publications > paper > summary of the hydrology of the floridan aquifer system... > the regional flow system
Summary of the Hydrology of the Floridan Aquifer System In Florida and In Parts of Georgia, South Carolina, and Alabama
By Richard H. Johnson and Peter W. Bush
Professional Paper 1403-A
The Regional Flow System
The existence of a regional flow system in the Floridan has been recognized since the early 1930's when Stringfield (1936) published his classic "Artesian Water in the Florida Peninsula." Stringfield was the first to identify a regional flow system in the carbonate rocks of Florida. His potentiometric-surface map of the Upper Floridan aquifer suggested the natural recharge and discharge areas and the general direction of ground-water movement from recharge to discharge areas. A major potential recharge area was implied in central Florida. Major discharge areas were implied in the coastal areas of Florida and Georgia.
Later Stringfield and others (1941) demonstrated the hydraulic continuity of this carbonate system across three States. They documented a widespread decline of head caused by large ground-water withdrawals at the coastal cities of Savannah and Brunswick, Ga., and Fernandina Beach and Jacksonville, Fla. By the early 1940's, the cones of depression around these cities coalesced to form a trough like depression in the potentiometric surface extending from Jacksonville to near Hilton Head, S.C., a distance of 150 mi. Since the early work of Stringfield and others (1941), there have been many hydrologic investigations of the Floridan aquifer system. However, the studies described in the Professional Paper 1403 series represent the first attempt to quantify regional rates of ground-water recharge and discharge.
A three-dimensional finite difference model (Trescott, 1975; Trescott and Larson, 1976) was used to study the regional flow system and to quantify flow rates. A regional model of the entire aquifer system and several subregional models were constructed for this purpose. The basic structure of the models is similar, although the subregional models depart from the regional model in detail to accommodate local conditions or features. The Upper and Lower Floridan aquifers were simulated as active layers; the surficial aquifer was treated as a source-sink bed; the lower confining unit (fine-grained clastic beds or bedded anhydrite) and the freshwater-saltwater interface (below and laterally) were assumed to be no-flow boundaries. However, the Upper Floridan, in a strip adjacent to its updip limit in Georgia, was bounded below by constant heads to simulate a small amount of upward leakage from the sand aquifer system beneath the Floridan. The updip limit of the system (p1.2) was considered a no-flow boundary, because it is generally a pinchout of the carbonate rocks.
The design, assumptions, and application of the regional flow model to simulating the predevelopment flow system have been described in a preliminary report by Bush (1982). Each node, or grid block, is 8 mi on a side and thus represents a 64-mi2 area. For simulation, the properties of an aquifer layer are assumed constant within each grid block. The value of a particular aquifer property in a grid block is an average over the block area. The subregional models have grids that are coincident with those of the regional model but smaller-4 mi on a side. Design of three of the four subregional models and preliminary simulation results are discussed by Krause (1982), Ryder (1982), and Tibbals (1981).
Regional simulation results including the major effects of present-day withdrawals are described in Professional Paper 1403-C. Results of subregional simulations and local pumping effects are described in Professional Papers 1403-D through F and 1403-H. (See fig. 1 for area described in each Professional Paper.)
A major assumption in applying the flow models is that flow in the Floridan behaves as flow in a porous medium. As discussed in the hydraulics section of Professional Paper 1403-C, this assumption is probably valid on a scale of several hundred feet (the scale of a typical aquifer test) except in the karstic spring areas and, thus, should be acceptable on the larger scale of the digital models. In the karst areas of central and northwest Florida, especially near major springs, conduit flow occurs on a local scale (hundreds to thousands of feet per Sinclair, 1978). At the scale of the regional and subregional models (with 4- or 8-mi grid-block spacing), the assumption of flow in the karst areas behaving as flow in a porous media probably is also valid.
The major features of the flow system can be illustrated and summarized by a potentiometric-surface map of the Upper Floridan aquifer. Plate 2 is an aquifer-wide potentiometric-surface map constructed from more than 2,700 water-level and pressure-head measurements made in May 1980. Superimposed on plate 2 are areas where the aquifer is unconfined, semiconfined, or confined.
The configuration of the potentiometric surface indicates that in South Carolina and Georgia, the direction of flow is generally east and southeast from the topographically high outcrop areas toward the Atlantic Coast and Florida. In Alabama and west Florida, flow is generally south from the outcrop areas toward the Gulf Coast. In peninsular Florida, the general flow direction is toward the Gulf and Atlantic Coasts from the central inland areas. Thus, it is implied that recharge occurs in the northern outcrop and peninsular inland areas, and that discharge occurs in the coastal areas.
The degree of confinement on the Upper Floridan is the characteristic of the system that most strongly influences the distribution of natural recharge, flow, and discharge. Most of the natural recharge, flow, and discharge occurs in unconfined and semiconfined areas. Potentiometric contours that are distorted as they cross streams indicate Upper Floridan discharge and typify unconfined and semiconfined aquifer conditions. Smoother, less-distorted contours are associated with confined parts of the system that are well "insulated" from surface drainage features.
The dominant feature of the Floridan flow system, both before and after ground-water development, is discharge from springs (the locations of which are shown on plate 2). Nearly all of the springs occur in unconfined and semiconfined parts of the aquifer system in Florida. Currently (early 1980's) the combined average discharge from about 300 known Upper Floridan springs probably ranges between 12,500 and 13,000 ft3/s-more than one-half of the total Floridan discharge. Potentiometric contours tend to be distorted around groups of springs, especially inland from the coast.
The impact of pumping from wells is evident in the confined areas as shown by cones of depression and areas of long-term water-level decline on plate 2. The steeper cones at Fort Walton Beach and Savannah are caused by lower transmissivity rather than higher withdrawal rates than other pumping centers. In contrast, large withdrawals near Orlando, northwest of Tampa, and in southwest Georgia-all located in areas of higher transmissivity and in unconfined or semiconfined areas-have produced only shallow localized cones of depression that cannot be shown at the scale of the regional potentiometric-surface map.
COMPARISON OF PREDEVELOPMENT AND CURRENT CONDITIONS
Before development, the flow system was in a state of dynamic equilibrium in which natural recharge to the Floridan aquifer system was balanced by natural discharge. It is estimated that about 67,000 mi2 was recharge area and about 27,000 mi2 was land discharge area (estimated total predevelopment discharge area, including offshore area, is 55,000 mi2). The total predevelopment recharge and, therefore, discharge simulated by the regional flow model was about 21,500 ft3/s. This is equivalent to 4.4 in./yr of water over the recharge area.
Springs and aquifer discharge to streams and lakes, nearly all of which occurs in unconfined and semiconfined areas (p1. 2), accounted for a very high percentage of the total predevelopment discharge. Simulated spring flow and aquifer discharge to streams and lakes were 88 percent of the 21,500 ft3/s simulated predevelopment discharge, or about 19,000 ft3/s. Diffuse upward leakage, which occurs primarily in confined areas, accounted for the remaining fraction of the total simulated predevelopment discharge, 12 percent or about 2,500 ft3/s.
Most of the recharge necessary to sustain spring flow and aquifer discharge to streams and lakes occurred relatively close to springs and to areas of point discharge to surface water bodies. Recharge to the Upper Floridan was highest in unconfined and semiconfined spring areas, averaging 10 to 20 in./yr. The proximity of high recharge to high discharge implies a vigorous and well developed shallow flow system in the unconfined and semiconfined parts of the Upper Floridan aquifer.
The estimated predevelopment discharge from the major ground-water areas of the Upper Floridan aquifer (as derived from simulation) is shown on figure 3. Regionally, and in every ground-water area except south Florida, the predominance of spring discharge and aquifer discharge to surface-water bodies over diffuse upward leakage is apparent. Not surprisingly, the five ground-water areas that are predominantly unconfined or semiconfined (Dougherty Plain-Apalachicola, Thomasville -Tallahassee, Suwannee, west-central Florida, and east-central Florida), although accounting for only about 50 percent of the Upper Floridan's area of occurrence, contribute nearly 90 percent of the simulated total predevelopment discharge. The Suwannee area is the most active part of the aquifer system in terms of ground-water flow; more than one-fourth of the total predevelopment discharge, close to 6,000 ft3/s, occurred in this area.
In the mostly confined Florida panhandle area, diffuse upward leakage occurred over the major part of the area. But confinement is lacking in the eastern third of the area and along the Upper Floridan outcrop area to the north, allowing direct aquifer discharge to streams. Similarly, in the southeast Georgia-northeast Florida-south South Carolina area, about three quarters of the simulated predevelopment discharge went to the four major rivers crossing the a really small northern outcrop. Diffuse upward leakage occurred over a much larger area, but accounted for a minor part of the area discharge. Only in south Florida, where no un-confined or semiconfined areas exist, was diffuse up-ward leakage the major form of predevelopment discharge; and the approximately 100 ft3/s simulated predevelopment discharge from south Florida represents less than 1 percent of the total regional discharge.
The general characteristics of the natural flow system have not been appreciably altered by ground-water development. Currently, as before development, discharge from Upper Floridan aquifer springs continues to be the dominant feature of the regional flow system; and the degree of confinement on the Upper Floridan is still the major hydrogeologic control on the distribution of recharge, discharge, and ground-water flow. However, similarity of the current flow system to the predevelopment flow system does not mean that ground-water development has not brought significant changes. In 1980, about 3 Bgal/d were pumped from the aquifer system (almost all from the Upper Floridan) for all uses, an amount equal to about 20 percent of the estimated predevelopment recharge or discharge. This pumpage has resulted in long-term regional water-level declines of more than 10 ft in three broad areas of the flow system, as shown on plate 2: coastal Georgia-- adjacent South Carolina--northeast Florida, west-central Florida, and panhandle Florida. The effect of ground-water development on the potentiometric surface is particularly evident at Savannah, Ga., and at Fernandina Beach and Fort Walton Beach, Fla., where deep cones of depression have formed. Saltwater has encroached as a result of pumping in some coastal areas, but its documented extent has been local.
Pumpage has been and continues to be supplied primarily by the diversion of natural outflow from the system and by induced recharge rather than by loss of water from aquifer storage. The aquifer system's transient response to changes in withdrawal rates dissipates fairly rapidly (days or weeks) in most areas. Thus on the average (that is, excluding the effects of seasonal changes in stresses), the current aquifer system is still considered to be approximately at equilibrium, except during periods following sustained increases in pumping.
Figure 4 shows the estimated current (early 1980's) discharge from the major ground-water areas of the Upper Floridan aquifer. Simulation suggests that current discharge is about 24,100 ft3/s, of which about three-fourths leaves the aquifer system as spring flow or discharge to surface-water bodies. The remaining one-fourth of the simulated discharge is split between pump-age (17 percent of total discharge) and diffuse upward leakage (8 percent of total discharge). Pumpage is now a major part of the aquifer discharge, especially in four of the ground-water areas. Diffuse upward leakage is markedly reduced from predevelopment rates in two ground-water areas where flow is sluggish (southeast Georgia-northeast Florida-south South Carolina, and south Florida); but regionally, diffuse upward leakage is not greatly reduced. Ground-water development has not resulted in significant movement of the divide separating the Suwannee area from the southeast Georgia--northeast Florida--south South Carolina area. The divide between the west-central Florida area and the east-central Florida and south Florida areas has shifted slightly southeast, thereby slightly enlarging the southern part of the west-central Florida area. The percentage of total system discharge that occurs in each area currently is not significantly different from the percentage contributed by the same area before groundwater development.
Simulation indicates that ground-water development has caused the rates of both downward and upward leakage between the Upper and Lower Floridan to increase about 16 percent. Lateral inflow to the system in the central Georgia outcrop area of the Upper Floridan has not changed significantly from its predevelopment rate. Simulation indicates a tremendous change from predevelopment in the rate of upward leakage from the Fernandina permeable zone. The change results from heavy pumping in the Jacksonville-Fernandina Beach-Brunswick areas, which has increased the vertical gradient from the Fernandina permeable zone toward the upper part of the Lower Floridan.
In summary, the major part of the flow system is largely unchanged from predevelopment conditions. Springs whose discharge is large are still the dominant feature of the system. Although pumping has caused recharge rates to increase locally, the greatest recharge still occurs near the springs. Even after development, ground-water flow remains sluggish in areas where the aquifer is deeply buried in comparison to flow in areas where the aquifer is close to the land surface.
U.S. Department of the Interior, U.S. Geological Survey
This page is: http://sofia.usgs.gov/publications/papers/pp1403a/regionalflow.html
Comments and suggestions? Contact: Heather Henkel - Webmaster
Last updated: 04 September, 2013 @ 02:04 PM(TJE)