not the result of bias or error in either method. Instead of great inaccuracies in one approach or the other, we believe that the order of magnitude disagreement between tritium modeling and Darcy flux calculations is more likely the result of comparing techniques that are sensitive to different timescales of interactions between surface water and ground water. The relatively short timescale calculations based on measurements in peat are good at characterizing high fluxes that occur periodically but are short-lived and switch direction frequently (Harvey et al., 2004). Those short-lived fluxes are mainly effective in causing exchange between wetland surface water and peat porewater. Tritium modeling in shallow ground water is insensitive to large and short-lived fluxes that frequently switch direction, because those events only have a minimal effect on tritium in ground water. Instead, tritium modeling is sensitive to the annual and longer term fluctuations associated with factors such as climatic variability, because those longer timescales are effective in exchanging surface water with ground water at depths up to 8 m in the Surficial aquifer. Thus, we believe that it is possible for two independent estimates of recharge and discharge to differ substantially because of different averaging timescales. The correct estimate to use for any particular investigation will depend on the particular problem of interest and its associated timescale. For example, recharge and discharge operating on short timescales could be highly relevant to understanding transport, storage, and re-release of phosphorus from peat porewater. In contrast, longer timescale interactions between surface water and the sand and limestone aquifer could be important in understanding the extent to which relict sea water and its associated dissolved salts are being mobilized deep in the aquifer and discharged to surface water.

5.1. Implications for Everglades water quality

The present study determined the long-term average rate at which the water and solutes presently flowing through the Everglades wetlands are exchanged for water discharging from the sand and limestone aquifer beneath the wetlands. In addition to estimating recharge and discharge as a vertical flux (0.01 cm d-1), the exchange of water can be expressed as a fraction of surface water exchanged per day (0.4%). Due to water management practices and agricultural runoff, surface waters in the central Everglades tend to be contaminated with excessive levels of nutrients, salts, and mercury (Harvey et al., 2002). In the past several decades, the application of best-management practices on farmlands adjacent to the Everglades has helped improve the quality of water flowing into the Everglades. The retention of recharged surface water and its solutes for decades in shallow ground water could have legacy effects for the future, because contaminants that were recently recharged potentially could be returned very slowly to surface waters over a period of decades. Of particular importance could be the recharge of phosphorus over the past few decades, which potentially could be returned to surface water in the next few decades with discharging ground water even if the quality of agricultural runoff continues to improve. The likely timescale at which contaminants now stored in peat porewater and the limestone and sand aquifer are returned to surface water could be decades. Our findings provide a reasonable hydrologic basis for models of that phenomena. One possible result of improved modeling of phosphorus transport is that it could become apparent that rapid initial improvements in water-quality restoration that might be achieved in routing incoming water through treatment wetlands might be difficult to sustain in future years. We stress that these ideas are highly preliminary, and that they need to be thoroughly tested by combining the hydrologic model presented here with biogeochemical data, along with further improvements in components of the coupled surface-sub-surface- biogeochemical model. Only through such improvements can predictions for future water quality be made more reliable.

< Results | Summary >

| Disclaimer | Privacy Statement | Accessibility |

U.S. Department of the Interior, U.S. Geological Survey
This page is: http://sofia.usgs.gov/publications/papers/timescaleint/discussion.html
Comments and suggestions? Contact: Heather Henkel - Webmaster
Last updated: 04 September, 2013 @ 02:04 PM(KP)