Hydrology

1996 Abstracts

Assessing the Origin and Fate of Ground Water in the Florida Keys

Eugene A. Shinn, Christopher D. Reich, and Donald T. Hickey, U. S. Geological Survey, St. Petersburg, FL; John Karl Böhlke, L. Niel Plummerv, Tyler B. Coplenv and Eurybiades Busenberg, U. S. Geological Survey, Reston, VA; Jeffrey Chanton, William Burnett, Kevin Dillon and Reide Corbett, Florida State University, Tallahassee, FL.

The origin and fate of nutrient-rich ground water is being addressed in three ways: 1) by direct subsurface measurement of direction and rate of flow using dye tracers (Shinn et al.), 2) by detection and measurement of methane, radon, and an artificial tracer, sulfur hexafluoride (Chanton et al.), and 3) by measurement of environmental isotopes (H, He, O, C, N, S) and tracers (chlorofluorocarbons, SF6, coprostanol) (B"hlke et al.). Combined results indicate a potential for rapid local ground-water movement of anthropogenic contaminants along with a tendency for export of Florida Bay ground water toward the Atlantic.

1) Two circular well clusters were core drilled in 1 to 2 m of water on either side of an undeveloped portion of Key Largo. Each cluster consists of 8 equally spaced wells arranged in a 200-ft-diameter circle. The wells contain two piezometers; one screened between -6 and -7.6 m and one between -12 and -13.7 m. The zones are separated by a Portland cement plug and a subaerial unconformity. An identical well is located in the center of each cluster. A fluorescein dye solution was pumped into the shallow zone of the central well and rhodamine dye was injected into the deep zone. These zones were periodically sampled in the 8 monitoring wells and dye concentrations were determined with a fluorometer. Movement of ground water was found to be most rapid in the shallow zone. Both rhodamine and fluorescein were detected in the shallow zone of two bay-side monitoring wells 18 days after injection. Rhodamine in the shallow zones of monitoring wells indicates upward movement of ground water and is consistant with seepage discussed later. Dyes were detected only in the seawardmost monitoring wells in each well cluster, confirming that net flow is toward the Atlantic on both sides of Key Largo. Rates of flow range from 0.5 to 2.0 m/day. Tidal pumping combined with elevated water level in Florida Bay apparently drive net flow toward the Atlantic.

Pressure measurements taken from piezometers at 15 minute intervals show that bay ground-water pressure, where tides are minimal, is precisely tuned to Atlantic tides and ground-water pressures in the Atlantic. Ground-water pressure is negative under the bay when the tide is low in the Atlantic and reverses during high tide. The difference in elevation, 1 m during low tide in the Atlantic, causes seaward flow of bay-side ground water through the permeable limestone underlying the Keys. The gradient is slightly less during high tide, resulting in reduced flow in the opposite direction. Tidal pumping also causes seepage of ground water into the overlying water column. Other chemical data, summarized below, were obtained from water collected from seepage meters, well clusters, and other monitoring wells.

2) Direct measurements of seepage confirmed that areas with high concentrations of natural tracers also exhibit high seepage rates. The natural tracers, 222Rn and CH4, are elevated in ground water relative to surface water and serve as indicators of the release of ground water into Florida Bay. Two independent surveys confirmed that concentrations of both tracers are significantly enriched in areas of Florida Bay near the Keys (especially near Key Largo), relative to the northern, middle, and northeastern portions of the bay. Direct measurements of seepage were consistent with this assessment. At a site near Key Largo, seepage was observed to change in harmony with Atlantic tidal stage. Water seeped into Florida Bay when ground-water pressure was positive and reversed when negative. Rates of seepage varied from 15 to 40 ml/m-2/min-1. Measurements over a tidal cycle indicate that variations of tracer concentrations within Florida Bay waters are controlled by tidal stage on the Atlantic side of Key Largo.

Sulfur hexafluoride, an inert, artificial, non-reactive, non-toxic tracer, was injected into a well at the Key Largo Ranger Station during a rising Atlantic tide. Located about 1 m above bay level, the well is screened from -1 to -12.2 m. The tracer moved ~50 m and was detected in Florida Bay surface waters 6 hours later. The experiment was replicated with similar results. During low Atlantic tide, the plume moved toward the Atlantic and was detected in a monitoring well 3 m away within 3 hours. After several more hours, the plume passed through the well a second time. The maximum extent of plume movement toward the Atlantic could not be measured due to lack of additional monitoring wells. These results, however, confirm oscillation and lateral movement of ground water driven by Atlantic tides. The well, which is open to the limestone 1 m below the surface, is considered more representative of a septic tank or cesspool system than a modern disposal well.

Sulfur hexafluoride was injected into a modern disposal well between -18 and -27.4 m at the Keys Marine Laboratory on Long Key. Within 1 hour, the tracer was detected at -18 m, 5 m away on the Atlantic side of the injection well. The tracer was observed 4 hours later at -4.5 m, 10 m away from the injection well. At other surrounding wells, the tracer was observed at a variety of depths. These results indicate mobility of the tracer associated with channels or conduits within the karst aquifer, rising of the freshwater plume in a saline aquifer, diffusion, and a hydrologic gradient dipping toward the Atlantic. These results are consistent with dye, microbial, and phosphate studies being conducted jointly at this site by the USGS, University of South Florida, and Pennyslvania State researchers.

3) Although small-scale tracer studies indicate rapid, local, lateral movement of water in the subsurface, and water level monitoring studies indicate hydraulic potential for ground-water flow from the bay side to the ocean side, those results do not address directly the large-scale extent of N-S ground-water transport and origin of nutrients observed in ground water far offshore. We are testing the use of a variety of environmental isotopes (He, H, C, O, S, and N) and tracers (chlorofluorocarbons, SF6, coprostanol) to see what can be learned from them about ground water sources, the scale of the ground-water flow systems, and the fate of injected contaminants from the Keys. In February 1996, 33 sites were sampled to provide a preliminary comprehensive survey of surface and shallow ground water representing the potential recharge sources and the major regional types of ground water in the vicinity of the Keys and offshore areas to the north and south. Analyses of those samples for isotopes and tracers, plus major ions, nutrients, and dissolved gases, are in various stages of completion in various laboratories.

Measurements of H and O isotope ratios and salinities indicate at least 4 mixing components: 1) seawater, 2) meteoric water, 3) evaporated seawater, and 4) evaporated meteoric water. Bay-side ground water generally was enriched in 2H, 18O, and salinity compared to offshore marine surface water. These data are consistent with recharge by evaporated marine bay water during times of relatively high bay salinity. Ocean-side ground water had _2H, _18O, and salinity values generally equal to those of offshore marine surface water and consistent with ground water recharge by normal seawater. Several ground water samples from under the Keys and from short distances (less than a few hundred meters) offshore had isotope compositions consistent with transport of bay water to the ocean side, and one nearshore sample indicated transport of seawater to the bay side. Isotope data for tap water, waste water, and injected waste water all are consistent with a common freshwater source on the Florida mainland, and with mixing of waste water and marine ground water in the subsurface near a waste-water injection site.

Concentrations of CFC-12 indicate that the residence times of marine ground water on both sides of the Keys range from years to decades or more, and that the apparent ages generally are stratified. Degradation apparently has altered the concentrations of CFC-11 and CFC-113 in some samples. Minor CFC contamination was detected in water from the Port Largo canal and in waste water at the Keys Marine Laboratory, but it does not appear to be widespread (though degradation may have altered some occurrences). Additional samples are being analyzed for tritium and He isotopes that, when combined with CFC results, should provide more insight into the ground-water age distributions in the region. Ambient concentrations of SF6 are also being investigated for comparison with CFC data.

Nutrient analyses confirm that reduced ground water throughout the area contains significant amounts of ammonium. Concentrations of sulfide, methane, and bicarbonate were also elevated. The concentrations and isotopic compositions of sulfate and sulfide in the reduced ground waters are consistent with minor sulfate reduction. The concentrations and isotopic compositions of dissolved inorganic carbon (DIC) indicate varying contributions from both organic carbon oxidation and carbonate mineral recrystallization. Additional samples are being analyzed for 14C, which may provide evidence about the ages of carbon sources for the DIC and methane. The concentrations and isotopic compositions of dissolved N2 were nearly consistent with atmospheric equilibration over a small range in temperature, but there was evidence for small amounts of excess N2 in many samples that may have been derived from denitrification (reduction of nitrate to N2). Relatively large amounts of excess N2 in waste water and in mixed ground water near an injection site apparently were the result of denitrification of nitrate in the wastewater. Isotope analyses of nitrate and ammonium are underway and should provide important additional constraints on the fate of anthropogenic nitrate and the origin of ground-water ammonium.

Evaluating the Precipitation and Evaporation Patterns in and Around Florida Bay

Paul Willis, Peter Dodge and Frank Marks, NOAA, Atlantic Oceanographic and Meterolocial Laboratory, Miami, FL; David Sikkema and DeWitt Smith, Everglades National Park, Homestead, FL; Dean Churchill, Rosenstiel School of Marine and Atmospheric Sciences, University of Miami, Miami, FL; and Thad Pratt, U. S. Army Corp of Engineers, Waterways Experiment Station, Vicksburg, MS.

The freshwater input from precipitation, directly, and through flows resulting from over-land rain immediately to the north, is a crucial parameter in many analyses; e.g., salinity studies, ecological studies, and hydrological and ecological modeling of Florida Bay. This study addresses the effect of rainfall inputs directly into Florida Bay, and the rainfall over land areas to the north over the Everglades which are of paramount importance in the short term freshwater flows into the Bay. Because of the tropical convective nature of the summer rainfall in South Florida, gage measurements of rainfall only give representative rainfall measurements for long averaging periods and can often miss, or erroneously assess the magnitude of, significant convective rainfall events. The new digitized and recorded next-generation WSR-88D Doppler weather radars (NEXRAD) at Miami and Key West are capable of producing rainfall estimates over the entire Florida Bay/Everglades area at a time and space resolution not previously possible. These recently installed NEXRAD radars certainly have great potential for providing the rain input required at the shorter time, and the smaller spatial scales. Because of the difficulty of operating gages over open water areas, the radar data will provide the only rain measurement over some areas.

However, despite the superb time and spatial resolution, the quantitative measurement of surface rainfall, or freshwater flux, with radar is not without problems. An initial comparison between radar rainfall estimates and gage measurements at four locations (NOAA/NCDC Tech Rept 96-03, 1996) show in some cases good agreement, but in some cases rainfall totals differing by a factor of two. Some of the error is in the methodology of the gage/radar comparison, and gages are not trouble free instruments, i.e. wind effects, exposure, etc. But, some of the difference is due to the interaction of the radar and rain drop size distributions, i.e. a few large drops and a very large number of very small drops look the same to the radar, but may have a very different rainfall rate. The NEXRAD algorithms used to convert radar reflectivity to surface rainfall have been developed largely for mid-latitude subtropical regimes, and may not be fully suitable for the more tropical summer convective conditions of South Florida. The purpose of this study is to assess the accuracy of the present NEXRAD rain products, and to refine the radar-rain algorithms as required.

The primary hydrological product provided in the suite of NOAA NEXRAD products is the hourly digital precipitation array. This has a time resolution of one hour and a spatial resolution of approximately 4 km. The first part of this study is a comparison between this product and the rain gage network over the Florida Bay Everglades area. This comparison is shown for the 1996 rainy season. In addition to the hourly comparisons, comparisons for longer periods, 6, 12, and 24 hours are shown.

Further, it is assessed whether modifications to this product can form the basis for an adequate rainfall product for Florida Bay, or whether higher resolution radar data has to form the basis of a Florida Bay rainfall product. To this end case studies are presented where a developmental algorithm is applied to the full resolution radar data (6 min, and 1 km) and the results compared to gage measurements. The tuned Z-R relations based on the drop size distributions made in support of this program are applied here. This algorithm includes a separation of the radar data into stratiform and convective groupings and the application of an appropriate tuned Z-R relation for each, and a preliminary assessment of the efficacy of a convective-stratiform rain separation method for this area is made.

In addition to the radar gage comparisons, several field measurements have been made to tune and refine the radar-rain algorithms for use in the Florida Bay/Everglades system. These include airborne drop size measurements from several NOAA P-3 flights over the area in the summers of 1995 and 1996. These airborne drop size measurements are compared to the airborne radar data collected during these flights, and to the NEXRAD radar data. During the summer of 1996 a near continuous record of drop size distributions was also measured using a surface drop distrometer at a station in the Everglades National Park, and with a mobile van. These measurements, and preliminary indications of their relevance to the radar- rain algorithm, are discussed.

 

Flow Within Florida Bay and Interaction With Surrounding Waters

Thomas N. Lee and Elizabeth Williams, University of Miami, RSMAS, Miami, FL; Elizabeth Johns, NOAA, Atlantic Oceanographic and Meterological Laboratory, Miami, FL; and Ned Smith, Harbor Branch Oceanographic Institution, Ft. Pierce, FL.

Observational studies of the interaction and exchange of Florida Bay with the connecting coastal waters of the Gulf of Mexico and the Atlantic in the Florida Keys have been initiated to address several of the key scientific questions presented in the NOAA/COP Florida Bay Implementation Plan as critical to understanding the functioning of the ecosystem and future evolution from restoration actions. In particular, the research will address the following questions:

1) To what degree is the circulation of water within Florida Bay coupled to that of the surrounding coastal and oceanic environments?

2) What is the relationship of surface and groundwater flows through the Everglades to the salinity of Florida Bay?

3) Is the quality of the water flowing from the Bay contributing to the degradation of corals along the reef tract of the Florida Keys in the Atlantic Ocean?

Observational methods consist of a combination of synoptic shipboard surveys, in-situ moorings and Lagrangian surface drifters to describe and quantify the circulation within the Bay as related to local forcing and coupling with the waters of the Atlantic and Gulf. These observations will also help to provide necessary boundary conditions for future physical and biological models.

Field Work: The University of Miami study of Florida Bay circulation and exchange has completed five hydrographic surveys of Florida Bay and the surrounding waters, consisting of seasonal surveys of winter, spring and summer conditions that include the hydrological cycle from dry to wet seasons, and surveys before and following the passage of a major winter storm. A 20 ft. shallow draft catamaran, the R/V Miller, owned by the NOAA Southeast Fisheries Center has been equipped with a continuous flow thermosalinograph/fluorometer and a broad band 600 KHz Acoustic Doppler Current Profiler (ADCP) for conducting the hydrographic surveys in Florida Bay and nearby waters. A five mooring current meter array has been maintained since December 1995. Time series of current, temperature, conductivity and bottom pressure are being collected in western Florida Bay and adjacent southwest Florida shelf and Florida Keys coastal waters. Harbor Branch has maintained current meter moorings between Cape Sable and Marathon and in the major passes between the Keys. These data are used to determine the degree and variability of coupling of Florida Bay to surrounding waters. Un. Miami has obtained 3 month surface drifter trajectories from two satellite tracked drifters over the winter/spring period and has estimated the tidal and net transports through the major exchange passages between the Florida Keys over tidal periods during winter and summer seasons using a 600 KHz broad band ADCP mounted between the hulls of R/V Miller.

Initial Scientific Results: The Shark River discharge forms a narrow low-salinity plume that is adverted towards Florida Bay and the Keys to the southeast by a net background flow of 1 to 4 cm/s. Plume widths range from about 2 to 20 km and vertical stratification can be up to 3 psu per 3 m water depth near the river mouth. Plume shape and salinity structure is a function of seasonal and episodic river discharge, local wind forcing and tidal mixing. Extensive exchange between the waters of the western Florida Bay and the Gulf of Mexico occurred during the passage of an intense cold front, but little exchange occurred in the northeast interior portion of Florida Bay. High chlorophyll concentrations are observed in Key West Harbor and in the low salinity plume downstream of the river mouth, indicative of planktonic uptake of riverborne nutrients. Drifter trajectories show evidence of a cyclonic recirculation between West Cape Sable and Cape Romano. Synthesis of drifter trajectories from this project and those deployed in a coordinated DEP project indicate a net southeastward flow from the Gulf of Mexico to the Florida reef track through western Florida Bay that varies with season, stronger in the winter (3 to 4 cm/s) and weaker in summer (1 to 2 cm/s). Drifter trajectories are strongly influenced by local tide and wind forcing. A multiple linear regression model is used to explain approximately 70 to 80% of the subtidal variance of drifter currents due to local wind forcing. The cause and variability of the residual background currents are unknown at this time.

Net volume transport through Channels 5, 2 and Long Key Channel combined is estimated at 1500 - 2000 m3/s toward the southeast. Net flow through Seven Mile Bridge is also estimated about 2000 m3/s towards t h e southeast. This is equivalent to the peak river discharge onto the southeast U. S. shelf by all the rivers between Florida and Cape Hatteras. Approximately 1000 - 1500 m3/s flows through Long Key Channel alone, which is about 200 times greater than the peak fresh water discharge out of Shark River. These net flows have been estimated for single tidal periods in winter and summer and have not as yet been corrected for tidal inequalities. Also we found that all surface drifters deployed near Shark River and in the western part of Florida Bay were observed to exit Florida Bay toward the reef track through Long Key Channel. Therefore it appears that waters from Shark River and the seagrass die-off region of western Florida Bay will be passively adverted and dispersed toward the Florida Keys Marine Sanctuary (FKNMS) primarily through Long Key Channel by this strong mean southeastward flow. The advective/dispersal time-scale for materials in the Shark River Plume to reach the FKNMS is estimated from drifter trajectories at one to two months.

Long-term current meter data indicate that the maximum cross-shelf flow in the FKNMS from Key Largo to the Dry Tortugas occurs in the outer shelf of the Long Key region. After entering Hawk Channel and the reef track the drifter trajectories were either northeast or southwest depending on the direction of the alongshore wind. Several drifters were entrained by the Florida Current and one was ejected out of the Florida Current northeast of Cape Hatteras into a warm-core ring and recirculated within this ring as it slowly progressed to the southwest in the Middle Atlantic Bight (MAB).

 

Measuring and Modeling the Freshwater Discharge at the Everglades-Florida Bay Ecotone

Eric Swain and Eduardo Patino, U. S. Geological Survey, Water Resources Division, Miami, FL.

With the current interest in the flow across the Buttonwood embankment, a hydrodynamic model of the embankment and surrounding area would yield useful information. The majority of the flows to Florida Bay have been thought in the past to occur through the numerous sloughs and creeks in the area. However, there is no doubt that at higher water levels, overland flows occur in the mangrove swamps. This flow must cross the Buttonwood Embankment to reach the bay. Sufficient data will be necessary to model the area accurately. Topography is especially important, since the times and frequency of the overtopping of the embankment are of crucial interest. The surveying effort must be done with great attention to accuracy, since the topological relief in the area is flat and small differences create large effects on flow. Frictional resistance coefficients, always difficult to parameterize, are needed for the mangroves along the embankment. Determining these values in an analytic sense will be difficult. In many models, the frictional resistance term is a calibration parameter, adjusted to make a best fit to field data. Since field data of flows across the Buttonwood Embankment are not available, only water-level data measured at stations an appreciable distance from the embankment can be used for calibration. Additionally, different field data collection and data analysis techniques will be explored to determine the most suitable technique for gaging flows in the shallow and slow moving estuarine streams discharging into Florida Bay.

Simulations of Regional Climatic Patterns Which Impact the Florida Bay Water Cycle

Craig A. Mattocks, NOAA, Atlantic Oceanographic and Meterolocial Laboratory, Hurricane Research Division, Miami, FL.

Persistent, locally forced weather regimes, such as sea (land/lake/urban heat island) breeze circulations, generate thunderstorm complexes over the Everglades and coastal areas that provide roughly one-third of Florida's annual rainfall. Heavy rain episodes have profound effects on the salinity and nutrient content in Florida Bay, cause sewage system overflows, and result in eutrophication by fertilizer/pesticide/contaminant runoff from agricultural and industrial areas. In addition, the horizontal distribution of saturated soil or standing water sets up a complicated surface heat/evaporation feedback cycle, which alters the location of subsequent thunderstorm formation.

In a quest to increase the precision and accuracy of estimating the rainfall/evaporation over Florida Bay, the Center for Analysis and Prediction of Storms' (CAPS) Advanced Regional Prediction System (ARPS) cloud-/mesoscale atmospheric numerical weather prediction model has been configured to simulate the sea breeze regime over the Florida Peninsula and surrounding waters at high resolution. During the past year of this project, the ARPS model's capability has been extended (by incorporating warm rain and ice-phase microphysics) to predict the amount and the distribution of rainfall. Previously, only moisture convergence and the locations of dry convective cells could be predicted. Realistic precipitation patterns have been replicated along the sea breeze front in an August 1975 case study from the Florida Area Cumulus Experiment (FACE).

A two soil-layer force/restore surface energy budget, with multiple categories of soil and vegetation, has been used to simulate the atmosphere's response to diurnal heating. A by-product of the calculation of radiation and the vertical flux of water vapor is the prediction of evaporation - at the ground surface, from the fraction of foliage covered by intercepted rainfall, and from transpiration by leaves.

Because the development of the sea breeze, the onset of convective systems, and the dispersion of atmospheric pollutants are sensitive to the depth of the planetary boundary layer (PBL), its determination is of great practical concern. ARPS has recently been enhanced to predict the PBL height as a function of time. This innovation prevents the artificially excessive storage of heat within the turbulent layer near the earth's surface and, thus, eliminates the unrealistic explosive growth of convective cells when they suddenly penetrate "capping" atmospheric inversions. Inclusion of a full 3-D radiation physics package, for simulating the absorption/transmission of solar radiation by the atmosphere and clouds, now prevents "runaway" (not self-limiting) convection.

Working closely with scientists at the South Florida Water Management District (SFWMD), ARPS-predicted rainfall and evapotranspiration patterns will be used to drive both the Natural System Model (NSM) and the South Florida Water Management Model (SFWMM), then assess the relative sensitivities of the natural vs. present-day surface/groundwater flows to the precipitation. Several classes of typical summer day weather scenarios (east/west coast sea breeze, upper-level cold low instability, hurricane) will be defined, GIS soil/vegetation and land cover/use databases will be incorporated, and the surface conditions (where ponding exists, depth to groundwater, etc.) will be prescribed for the natural system vs. present-day conditions. By selectively reverting isolated areas of urbanization and drainage in the simulations, any distinctive microclimates (urban heat islands and associated shifts in the rainfall distribution, for example) which have emerged over the past century will be identified. The impact of proposed modifications to the groundwater flow and the future urbanization/development of pristine areas will also be investigated. Results from these experiments will provide a basis for the development of a fully-coupled hydrometeorological model in subsequent research efforts.

The achievement of these important milestones in the development of the ARPS atmospheric numerical weather prediction model makes the generation of high-resolution simulations of rainfall and surface winds, and their application as tactical decision aids (TDAs) in Everglades restoration management, a near-term possibility.

Last updated: 2/26/98
by: Monika Gurnée
gurnee@aoml.noaa.gov