Purpose and Scope

The purpose of this report is to document the use of data compiled since the report by Hutchinson (1983) to analyze the potential effects that modifications to the shipping channels have had or may have had on the potentiometric surface and groundwater salinity in the coastal areas surrounding Tampa Bay (fig. 1). Existing groundwater level and geochemical data from the USGS National Water Information System (USGS, 2016) and data portal for the Southwest Florida Water Management District (SWFWMD, 2018a, b) were examined to investigate the effects of deepening or widening of navigation channels on the interconnection of the saline bay-water and groundwater, potential downward leakage of saline bay-water into the underlying aquifer, and the location and thickness of the saltwater front in the groundwater systems surrounding the bay. Hydraulic head and chloride concentration, in milligrams per liter, were examined in wells in the coastal areas surrounding the bay spanning the period 1976–2022.

The scope of this report includes analyses of available groundwater well data, including hydraulic head, geochemical, and well construction reports for wells within coastal areas surrounding the bay. Available data were analyzed to identify the factors that influence hydraulic heads and the locations and movement of the saltwater front. These factors include changes to mean tidal stage, groundwater withdrawals, and variations in rainfall and evapotranspiration. A second phase of this study, which is proposed but not part of this report, includes the development of a more refined groundwater model that incorporates variable density (freshwater to saltwater) to simulate the response of the groundwater system to breaches in the confining surficial sediments underlying Tampa Bay.

Groundwater data are available from several sources in the Tampa Bay region. Key sources of data used for this study are records from the Regional Observation and Monitor-well Program (ROMP) maintained by the SWFWMD and additional records maintained by the USGS (SWFWMD, 2018b; USGS, 2016). An inventory of available well data was compiled, and the groundwater data at well sites were analyzed for trends in water level and water quality. Statistical analysis of trends was based on the Kendall tau (τ) test (Helsel and others, 2020), using the R statistical package (R Core Team, 2022). The water-quality parameter of interest was chloride concentration, which is related to movement of the freshwater/saltwater interface.

Maps of the elevation of the potentiometric surface of the Floridan aquifer system (FAS) were also analyzed for trends related to hydrologic changes. A potentiometric surface is an imaginary surface representing the interpolation of static heads of groundwater within tightly cased wells that tap a confined aquifer (USGS, 2013). The potentiometric surface is equivalent to the water-table level within an unconfined aquifer and is typically mapped with contours for a particular time. The slope of the potentiometric surface is referred to as the “hydraulic gradient” and can be used to infer the direction of groundwater flow, given that groundwater flows down-gradient. The magnitude of the hydraulic gradient and an aquifer’s hydraulic properties determine the actual velocity and volumetric rate of the groundwater flow. The direction of groundwater flow is oriented perpendicular to the potentiometric contours, with flow traveling across them from higher areas of the potentiometric surface toward lower areas. The location, movement, and thickness of the saltwater front is affected by the hydraulic gradient and resulting groundwater flow rate. Trends in saltwater intrusion can be slowed, halted, or reversed by changes to the potentiometric surface and hydraulic gradient.

The relations between these trends and previous dredging activities within the bay, sea-level rise (SLR), and changes in water use were examined and are discussed. Regional changes within the hydrologic system are also compared to the predictive results of the related study conducted by Hutchinson (1983) surrounding previous channel modification activities.

Description of Study Area

Tampa Bay is adjacent to the Gulf of Mexico and is Florida’s largest open-water estuary (Goodwin, 1987). The bay is composed of several interconnected bays and lagoons grouped into seven segments: Hillsborough Bay, Old Tampa Bay, middle and lower Tampa Bay, Boca Ciega Bay, Terra Ceia Bay, and the tidal reach of Manatee River (fig. 1). Tampa Bay has a surface area of approximately 329 square miles and has an average bottom depth of 12.8 ft below the North American Vertical Datum of 1988 (NAVD 88) (fig. 2) (CIRES, 2014; SWFWMD, 2021a). The topography and bathymetry data within the drainage basin defined as Tampa Bay have elevations that range from −64.2 to 35.2 ft (NAVD 88), with areas above sea level consisting of Tampa Port Authority spoil islands 2D and 3D within Hillsborough Bay (fig. 2). The spoil islands serve as nesting areas for rare species of birds. Bathymetry values near the mouth of the bay include depths greater than 90 ft below NAVD 88. Locations of shipping channels are clearly indicated by linear features in the bathymetric map of the bay. Currently, the deepest channels are dredged to a depth of 43 ft below the bottom of the bay, corresponding to depths of approximately 50 to 60 ft below NAVD 88.

The area surrounding Tampa Bay is largely urbanized and includes the cities of Tampa, Saint Petersburg, Clearwater, and Bradenton, with Macdill Air Force Base situated on a peninsula that extends into the center of Tampa Bay (fig. 1). The four counties that surround the bay are Hillsborough, Pinellas, Pasco, and Manatee Counties. The combined population of these four counties has increased from 1.52 million in 1975 to 3.38 million in 2020—an increase of 122.4 percent (fig. 3) (Leach, 1978; U.S. Census Bureau, 1982, 2020, 2022). This population growth greatly outpaced that of the entire United States, whose population increased by 53.8 percent during the same period. Commercial, industrial, and recreational activities associated with the bay are a critical component of the regional economy, with commercial shipping also being an important component of the local economy.

Previous Studies

A previous study of the hydrology of Tampa Bay was conducted by the USGS (Hutchinson, 1983) and included the development and testing of a groundwater model encompassing the Hillsborough Bay area in northeastern Tampa Bay (fig. 1) to test the possible effects of planned modifications to the channel network. The 97-square-mile domain of the groundwater model was oriented to align the vertical grid along the main shipping channel that was to be deepened under a proposed dredging project, which also included modifications to the Alafia River Channel (fig. 2). Five different scenarios were simulated: (1) existing channels with pumping; (2) dredged main shipping channel only with pumping; (3) dredged main shipping channel and Alafia River Channel with pumping; (4) existing channels without pumping; and (5) dredge main shipping channel and Alafia River Channel without pumping. The greatest hydrologic effects were expected to occur near a large industrial pumping center east of the main shipping channel along the Alafia River Channel where the dredging was likely to breach the upper confining bed. Historical potentiometric contours indicate a cone of depression once existed near this pumping center. The underlying Upper Floridan aquifer (UFA) already contained saltwater in the area but was believed to not be directly exposed to the bay prior to channel modifications.

The simulations by Hutchinson (1983) indicated that channel deepening would likely increase downward leakage into the FAS near the Alafia River Channel. However, the change in downward leakage with or without the channel modifications was small compared to the change in downward leakage with and without industrial pumping. Generally, when pumping and channel modifications were considered, the increased connectivity between the bay and the FAS would reduce the cone of depression near the pumping center and increase the potentiometric surface in the area by approximately 10 ft. The increased saltwater leakage through the channel cut would likely be drawn into the pumping center, reducing the stress on nearby freshwater resources formerly drawn to the well field. Additionally, the modifications to the main shipping channel would likely result in the potentiometric surface elevation equilibrating to bay water levels within Hillsborough Bay, resulting in slight increases in some areas and decreases in others. Dredging the Alafia River Channel had a much greater effect on the leakage and potentiometric surface than dredging the main shipping channel. These results not only indicated the potential effects of additional dredging within the bay, but also illustrated the affect that groundwater withdrawals had on regional groundwater flow.

Additional studies conducted by HydroGeoLogic, Inc., initially involved developing a sharp-interface model, then subsequently developing a density-dependent groundwater flow and solute transport model to simulate the position and movement of saline water in the UFA within the Southern Water Use Caution Area (HydroGeoLogic, Inc., 1994, 2002). The focus of these efforts was to analyze the effects of water extraction, rather than channel modifications, along the eastern coast of Tampa Bay and in areas south of the bay. Using the density-dependent groundwater flow and transport model, pumping scenarios of 400, 600, 800, and 1,000 million gallons per day (Mgal/d) were simulated with the Southern Water Use Caution Area, compared to a range of 562–832 Mgal/d reported within the same area during 1989–98. A total of 1,806 wells were included in the model domain. The model predicted 8.5 percent of them were affected by chloride concentrations exceeding 500 milligrams per liter (mg/L) under the then “current” conditions for the year 2000. After 50 years of water extraction, 12.4 percent could be affected by chloride concentrations exceeding 500 mg/L with pumping rates of 1,000 Mgal/d. In the highly permeable zone of the Avon Park Formation, the predicted inland movement of the saltwater front ranged from approximately 2.5 to 9 mi for the 400 and 1,000 Mgal/d pumping scenarios, respectively, after 50 years of simulation.

Collocated Unit Terminology Used in This Report

To help provide context to discussions in this report, collocated hydrogeologic and geologic units are mentioned where applicable. For brevity, the shortened designations shown in table 1 for these collocated units are used because most are referred to frequently in the text and because the full names of some hydrogeologic and geologic units are lengthy.

Groundwater Withdrawals

The total groundwater withdrawals from all aquifers, which include withdrawals for industrial, agricultural, and public supply, averaged 408 Mgal/d for the four counties surrounding Tampa Bay for 1975–2015 (fig. 5, table 2) (Leach, 1978, 1983; Leach and Healy, 1980; Marella, 1988, 1992, 1999, 2004, 2009, 2014, 2020). The peak withdrawal for this period occurred in 2000 and was 502 Mgal/d, which decreased to 308 Mgal/d by 2015. For these four counties, 294 Mgal/d of the total 308 Mgal/d of groundwater withdrawals for 2015, which corresponds to 95 percent, were withdrawn from the FAS (Marella, 2020). The reduction in total withdrawals from the peak in 2000 to 2015 can be attributed to several factors, which include decreased withdrawals for agricultural and public supply.

Hutchinson (1983) reported that in 1977, groundwater withdrawals from the FAS averaged approximately 330 Mgal/d for the Tampa Bay region. Marella (2020; table 2) reported groundwater withdrawals of 389 Mgal/d. The discrepancy between estimates results from the exclusion of Pasco County by Hutchinson (1983) from the Tampa Bay region. An estimated 271 Mgal/d of the 330 Mgal/d was considered freshwater and 59 Mgal/d was considered saline water, of which 55 Mgal/d was withdrawn from wells at a phosphate processing plant at the mouth of the Alafia River. Leach and Healy (1980) provided a similar estimate of industrial self-supplied saline groundwater for Hillsborough County of 57 Mgal/d in 1977 (table 3).

A cone of depression in the potentiometric surface of an aquifer can result from water withdrawals from wells. Changes in the potentiometric surface over time are discussed later in the “Historical Potentiometric Surface Data” section of this report. A cone of depression near the mouth of the Alafia River in Hillsborough County was present on the potentiometric surface map for May 1978 and was attributed by Hutchinson (1983) to these industrial withdrawals of saline water from the FAS at a depth of approximately 1,000 ft, which would likely be within the UFA/Avon Park Formation. Between the 1980 and 1985 water use estimates for the State of Florida, the industrial supply of saline water from the FAS for Hillsborough County decreased to zero, where it remained through 2015 (table 3) (Leach, 1978, 1983; Leach and Healy, 1980; Marella, 1988, 1992, 1999, 2004, 2009, 2014, 2020). These data indicate the phosphate processing plant at the mouth of the Alafia River no longer relied on self-supplied saline groundwater withdrawals from the FAS after 1985 and perhaps as early as 1981.

Hutchinson (1983) noted two additional cones of depression in the Tampa Bay region in the May 1978 potentiometric surface map for the FAS. One depression in Pasco County was attributed to municipal withdrawals, and a second depression in southern Hillsborough and northern Manatee Counties was attributed to agricultural withdrawals. The groundwater withdrawals for public supply within Hillsborough, Pinellas, Pasco, and Manatee Counties averaged 168 Mgal/d for the 1975–2015 period and peaked at 242 Mgal/d in 2000 (fig. 6) (Leach, 1978, 1983; Leach and Healy, 1980; Marella, 1988, 1992, 1999, 2004, 2009, 2014, 2020). In 1975, 88 percent of the groundwater withdrawals for public supply occurred in Pinellas County. By 1985, public-supply groundwater withdrawals occurring in Hillsborough and Pasco Counties exceeded those in Pinellas and Manatee Counties, and this remained the case through 2015 (fig. 6).

Agricultural water withdrawals can vary greatly from year to year and depend on climatic conditions. Such conditions reduce the recharge to aquifers and often lead to lower regional water-table levels and potentiometric surfaces, while simultaneously increasing demand for agricultural withdrawals. These factors work to increase the magnitude and spatial extent of drawdowns during drier years, which can then lead to further inland movement of the saltwater front. The agricultural groundwater withdrawals for 1975–2015 averaged 161.6 Mgal/d and reached a peak of 229.6 Mgal/d in 1985 for the four counties that surround the bay, with most of the withdrawals occurring in Hillsborough and Manatee Counties (fig. 7A) (Leach, 1978, 1983; Leach and Healy, 1980; Marella, 1988, 1992, 1999, 2004, 2009, 2014, 2020). Annual rainfall for the four counties during these years ranged from 31.3 to 66.2 inches (in.) (fig. 7B) (SWFWMD, 2022).

Overall, in the Tampa Bay region, total groundwater withdrawals and public-supply withdrawals of groundwater peaked around 2000, and agricultural withdrawals peaked in the mid-1980s (Leach, 1978, 1983; Leach and Healy, 1980; Marella, 1988, 1992, 1999, 2004, 2009, 2014, 2020). Additionally, the industrial saline groundwater withdrawals at the mouth of the Alafia River ceased between 1980 and 1985. These withdrawals placed stress on the aquifer system, which is evident on the potentiometric surface maps presented and discussed later in the “Historical Potentiometric Surface Data” section.

Sea-Level Rise

SLR is expected to increase groundwater elevations in coastal aquifers with the effects extending some distance inland (Bjerklie and others, 2012; Rotzoll and Fletcher, 2013; Decker and others, 2019; Decker, 2022). The distance inland over which SLR affects aquifers depends on many factors, including aquifer characteristics, surface-water controls, groundwater recharge, and groundwater extraction. In coastal areas, increased sea-level elevation can decrease gradients in the potentiometric surface, causing inland migration of the saltwater front. Intrusion of saline water into areas of groundwater extraction can necessitate mitigation strategies such as decreasing well production or possibly abandoning or relocating well fields.

Historical

Although not in the area of Tampa Bay, the Key West, Florida tidal gage (National Oceanic and Atmospheric Administration, [NOAA], 8724580) (figs. 1 [inset map] and 8) in southernmost Florida has one of the longest continuous periods of record within the State and is often used to analyze long-term trends in mean sea level (NOAA, 2023a). The trend in monthly mean tidal level can vary based on the period of record considered. The linear trend in monthly mean tidal level at Key West, Fla., when using all 110 years of available data from 1913 to 2022 is an increase in sea level of 0.101 inch per year (in/yr) (fig. 8). The linear trend over the 37-year period from 1986 to 2022 is an increase of 0.175 in/yr, indicating an increased rate of SLR in recent decades. Similar trends have been measured at the St. Petersburg, Fla. tidal gage (NOAA 8726520) (figs. 1, 2, and 9) (NOAA, 2023b). The linear trend in monthly mean tidal level was an increase of 0.120 in/yr for the 76-year period 1947–2022 and an increase of 0.181 in/yr for 1986–2022, which indicates approximately 6.7 in. or 0.56 ft of SLR over the last 37 years (fig. 9). The mean tidal level at the St. Petersburg, Fla. tidal gage during 1986–2022 was −0.13 ft (NAVD 88). The increased elevation of mean sea level within the bay likely caused a similar increase in hydraulic heads in the underlying aquifer and for some distance inland. This likely caused the saltwater/freshwater interface to move farther inland, but the amount of movement has not been quantified.

Projected

Projections of sea-level rise in the future are used during planning processes and analyses of alterations to the hydrologic system within a coastal environment (Sweet and others, 2017; Decker, 2022). The NOAA projections of global mean sea-level rise include six projection scenarios that range from 0.98 ft to 8.20 ft of SLR by 2100 (Sweet and others, 2017). These scenarios can be adjusted by incorporating nearby tidal gages to create regional projections of future sea level. The likelihood of exceeding these SLR projections has been quantified under different representative concentration pathways (RCPs) that correspond to possible future greenhouse gas concentrations and radiative forcing through 2100: RCP 2.6, RCP 4.5, and RCP 8.5 (table 4). The RCP 2.6 scenario is considered unlikely and would require strong mitigation activities with net-negative greenhouse gas emissions toward the end of the 21st century (Sweet and others, 2017). The RCP 4.5 scenario is an intermediate and most likely scenario in which emissions are stabilized by 2050 then begin to decline in the latter half of the 21st century. The RCP 8.5 scenario is generally taken as the worst-case scenario in which emissions continue to rise throughout the 21st century. The range and likelihood of these projections is intended to be incorporated into a risk-based approach to coastal planning and management.

Table 4.    

Probability of exceeding global mean sea-level rise scenarios under different representative concentration pathways (Sweet and others, 2017).

[NOAA, National Oceanic and Atmospheric Administration; GMSL, global mean sea level; ft, foot; SLR, sea-level rise; RCP, representative concentration pathway; %, percent]

Table 4.    Probability of exceeding global mean sea-level rise scenarios under different representative concentration pathways (Sweet and others, 2017).

NOAA GMSL rise scenario	2000–2100 SLR (ft)	Probability of exceedance by 2100
		RCP 2.6	RCP 4.5	RCP 8.5
Low	0.98	94%	98%	100%
Intermediate-low	1.64	49%	73%	96%
Intermediate	3.28	2%	3%	17%
Intermediate-high	4.92	0.4%	0.5%	1.3%
High	6.56	0.1%	0.1%	0.3%
Extreme	8.20	0.05%	0.05%	0.1%

The Tampa Bay Climate Science Advisory Panel (TBCSAP) recommends using the NOAA intermediate-low, intermediate, and high SLR projection curves for planning projects within the Tampa Bay region (TBCSAP, 2019). These three SLR curves are a subset of the six SLR curves presented by NOAA and exclude the low, intermediate-high, and extreme scenarios (table 4). When the global mean sea-level rise projections were adjusted regionally using data from the St. Petersburg, Fla. tidal gage (NOAA 8726520) and the USACE Sea Level Change Curve Calculator Tool (USACE, 2022), the intermediate-low, intermediate, and high NOAA SLR projections result in 1.9, 3.9, and 8.5 ft of SLR between 2000 and 2100, respectively (table 5) (TBCSAP, 2019).

These regional projections can be reproduced using the Sea Level Analysis Tool (USACE, 2023), which superseded the USACE Sea Level Change Curve Calculator Tool. The 5-year mean sea level centered on the year 2000 was calculated to be −0.25 ft (NAVD 88) based on St. Petersburg Fla. tidal gage data for 1998–2002. Adding the SLR projections to mean sea level for 2000 results in projected sea levels for 2100 of 1.65, 3.65, and 8.25 ft (NAVD 88) for the intermediate-low, intermediate, and high SLR scenarios, respectively (table 5, fig. 10). These scenarios for projected sea levels can be compared to a calculated sea level of 1.26 ft (NAVD 88) for 2100, obtained by adding the 0.181 in/yr averaged linear trend observed for 1986–2022 at the St. Petersburg Fla. tidal gage (NOAA 8726520) to the mean sea level for 2000. Based on the published data in table 4, SLR will likely exceed the intermediate-low SLR scenario but will likely not exceed the intermediate scenario under the most likely RCP 4.5 scenario. The 5-year mean sea level at the St. Petersburg Fla. tidal gage has trended near or slightly above the NOAA intermediate-low SLR projections during 2000–20, with an increasing rate toward the end of the available record (fig. 11). The 5-year mean sea level at the St. Petersburg Fla. tidal gage increased 0.43 ft between 2000 and 2020 compared to projected values of 0.37, 0.53, and 0.83 ft interpolated from the intermediate-low, intermediate, and high NOAA SLR projections, respectively. If sea level continues to trend toward the intermediate-low and intermediate SLR scenarios, then the likely SLR for the Tampa Bay region, referenced to the year 2000, will be between 0.95 and 1.44 ft for 2050 and between 1.90 and 3.90 ft for 2100. The resulting expected mean sea level would be between 0.7 and 1.19 ft (NAVD 88) for 2050 and between 1.10 and 2.08 ft for 2100 (NAVD 88) (table 5).

Table 5.    

Projected regional sea-level rise and regional mean sea level for the Tampa Bay region for National Oceanic and Atmospheric Administration scenarios (Tampa Bay Climate Science Advisory Panel [TBCSAP], 2019).

[NOAA, National Oceanic and Atmospheric Administration; ft, foot; NAVD 88, North American Vertical Datum of 1988; int-low, intermediate-low]

Table 5.    Projected regional sea-level rise and regional mean sea level for the Tampa Bay region for National Oceanic and Atmospheric Administration scenarios (Tampa Bay Climate Science Advisory Panel [TBCSAP], 2019).

Year	Sea-level rise based on NOAA scenario (ft)			Regional mean sea-level based on NOAA scenario (ft relative to NAVD 88)
	Int-low	Intermediate	High	Int-low	Intermediate	High
2000	0.00	0.00	0.00	−0.25	−0.25	−0.25
2030	0.56	0.79	1.25	0.31	0.54	1.00
2040	0.72	1.08	1.77	0.47	0.83	1.52
2050	0.95	1.44	2.56	0.70	1.19	2.31
2060	1.15	1.87	3.48	0.90	1.62	3.23
2070	1.35	2.33	4.56	1.10	2.08	4.31
2080	1.54	2.82	5.71	1.29	2.57	5.46
2090	1.71	3.38	7.05	1.46	3.13	6.80
2100	1.90	3.90	8.50	1.65	3.65	8.25

Rainfall and Evapotranspiration

Rainfall recharges groundwater in an aquifer, whereas evapotranspiration returns rainfall or surface water back to the atmosphere (Freeze and Cherry, 1979; Davie, 2008). Rainfall that does not recharge groundwater or return to the atmosphere can become surface runoff to rivers or directly to the bay. Aquifers can also discharge water to the land surface or sea floor given the right hydrologic conditions, such as at springs. In the Tampa Bay region, the FAS is recharged in three primary ways: inland areas are recharged by downward leakage of rainfall through the overlying SAS and confining unit or directly where the FAS is exposed at land surface; areas under the bay are recharged by leakage from the bay that occurs through surficial sediments or directly where the FAS is exposed at the sea floor; and the FAS can receive water through interflow from adjacent aquifers (Hutchinson, 1983).

The amount of precipitation minus evapotranspiration (ET) occurring over an area represents the net water flux available for recharge to an aquifer, with increases in rainfall and decreases in ET likely leading to increased recharge (Freeze and Cherry, 1979; Davie, 2008). Conversely, decreases in rainfall and increases in ET would likely lead to decreased recharge to an aquifer. The elevation of a potentiometric surface is controlled by recharge to an aquifer as well as other factors such as changes in storage and discharge (Taylor and Alley, 2001). Years for which precipitation minus ET totals are higher than average will likely also have higher potentiometric levels in inland areas where the FAS is exposed and the aquifer is recharging and, as a result, increased potentiometric gradients toward the coast. The increased gradient can decrease or stop the inland movement of the saltwater front or even lead to coastward movement of the front. Seasonal variations in precipitation minus ET will also alter the potentiometric surface, with elevations typically being higher during the wet season than during the dry season.

From 1915 to 2022, the mean annual rainfall totals for the four counties surrounding Tampa Bay were 52.7, 51.7, 53.4, and 54.0 in. for Hillsborough, Pinellas, Manatee, and Pasco Counties, respectively (fig. 12, table 6) (SWFWMD, 2022). The minimum annual rainfall totals over the same period were 34.1, 33.8, 34.4, and 31.3 in. for these counties, respectively, and the maximum annual rainfall totals over the same period were 80.7, 85.2, 84.8, and 75.7 in., respectively. The mean annual rainfall totals averaged for all four counties for 1915–2022 was 53.0 in., with a standard deviation of 7.8 in. The minimum and maximum annual rainfall for the same period were 34.1 and 81.6 in., respectively (fig. 13, table 6).

Prolonged periods of below-normal rainfall could lower inland groundwater levels altering the potentiometric gradient toward the bay (Taylor and Alley, 2001). Reduced elevations within the potentiometric surface could increase the intrusion of saltwater from Tampa Bay into aquifers surrounding the bay (Hutchinson, 1983). In areas underlying the bay, a potentiometric surface with higher mean values relative to mean tidal levels will generally result in upward groundwater discharge from the FAS rather than downward leakage of saline water into the FAS. Net leakage of saline water from the bay into the FAS will likely occur when the mean potentiometric surface falls below mean tidal levels (Hutchinson, 1983). Cones of depression in the potentiometric surface caused by groundwater extraction typically enlarge during drier years and shrink or disappear during wetter years. ET tends to exhibit less year-to-year variation, but years of increased actual ET can produce similar effects by reducing the net recharge to the aquifer systems and lowering the potentiometric surface.

For the period 1986–2022 (37 years), which corresponds to the years following the last major channel modifications, the mean of the yearly rainfall for the four counties surrounding Tampa Bay was 51.9 in., with a standard deviation of 6.8 in. For the 37 years (1949–85) preceding this period, the mean of the yearly rainfall was 54.2 in., which is 4.3 percent greater than the subsequent period, with a standard deviation of 6.8 in. (table 6). The minimum and maximum of the mean yearly rainfall for 1986–2022 were also lower than the rainfall for 1949–85: 34.1 in. compared to 38.4 in., and 66.5 in. compared to 81.6 in. Based on the entire measurement period, the number of years considered “drier than average,” which fall below the 25th percentile (47.8 in.), and those are considered a “dry” year, which fall below the 10th percentile (43.9 in.), were both higher for 1986–2022 than for 1949–85: 9 “drier than average” years during 1986–2022 compared to 8 “drier than average” years during 1949–85, and 4 “dry” years during 1986–2022 compared to 3 “dry” years during 1949–85. The number of years considered “wetter than average”, which exceed the 75th percentile (56.6 in.), and those considered a “wet” year , which exceed the 90th percentile (63.2 in.), were lower during 1986–2022 compared to 1949–85: 8 “wetter than average” years during 1986–2022 compared to 10 “wetter than average” years during 1949–85, and 3 “wet” years during 1986–2022 compared to 7 “wet” years during 1949–85. The increased number of “drier than average” years and an overall reduction in rainfall after 1986 might be contributing factors in saltwater intrusion in the coastal areas surrounding the bay.

Estimates of actual ET are available for the conterminous United States using an operational Simplified Surface Energy Balance (SSEBop) model (Senay and others, 2013) for 2000–18. These are estimates of the net depth of water removed from the system and are a fraction of the calculated potential ET, which represents the atmospheric demand if sufficient water were available. For 2000–18, the average annual actual ET totals for the four counties surrounding the bay were 36.6, 37.3, 38.2, and 30.4 in/yr for Hillsborough, Pinellas, Manatee, and Pasco Counties, respectively (table 7). Actual ET has less yearly variability when compared to other atmospheric forcings like rainfall (fig. 14). The mean of the estimated precipitation minus ET for the four counties, calculated by subtracting actual ET from the rainfall totals, ranged from 1.0 to 29.2 in. for 2000–18 (fig. 14, table 7). A value of 0.0 was assigned for a given county’s effective yearly rainfall total when estimated actual ET was larger than that county’s yearly rainfall total, which occurred only in 2000 for Hillsborough County.

Summary of Historical Data From Groundwater and Chloride Monitoring Wells

Twenty hydraulic head datasets and two chloride concentration datasets were evaluated for wells with an SAS hydrogeologic unit designation (tables 9 and 10, fig. 23). No significant regional trends in hydraulic head or chloride concentration were indicated by Kendall’s τ and p-values within these datasets that could be related to the stressors of the aquifer system being considered.

All being part of the Hawthorn Group, the 21 hydraulic head datasets having hydrogeologic/geologic designations of UCU/Hawthorn Group, UFA/upper Arcadia Formation, UFA/lower Arcadia Formation, and UFA/Tampa Member were combined for the analysis (SWFWMD, 2018b; USGS, 2016). An increasing trend in hydraulic head was indicated using Kendall’s τ and p-values along the eastern and northern coast of Tampa Bay, and a likely decreasing trend was indicated using Kendall’s τ and p-values along the coast in the areas south of Tampa Bay (table 9, fig. 24). Eighteen chloride concentration datasets were evaluated for wells having hydrogeologic/geologic designations of UCU/Hawthorn Group, UFA/upper Arcadia Formation, UFA/lower Arcadia Formation, and UFA/Tampa Member. The calculated Kendall’s τ and p-values of data from 14 wells indicate decreasing or no significant trends in chloride concentration along the eastern and northern coast of the bay as well as along the coast south of the bay, although chloride concentrations are relatively low at these depths (table 10, fig. 25). Changes seen within the potentiometric surface of the UFA, likely due to changes in groundwater extraction, during the study period are one potential cause for these regional trends in hydraulic head and chloride concentration (figs. 16–21).

Nineteen hydraulic head datasets were evaluated for wells with a hydrogeologic/geologic unit designation of UFA/Suwannee Limestone. Using Kendall’s τ and p-values, an increasing trend in hydraulic head was indicated along the eastern coast of Tampa Bay and a decreasing trend in hydraulic head was indicated along the northern coast of the bay (table 9, fig. 26). Twelve chloride concentration datasets were evaluated for wells with a hydrogeologic/geologic unit designation of UFA/Suwannee Limestone. Increasing trends in chloride concentration were indicated by Kendall’s τ and p-values along the coast in the northern part of the bay and south of the bay (table 10, fig. 27). Limited chloride concentration data were available along the eastern coast of Tampa Bay with a hydrogeologic/geologic unit designation of UFA/Suwannee Limestone.

Six hydraulic head datasets were evaluated for wells having a hydrogeologic/geologic unit designation of UFA/Ocala Limestone. Because only a few datasets were available, regional trends in hydraulic head could not be determined (table 9, fig. 28). Only three chloride concentration datasets were available for wells having a hydrogeologic/geologic unit designation of UFA/Ocala Limestone. Using Kendall’s τ and p-values, all three datasets indicate an increasing trend, and all locations are along the eastern coast of Tampa Bay; however, with so few datasets, no significant regional trends could be determined (table 10).

Fourteen hydraulic head datasets and 15 chloride concentration datasets were evaluated for wells with a hydrogeologic/geologic unit designation of UFA/Avon Park Formation. Kendall’s τ and p-values indicate increasing trends in both hydraulic head and chloride concentration along the eastern coast of Tampa Bay and south of the bay (tables 9 and 10, figs. 29 and 30).

Three wells north of Tampa Bay had available data before and after the last major channel modifications completed in the mid-1980s: TR 15-1, TR 15-2, and TR 15-3 (tables 9 and 10, fig. 22) (SWFWMD, 2018b; USGS, 2016). Each well had hydraulic heads monitored at a single depth beginning in 1978 with hydrogeologic/geologic unit designations of UFA/Suwannee Limestone, UFA/Tampa Member, and UFA/Suwannee Limestone for wells TR 15-1, TR 15-2, and TR 15-3, respectively. Kendall’s τ and p-values for measured hydraulic heads indicated increasing trends throughout the period of record. Chloride concentrations show no significant trend, an increasing trend, and a decreasing trend for wells TR 15-1, TR 15-2, and TR 15-3, respectively.

Three wells along the northern coast of Tampa Bay had available data before and after the last major channel modifications completed in the mid-1980s: TR 13-3, TR 11-2, and TR 10-2 (tables 9 and 10, fig. 22) (SWFWMD, 2018b; USGS, 2016). Well TR 13-3 had hydraulic heads monitored at two depths with hydrogeologic/geologic unit designations of SAS and UFA/Avon Park Formation beginning in 1977 and 1982, respectively. Kendall’s τ and p-values of hydraulic head data indicate an increasing trend in the SAS and a decreasing trend in the UFA/Avon Park Formation. Kendall’s τ and p-values of hydraulic head data at well TR 11-2 monitored at a single depth in the UFA/Suwannee Limestone beginning in 1980 indicate an increasing trend. Well TR 10-2 has hydraulic heads and chloride concentrations monitored within the Arcadia Formation beginning in 1980. Kendall’s τ and p-values for both hydraulic head and chloride concentration datasets for this well indicate increasing trends.

Four wells located along the eastern coast of Tampa Bay had available data before and after the last major channel modifications completed in the mid-1980s: TR 9-1, TR 9-3, 50, and 51 (tables 9 and 10, fig. 22) (SWFWMD, 2018b; USGS, 2016). Well TR 9-1 had hydraulic heads and chloride concentrations monitored at a single depth within the lower part of the Arcadia Formation beginning in 1976 and 1978, respectively. Kendall’s τ and p-values for hydraulic head data indicated an increasing trend, whereas Kendall’s τ and p-values for chloride concentration data indicated a decreasing trend. Well TR 9-3 had hydraulic heads and chloride concentrations measured within the SAS beginning in 1976 and 1978, respectively. Kendall’s τ and p-values for the hydraulic head and chloride concentration datasets indicated an increasing trend and no significant trend, respectively. Well 50 had hydraulic heads and chloride concentrations monitored within the SAS beginning in 1976. Kendall’s τ and p-values for the hydraulic head and chloride concentration datasets within the SAS monitored depth indicated an increasing trend and a decreasing trend, respectively. Well 51 had chloride concentrations monitored within the UFA/Avon Park Formation beginning in 1976, with Kendall’s τ and p-values indicating an increasing trend.

Two wells along the southern coast of Tampa Bay had available data before and after the last major channel modifications completed in the mid-1980s: TR 7-1 and TR 6-1 (tables 9 and 10, fig. 22) (SWFWMD, 2018b; USGS, 2016). Well TR 7-1 had hydraulic heads and chloride concentrations measured at a single depth within the lower part of the Arcadia Formation beginning in 1983 and 1981, respectively, with Kendall’s τ and p-values for both datasets indicating no significant trends. Well TR 6-1 had hydraulic heads and chloride concentrations measured at a single depth within the lower part of the Arcadia Formation beginning in 1979. Kendall’s τ and p-values for the hydraulic head and chloride concentration datasets indicated a decreasing trend and no significant trend, respectively.

Groundwater Monitoring Wells North of Tampa Bay

The trends and changes in hydraulic head and chloride concentration observed at wells north of Tampa Bay (fig. 22) cannot be attributed to channelization within the bay because these locations are too far removed from the bay. Data from these wells are presented to illustrate changes in potentiometric levels within the region that are likely the result of alterations in groundwater withdrawals, aquifer recharge, SLR, and interaction with surface-water features such as Lake Tarpon (fig. 22). Table 11 lists the ROMP wells whose data were analyzed for trends in hydraulic head and chlorine concentration.

ROMP wells TR 16-2, TR 16-3, and TR 16-4 are in the northern part of the study area, approximately 15 mi north of northernmost extent of Tampa Bay, and are oriented west to east along a transect that extends inland from the gulf coast (fig. 22). Well TR 16-2 is located approximately 2 mi east of the gulf coast and has hydraulic heads monitored within the UFA/Tampa Member (68-ft depth), UFA/Suwannee Limestone (210-ft depth), UFA/Ocala Limestone (350-ft depth), and the UFA/Avon Park Formation (455-ft depth) (table 11). Hydraulic head data from the ROMP TR 16-2 well for 1992–2022 are shown in figure 31. Kendall’s τ and p-values for these data indicate increasing trends for this period (table 12), with mean heads of 5.92, 5.84, 0.60, and −0.12 ft (NAVD 88) for the UFA/Tampa Member (68-ft depth), UFA/Suwannee Limestone (210-ft depth), UFA/Ocala Limestone (350-ft depth), and UFA/Avon Park Formation (455-ft depth) wells, respectively. The measured hydraulic heads within the shallower locations (68- and 210-ft depths) have mean values approximately 5 ft higher than hydraulic heads within the deeper locations (350- and 455-ft depths), suggesting these aquifer units are not well-connected and indicating an upward hydraulic head gradient. There likely exists a confining unit between the two shallower formations, UFA/Tampa Member and UFA/Suwannee Limestone, and the two deeper formations, UFA/Ocala Limestone and UFA/Avon Park Formation, resulting in the differences in measured hydraulic heads. The mean chloride concentration of 330 mg/L within the UFA/Tampa Member (68-ft depth) indicates freshwater within the UCU, and the mean chloride concentration of 1,242 mg/L within the UFA/Suwannee Limestone (210-ft depth) indicates brackish water within the upper portions of the UFA. There is a significant increasing trend in chloride concentration within the Tampa Formation (68-ft depth) and a decreasing trend in chloride concentration within the UFA/Suwannee Limestone (210-ft depth), indicated by the Kendall’s τ and p-values (table 13); however, the magnitude of change in chloride concentration within the UFA/Tampa Member (68-ft depth) is much smaller than within the UFA/Suwannee Limestone (210-ft depth). As noted, hydraulic heads at each depth have increased during this same period.

ROMP wells TR 16-3 and TR 16-4 each have hydraulic heads measured at two depths and are located approximately 3.5 and 4.5 mi from the gulf coast, respectively (fig. 22). In order of increasing depth, the wells within TR 16-3 are finished within the SAS (5-ft depth) and UFA/interface with Avon Park (494-ft depth) (table 11). The vertical gradient typically was downward during 2002–22, with measured hydraulic heads within the SAS higher than those within the UFA/Avon Park Formation (fig. 32) and mean values of 14.80 and 13.34 ft (NAVD 88), respectively, averaged for 2002–22 (table 12). Data records indicate the downward gradient was reversed during several measurement periods when heads within the UFA/Avon Park Formation were higher than land surface elevation (17.7 ft NAVD 88) (SWFWMD, 2018b; USGS, 2016). These periods correspond to large rainfall events associated with tropical disturbances, which include Hurricane Frances (September 5, 2004), Hurricane Jeanine (September 26, 2004), an unnamed tropical disturbance (August 3, 2015), and Hurricane Hermine (September 2, 2016) (Kane, 2005; Lawrence and Cobb, 2005; Marrero and others, 2016).

The 4- and 740-ft-depth locations within well TR 16-4 are finished within the SAS and UFA/Avon Park Formation, respectively (table 11). There is typically a downward gradient in hydraulic head (fig. 33), with mean values of 17.66 and 15.92 ft (NAVD 88) averaged for 2002–22 within the SAS and UFA/Avon Park Formation, respectively (table 12). The similarity in mean values and trends in the data from TR 16-4 suggests that these aquifer units are well-connected.

ROMP wells TR 16-2, TR 16-3, and TR 16-4 have a pattern of decreasing hydraulic head toward the gulf shore, indicating groundwater flow toward the coast. ROMP wells TR 16-3 and TR 16-4 are farther inland than TR 16-2 (fig. 22) and have lower chloride concentrations, with mean values of 349 and 45 mg/L within the UFA/Avon Park Formation for TR 16-3 (494-ft depth) and TR 16-4 (740-ft depth), respectively, which indicate fresh groundwater (figs. 32 and 33, table 13). Kendall τ and p-values indicate a significant upward trend in chloride concentration over time within the UFA/Avon Park Formation at TR 16-3 (494-ft depth), and no significant trend over time occurs within the UFA/Avon Park Formation at TR 16-4 (740-ft depth).

ROMP wells TR 15-1, TR 15-2, and TR 15-3 are located farther south and closer to the Gulf of Mexico than TR 16-2, TR 16-3, and TR 16-4. The first three wells are oriented west to east along a transect at approximate distances of 0.1, 1.5 and 2.5 mi from the gulf coast, respectively, with TR 15-3 near Lake Tarpon (fig. 22, table 11). Each well has a single monitored depth, with hydrogeologic/geologic unit designations of UFA/Suwannee Limestone, UFA/Tampa Member, and UFA/Suwannee Limestone for TR 15-1 (80-ft depth), TR 15-2 (50-ft depth), and TR 15-3 (137-ft depth), respectively. The mean hydraulic heads for the period 1978–2022 indicate a downward gradient toward the coast with values of 2.13, 2.79, and 4.54 ft (NAVD 88) for TR 15-1, TR 15-2, and TR 15-3, respectively (table 12). Kendall τ and p-values indicate increasing trends in hydraulic head during this period for these three wells (table 12).

Fresh groundwater is indicated within ROMP wells TR 15-1, TR 15-2, and TR 15-3 at depths of 80, 50, 137 ft, respectively. For 1978–2022, chloride concentration data at TR 15-1 (UFA/Suwannee Limestone, 80-ft depth) have no significant trend and a mean value of 576 mg/L (fig. 34, table 13). Kendall’s τ and p-values for chloride concentration data for TR 15-2 (UFA/Tampa Member, 50-ft depth) indicate an increasing trend over time, with a mean value of 242 mg/L (fig. 35, table 13). The cause of the increasing chloride concentrations at TR 15-2 is not clear; however, groundwater withdrawals in the area could pull higher salinity water upward into the shallower aquifer. The chloride concentrations at TR 15-3 (UFA/Suwannee Limestone, 137-ft depth), the farthest inland well, significantly decreased from approximately 700 to 100 mg/L during 1978–2022, with a mean value of 432 mg/L (fig. 36, table 13). This decrease in chloride concentration was likely associated with the salinity conditions in nearby Lake Tarpon (fig. 2), located 1/8 mi east of TR 15-3. Lake Tarpon sink is located on the northwestern shoreline of Lake Tarpon and is hydrologically connected to the Gulf of Mexico (Dooris and Bartos, 1980). The Lake Tarpon sink previously acted as both an outflow and inflow source depending on tidal and lake conditions, causing high salinity conditions. During the 1960–70s, salinity mitigation activities were undertaken to freshen the lake, which included construction of a berm to enclose the sinkhole and the construction of a canal and water-control structure to discharge waters from the lake southward to Old Tampa Bay. Following these mitigation activities, chloride concentrations within the lake decreased from as high as 5,000 mg/L to values near the 250-mg/L drinking water standard (Dooris and Bartos, 1980). Changes in groundwater chloride concentrations within TR 15-3 illustrate the local hydraulic connection between the Gulf of Mexico and Lake Tarpon by means of groundwater transport through karst features.

In summary, wells north of Tampa Bay are sufficiently removed from the bay for evaluating trends in the region not affected by dredging. Hydraulic heads indicate flow directions westward toward the gulf coast. Wells near the gulf coast and near Lake Tarpon had hydraulic heads approaching those of tidal levels in the bay (having means as low as 2.13 ft) that still remained higher than tidal levels. At Lake Tarpon, the local effect of karst is illustrated by the changes in salinity in the lake associated with the Lake Tarpon sink that hydraulically connected the lake with the Gulf of Mexico, and subsequent mitigation measures that disconnected the lake from the sinkhole. Prior to mitigation measures, salinity was elevated despite the regional gradients in hydraulic head indicating flow from inland areas westward toward the Gulf of Mexico. This illustrates the difficulty in predicting the effect of karst features on local hydrology.

Groundwater Monitoring Wells Along the Northern Coast of Tampa Bay

Data from 11 wells within 5 mi of the northern coast of Tampa Bay were analyzed for trends (fig. 22, table 14) in hydraulic head and chloride concentrations. ROMP wells TR 14-1, TR 14-3, and TR 14-2 are oriented along a transect extending southeast to northwest between Old Tampa Bay and the Gulf of Mexico at distances of 0.5, 2.5, and 4 mi from the bay, respectively (fig. 22). Well TR 14-2 is approximately 2 mi from the Gulf of Mexico. Three depths were monitored at TR 14-1 within the SAS (9-ft depth), UFA/Tampa Member (70-ft depth), and UFA/Suwannee Limestone (268-ft depth), and the data collected at depths indicate similar hydraulic heads and responses, suggesting these aquifer units are well-connected at this location (fig. 37, table 14). The mean measured hydraulic heads at these three depths indicated a downward gradient, with values of 9.21, 8.43, and 8.07 ft (NAVD 88) during 1987–2022 for the SAS (9-ft depth), UFA/Tampa Member (70-ft depth), and UFA/Suwannee Limestone (268-ft depth) designations, respectively (table 15). Given that hydraulic heads are well above the tidal levels, groundwater would likely be discharging from an underlying aquifer into the bay if exposed in this area. There was no long-term monitoring of chloride concentration at TR 14-1.

TR 14-2 and TR 14-3 are located at higher land surface elevations (54- and 94-ft depths) compared to TR 14-1 (16-ft depth) and therefore have higher hydraulic heads within the SAS (figs. 38 and 39, table 14). Mean hydraulic heads for 1987–2022 within TR 14-2 were 48.10, 4.59, and 1.27 ft (NAVD 88) within the SAS (18-ft depth), UFA/Tampa Member (213-ft depth), and UFA/Ocala Limestone (440-ft depth), respectively (fig. 38, table 15). Mean hydraulic heads for 1987–2022 within well TR 14-3 were 90.04, 4.76, and 5.65 ft (NAVD 88) within the SAS (10-ft depth), UFA/Tampa Member (125-ft depth), and UFA/Suwannee Limestone (299-ft depth) (fig. 39, table 15), respectively. These data suggest that the SAS at TR 14-2 and TR14-3 does not appear to be as well-connected to the underlying aquifer units compared to the SAS at TR 14-1, with a confining unit likely present at TR 14-2 and TR 14-3. Well construction logs for wells TR 14-3 and TR 14-2 indicate approximately 70 ft and 43 ft of clay within the UCU/Hawthorn Group between the sand from the SAS and limestone from the UFA/Tampa Member, respectively (Decker, 1984b; Henderson and Decker, 1985). Data suggest the UFA/Tampa Member (213-ft depth) and UFA/Ocala Limestone (440-ft depth) hydrogeologic/geologic unit designations at TR 14-2 are not as well-connected when compared to the UFA/Tampa Member (125-ft depth) and UFA/Suwannee Limestone (299-ft depth) hydrogeologic/geologic unit designations at TR 14-3.

Williams and Kuniansky (2015) described the Suwannee permeable zone as including both the Tampa Member and Suwannee Limestone in west-central Florida, possibly explaining why these formations appear to be well-connected while the Tampa Member and the Ocala Limestone within these wells appear to not be. There were no significant long-term trends in hydraulic head at these three well locations. During the same period, chloride concentrations at TR 14-2 within the UFA/Tampa Member (213-ft depth) decreased from approximately 1,700 to 500 mg/L, while chloride concentrations at TR 14-3 within the UFA/Suwannee Limestone (299-ft depth) increased from approximately 20 to 1,700 mg/L (figs. 38 and 39, table 16). Chloride concentrations at TR 14-2 indicated saline groundwater present within the UFA/Ocala Limestone (440-ft depth), with values between 10,000 and 11,000 mg/L measured during 1988–91 (fig. 38). Chloride concentrations at TR 14-3 within the UFA/Suwannee Limestone (299-ft depth) increased from below 50 mg/L to over 1,700 mg/L during this period, transitioning from fresh to brackish. The TR 14-3 monitoring well is located near public-supply production wells cased in the Tampa Limestone, which together with the increase in chloride concentration, suggests higher salinity water being drawn upwards into the Suwannee Limestone from deeper within the UFA (SWFWMD, 2023).

Monitoring wells TR 13-1X and TR 13-2X are located west of Old Tampa Bay on the Pinellas Peninsula (fig. 22). Well TR 13-1X is located approximately 4 mi from the bay and 3 mi from the gulf coast and has four monitored depths finished within the SAS (10-ft depth), UFA/Tampa Member (173-ft depth), UFA/Suwannee Limestone (254-ft depth), and UFA/Ocala Limestone (520-ft depth) (table 14). The highest hydraulic heads in TR 13-1X are measured within the UFA/Tampa Member (173-ft depth) and UFA/Suwannee Limestone (254-ft depth), having mean hydraulic heads for 1988–2022 of 5.89 and 5.62 ft (NAVD 88), respectively (fig. 40, table 15).

The similar hydraulic heads and responses for TR-13-1X suggest that the UFA/Tampa Member (173-ft depth) and UFA/Suwannee Limestone (254-ft depth) hydrogeologic/geologic unit designations are well-connected. During the same period, hydraulic heads measured within the shallower SAS (10-ft depth) at TR 13-1X have a mean value of 3.24 ft (NAVD 88), indicating the likely presence of a confining unit between the SAS and the UFA/Tampa Member at this location (table 15). Well construction logs indicate the presence of clay from the UCU/Hawthorn Group between depths of 22 and 78 ft, likely resulting in confinement (Clayton, 1985b). The hydraulic heads within the SAS (10-ft depth), UFA/Tampa Member (173-ft depth), and UFA/Suwannee Limestone (254-ft depth) at TR 13-1X are each above mean tidal levels, indicating that there is an upward vertical gradient and that groundwater would likely discharge (rather than recharge) in the northwestern part of the bay. The lowest hydraulic heads at this location were measured within the UFA/Ocala Limestone (520-ft depth), with a mean value of −0.77 ft (NAVD 88) suggesting the UFA/Suwannee Limestone (254-ft depth) and the UFA/Ocala Limestone (520-ft depth) are not well-connected (table 15).

Well TR 13-2X is located 1.5 mi from the bay and had mean hydraulic heads of 13.18, 5.46, 4.04, and −1.62 ft (NAVD 88) within the SAS (10-ft depth), UFA/Tampa Member (146-ft depth), UFA/Suwannee Limestone (269-ft depth), and UFA/Suwannee Limestone (530-ft depth), respectively (fig. 41, table 15). The hydraulic heads within the UFA/Tampa Member (146-ft depth) and UFA/Suwannee Limestone (269-ft depth) are similar during the period 1988–2022, suggesting these aquifers are well-connected, as is the case at TR 13-1X. The lowest measured hydraulic heads at this location were measured within the UFA/Suwannee Limestone (530-ft depth), with a mean value of −1.62 ft. The differences in head between the two UFA/Suwannee Limestone monitored depths of 269 and 530 ft suggests these depth locations within the same aquifer designation are not well-connected. The differences in mean hydraulic heads between the SAS (10-ft depth) and UFA/Tampa Member (146-ft depth) would also suggest these aquifer units are not well-connected with the likely presence of a confining unit. Well construction logs for TR 13-2X indicate the presence of clay from the UCU/Hawthorn Group between depths of 39 and 123.5 ft, likely resulting in confinement (Clayton, 1985a). These results are like those from TR13-1X except for the higher hydraulic heads measured within the SAS at TR 13-2X. The likely cause for the differences in hydraulic heads is the higher land surface elevation at TR 13-2X, 16.6 ft (NAVD 88), compared to TR13-1X, 9.1 ft (NAVD 88) (table 14). In all but the UFA/Suwannee Limestone (530-ft depth), the mean hydraulic heads during the periods of record are greater than the mean tidal levels. Kendall’s τ and p-values indicate significant increasing trends in hydraulic head at TR 13-1X within the SAS (10-ft depth) and UFA/Ocala Limestone (520-ft depth) and at TR 13-2X within the UFA/Tampa Member (146-ft depth) and UFA/Suwannee Limestone (530-ft depth) (table 15). A significant decreasing trend in hydraulic head was indicated at TR 13-2X within the UFA/Suwannee Limestone (269-ft depth). Significant trends in hydraulic head were not indicated at the other monitored depths within either well.

Chloride concentrations at TR 13-1X within the UFA/Tampa Member (173-ft depth) have a mean value of 89 mg/L for 1988–99 and no significant trend (fig. 40, table 16); however, Kendall’s τ and p-values for chloride concentrations within the UFA/Suwannee Limestone (254-ft depth) indicate a significant increasing trend during the period 1988–2022, transitioning from fresh to brackish (greater than 1,000 mg/L), with a mean value of 574 mg/L. A significant increasing trend in chloride concentration at TR 13-2X within the UFA/Suwannee Limestone (269-ft depth) is also indicated by Kendall’s τ and p-values. However, chloride concentrations at this location were higher than those measured at TR 13-1X, with a mean value of 3,462 mg/L (fig. 41, table 16). A search of water use permits in this area did not reveal any nearby locations of groundwater production or withdrawal (SWFWMD, 2023).

Groundwater well TR 13-3 is located approximately 3 mi north of Old Tampa Bay and has measured hydraulic heads within the SAS (10-ft depth) and UFA/Avon Park Formation (724-ft depth) (fig. 22). The mean hydraulic head for 1977–2022 within the SAS (10-ft depth) was 14.46 ft (NAVD 88) and the mean hydraulic head for 1982–2022 within the UFA/Avon Park Formation (724-ft depth) was 14.84 ft (NAVD 88) (fig. 42, table 15). The similarity in hydraulic heads and groundwater response between measurements suggest these aquifers are well-connected across these depths (table 15). Well construction logs do not indicate the presence of clays from the UCU/Hawthorn Group between these two depths, whereas Williams and Kuniansky (2015) indicate that within this geographic area, a well depth of 724 ft would likely be within the Avon Park permeable zone. Mean chloride concentration for 1991–2022 within the UFA/Avon Park Formation (724-ft depth) was 913 mg/L (table 16). Prior to approximately 1997, there was an upward vertical gradient in hydraulic head between the SAS and UFA/Avon Park Formation. After 1997, there is an increase in the frequency of periods of downward vertical gradient, which coincides with increased chloride concentrations within the UFA/Avon Park Formation. Kendall’s τ and p-values indicate an increasing trend in hydraulic head within the SAS (10-ft depth) and decreasing trend within the UFA/Avon Park Formation (724-ft depth), and an increasing trend in chloride concentration within the UFA/Avon Park Formation (724-ft depth), transitioning from fresh to brackish (tables 15 and 16).

Monitoring wells TR 12-1 and TR 12-3 are located northeast of Old Tampa Bay approximately 0.3 and 3 mi from shore, respectively (fig. 22). The mean measured hydraulic head at TR 12-1 was 3.88 ft (NAVD 88) for 1996–2022 within the SAS (22-ft depth) and 4.48 ft (NAVD 88) for 1993–2022 within the UFA/Tampa Member (132-ft depth) (fig. 43, table 15). The similarities in hydraulic heads and responses during the measurement periods suggest the SAS and UFA/Tampa Member are well-connected at this location with a slight upward hydraulic head gradient (table 15). Well construction logs indicate approximately 15.5 ft of UCU/Hawthorn Group present between depths of 24 and 39.5 ft, which could indicate some confinement (SWFWMD, undated a). Monitoring well TR 12-3 has less available hydraulic head data than TR 12-1 because of an incomplete dataset during the period 1993–2022 at a single depth of 294 ft within the UFA/Suwannee Limestone (fig. 44, table 15). For this period, the mean hydraulic head was 11.05 ft (NAVD 88). Kendall’s τ and p-values indicate significant increasing trends in hydraulic head within the SAS (22-ft depth) and UFA/Tampa Member (132-ft depth) at TR 12-1 and within the UFA/Suwannee Limestone (294-ft depth) at TR 12-3 (table 15). Measured chloride concentrations within the UFA/Suwannee Limestone (132-ft depth) at TR 12-1 had a mean value of 548 mg/L for 1992–2022 and a significant increasing trend (fig. 43, table 16). Measured chloride concentrations within the UFA/Suwannee Limestone (294-ft depth) at TR 12-3 had a mean value of 67 mg/L for 1992–2022 and a significant decreasing trend, but with very low chloride concentrations (fig. 44, table 16). Groundwater within both TR 12-1 and TR 12-3 was considered fresh (less than 1,000 mg/L) at these depths during these periods.

Groundwater monitoring wells TR 11-2, TR 10-2, and 62 are located along the northeastern coast of Hillsborough Bay (fig. 22). Well TR 11-2 has a single monitored depth within the UFA/Suwannee Limestone (300-ft depth) and is located near the Tampa Bypass Canal approximately 3 mi inland and downstream from all control structures on the canal. Hydraulic heads within the UFA/Suwannee Limestone (300-ft depth) at TR 11-2 had a mean value of 14.34 ft (NAVD 88) for 1980–2022 and an increasing trend with an approximate 2.5-ft rise by the end of this period (fig. 45, table 15). Chloride concentration data for 1982–2022 within the UFA/Suwannee Limestone (300-ft depth) at TR 11-2 had a mean concentration of 204 mg/L and a significant increasing trend (table 16). Multiple depths were not monitored at this location; however, well construction logs specify 33 ft of UCU/Hawthorn Group between depths of 7 and 40 ft, which could indicate some confinement (SWFWMD, undated b). Additionally, Williams and Kuniansky (2015) suggest that the monitored depth of 300 ft within this area likely lies within the Suwannee permeable zone.

Hydraulic heads within well TR 10-2 were monitored within the SAS (3-ft depth) and UFA/Tampa Member (115-ft depth). Hydraulic heads within the SAS had a mean value of 11.67 ft for 1995–2022 (fig. 46, table 15). Hydraulic heads within the UFA/Tampa Member (115-ft depth) for 1980–2022 had a lower mean value of 9.16 ft (NAVD 88) and exhibited an increase of approximately 10 ft from 1980 to 1996. Hydraulic heads within the SAS (3-ft depth) at well TR 10-2 were approximately 3-4 ft higher compared to those in the UFA/Tampa Member (115-ft depth). After 1996, hydraulic heads demonstrated a steady increasing trend within the UFA/Tampa Member (115-ft depth) and no significant trend within the SAS (3-ft depth). Two possible reasons for the lack of change in the shallower aquifer are that either discharge to the surface could occur as hydraulic heads in the shallow aquifer approach land surface (14.0-ft depth) or the measurement system was not able to accurately capture the hydraulic heads beyond this range. The approximate 3-ft difference in measured hydraulic heads between the SAS and UFA/Tampa Member suggests the two aquifer units are not well-connected between the two monitored depths (table 14). Well construction logs indicate approximately 6 ft of UCU/Hawthorn Group present between depths of 19 and 25 ft and 28.5 ft of low permeability limestone present between depths of 34.5 and 63 ft (SWFWMD, undated c).

Chloride concentration data for 1980–2021 within the UFA/Tampa Member Formation at TR 10-2 had a mean value of 174 mg/L and indicated an abrupt increase of approximately 80 mg/L from 1980 to 1983 (fig. 46, table 16). The timing of this increase coincides with the cessation of industrial pumping of saline water from the UFA near the mouth of the Alafia River that was discussed previously in the “Groundwater Withdrawals” section (table 3). One hypothesis is that the groundwater withdrawals could have pulled freshwater down from the shallower aquifer and once withdrawals ceased, the salinity increased. Chloride concentrations at this depth seem to have stabilized at this increased level once the system was allowed to recover.

Groundwater monitoring well 62 is located south of the Alafia River approximately 5 mi inland from the bay (fig. 22, table 14). Measured hydraulic heads at this well for 2011–22 had mean values of 56.87 and 12.95 ft (NAVD 88) within the SAS (13-ft depth) and UFA/Suwannee Limestone (218-ft depth), respectively (fig. 47, table 15). Measured hydraulic heads for 1995–2022 had a mean value of 11.5 ft (NAVD 88) within the UFA/Avon Park Formation (625-ft depth). The higher hydraulic heads within the shallow aquifer at this well relative to those in nearby wells are likely due to the higher land-surface elevation, 59.4 ft (NAVD 88), at this measurement site relative to other sites along the northern coast of Tampa Bay, and likely presence of a confining unit between the SAS and UFA/Suwanee Limestone designations. Well construction logs in the area indicate approximately 150 ft of UCU/Hawthorn Group confining unit present between the SAS and UFA/Suwannee Limestone (Gates, 2013). The hydraulic heads measured at the deeper depths had very similar values and responses, suggesting the UFA/Suwannee Limestone (218-ft depth) and UFA/Avon Park Formation (625-ft depth) hydrogeologic/geologic unit designations are well-connected at this location. Hydraulic heads demonstrated a significant increasing trend for the period 1995–2021 at all depths. Measured chloride concentrations within the UFA/Avon Park Formation (625-ft depth) had a mean value of 27 mg/L for this period and exhibited a significant increasing trend, but the magnitude of change was slight (fig. 47; table 16).

In summary, wells considered to be located along the northern coast of Tampa Bay are located on the Pinellas Peninsula and along the northern coast of Old Tampa Bay and Hillsborough Bay. On the Pinellas Peninsula (TR 14-3, TR 13-1X, and TR 13-2X), deeper levels of the UFA (>250 ft) had upward trends in chloride concentration over time, but water withdrawal records that could explain this trend were not identified. That does not preclude the fact that water withdrawals occur in this area that affect salinity at the deeper monitoring levels. Hydraulic heads in this area were above tidal levels, indicating discharge to the bay if the UFA were exposed.

Along the northern coast of Old Tampa Bay, Kendall’s τ and p-values for data from well TR 13-3 indicate a decreasing trend in hydraulic head and an increasing trend in chloride concentrations at a depth of 724 ft, suggesting possible increased stresses on the groundwater system such as groundwater extraction. Data collected at wells TR 12-1 and TR 12-3 indicate increasing trends in hydraulic heads at depths of 10, 132, and 294 ft. Along the eastern coast of Hillsborough Bay (TR 11-2, TR 10-2 [115 ft], and 62), a similar pattern of increasing trend in hydraulic head was attributed to the cessation of saltwater withdrawals from the UFA in this region and recovery of cone of depression A (figs. 16 and 17). This area has also been affected by construction of the Tampa Bypass Canal in 1978. As previously discussed, base-flow discharge in the canal after construction was estimated to be approximately twice that of preconstruction discharge, indicating an upward flow of groundwater (discharge). Hydraulic heads inland along the canal system increased by as much as 4 ft because of the impoundment of water within the canal.

Groundwater Monitoring Wells Along the Eastern Coast of Tampa Bay

Hydraulic head and chlorine concentration data from 10 wells within 10 mi of the eastern coast of Tampa Bay were analyzed for trends (fig. 22, table 17). ROMP wells TR 9-1, TR 9-3, TR 9-2, and TR AB-3 are roughly oriented southwest to northeast along a transect extending inland from the bay (fig. 22). Well TR 9-1 is approximately 1 mi from the shoreline and has a single measurement depth within the UFA/lower Arcadia Formation (124-ft depth). Hydraulic heads measured 1976–2022 had a mean value of 11.82 ft (NAVD 88), which is 8.6 ft above land surface elevation, and exhibited a significant increasing trend during this period (fig. 48, table 18). Well construction logs noted that an artesian aquifer was encountered approximately 86 ft below land surface within the UCU/Hawthorn Group and was overlain by approximately 40 ft of a mixture of clay, marl, and limestone (SWFWMD, undated d). The mean measured chloride concentration during 1978–2022 was 26 mg/L, indicating fresh groundwater (table 19). Except for an approximate 4-year period when chloride concentrations approached 50 mg/L, the data indicate steady levels near 25 mg/L with a statistically significant decreasing trend.

ROMP well TR 9-3 is located approximately 2 mi east of the bay and has hydraulic head data for the SAS (20-ft depth), UFA/Suwannee Limestone (289-ft depth), and UFA/Avon Park Formation (764-ft depth) (table 17). Mean hydraulic heads for 1976–2022 were 3.16, −1.28, and 7.21 ft (NAVD 88) for the SAS, UFA/Suwannee Limestone, and UFA/Avon Park Formation, respectively (table 18). The differences in mean hydraulic head suggest these three hydrogeologic/geologic unit designations are likely separated by a confining unit and thus not well-connected. Well construction logs indicate 132 ft of UCU/Hawthorn Group present between depths of 28 and 160 ft and moderately high porosity and permeability at depths of 289 and 764 ft (Decker, 1984a). Kendall’s τ and p-values for the hydraulic head data from the UFA/Suwannee Limestone well (289-ft depth) indicate a significant downward trend after approximately 2005, whereas Kendall’s τ and p-values for the hydraulic head data for the SAS (20-ft depth) and UFA/Avon Park Formation (764-ft depth) indicated increasing trends (table 18). The lower mean hydraulic heads measured within the UFA/Suwannee Limestone indicate groundwater extractions likely are occurring in the area at these depths; however, the potentiometric-surface maps for the UFA in May 2019 do not show a cone of depression near this well (fig. 21). The lack of a depression on the map may be an artifact of the selection of deeper wells within the area to represent the potentiometric surface.

Chloride concentration data from TR 9-3 within the SAS (20-ft depth) and UFA/Suwannee Limestone (289-ft depth) indicate fresh groundwater and Kendall’s τ and p-values indicate no long-term trends, with mean chloride concentrations for 1990–2022 of 53 and 863 mg/L, respectively (fig. 49, table 19). The calculated mean chloride concentration of 863 mg/L within the UFA/Suwannee Limestone (289-ft depth) is likely higher than the central tendency because of 20 values within the dataset that are above 3,000 mg/L (not shown on figure). The higher values are the same magnitude as those of samples collected from the deeper UFA/Avon Park Formation (764-ft depth) at the same time of measurement and well above the values for the samples collected before and after those collected from the UFA/Suwannee Limestone. Removing these values from the calculation of the mean chloride concentration yields a value of 550 mg/L, which is likely more representative of the actual chloride concentration at this depth. Within the UFA/Avon Park Formation (764-ft depth), Kendall’s τ and p-values for chloride concentration data indicate a long-term increasing trend beginning in 1985, resulting in saline groundwater after approximately 2010 (fig. 49, table 19). Given the lower mean groundwater levels within the UFA/Suwannee Limestone (289-ft depth), any upward groundwater flow in this area from the deeper aquifer would result in increased salinity at shallower depths.

ROMP well TR 9-2 is located approximately 2.5 mi from the bay and has data measured within the SAS (10-ft depth), UFA/lower Arcadia Formation (118-ft depth), UFA/Suwannee Limestone (247-ft depth), UFA/Ocala Limestone (622-ft depth), and the UFA/Avon Park Formation (714-ft depth) (table 17). Mean hydraulic heads for 1992–2022 within all but the deepest aquifer vary by only 1.72 ft, having values of 6.92, 8.23, 8.64, and 7.89 ft (NAVD 88) for the SAS, UFA/lower Arcadia Formation, UFA/Suwannee Limestone, and UFA/Ocala Limestone, respectively (figs. 50 and 51, table 18). The relative similarity in heads suggests these hydrogeologic/geologic unit designations are likely well-connected. However, well construction logs document the presence of 230.5 ft of UCU/Hawthorn Group formations between depths of 38 and 268.5 ft, indicating the presence of a confining unit between the SAS and lower aquifer units (Decker, 1990). The measured hydraulic heads for these four hydrogeologic/geologic unit designations and depths have an increasing trend during this period (table 18). The deepest well at ROMP well TR 9-2 is within the UFA/Avon Park Formation (714-ft depth) and generally has lower measured hydraulic heads than those within the shallower units, with a mean value of 4.35 ft (NAVD 88) during the same period (fig. 51, table 18). Kendall’s τ and p-values for measured hydraulic heads within the UFA/Avon Park Formation at TR 9-2 indicate a significant decreasing trend during this period. The larger differences in hydraulic heads suggest the UFA/Avon Park Formation is not well-connected with the shallower units and the likelihood of a confinement layer between them. Well construction logs describe approximately 239.5 ft of core as having low to low-moderate permeability between depths of 457.5 and 697 ft (Decker, 1990).

Chloride concentrations within the shallower wells at ROMP well TR 9-2 for 1990–2022 indicate freshwater conditions, and Kendall’s τ and p-values indicate no significant trends within the SAS (10-ft depth), UFA/lower Arcadia Formation (118-ft depth), and UFA/Suwannee Limestone (247-ft depth), with mean values of 37, 23, and 199 mg/L, respectively (fig. 50, table 19). Kendall’s τ and p-values for the chloride concentration data at the greater depths indicate significant increasing trends, with mean values of 934 and 4,971 mg/L within the UFA/Ocala Limestone (622-ft depth) and UFA/Avon Park Formation (714-ft depth), respectively (fig. 51, table 19). Groundwater within the UFA/Ocala Limestone transitions from fresh to brackish, whereas groundwater within the UFA/Avon Park Formation nearly completes the transition from brackish to saline, with a final value above 9,000 mg/L (fig. 51).

Well TR AB-3 is located approximately 4 mi from the bay and has chloride concentrations measured within the UFA/Avon Park Formation at a single depth of 865 ft with no hydraulic head data (fig. 52, table 19). The mean measured chloride concentrations for 1993–2022 was 3,941 mg/L, and the data indicate an approximately 15-year increasing trend ending in 2008, after which chloride concentrations stabilize at a mean value of 7,900 mg/L. Kendall’s τ and p-values indicated a significant increasing trend in chloride concentration (table 19).

ROMP wells 49, 50, and 51 are located farther east of the bay at approximate distances of 10, 6, and 5 mi from the shoreline, respectively (fig. 22). Well 49 has hydraulic heads measured within the SAS (37-ft depth), UFA/lower Arcadia Formation (230-ft depth), UFA/Suwannee Limestone (410-ft depth), and UFA/Avon Park Formation (792-ft depth) (table 17). The mean hydraulic heads for 1992–2022 measured within the UFA/Suwannee Limestone (410-ft depth) and UFA/Avon Park Formation (792-ft depth) have similar values of 11.67 and 11.50 ft (NAVD 88), respectively (fig. 53, table 18). This similarity suggests these hydrogeologic/geologic unit designations are well-connected. The shallower monitoring well within the SAS (37-ft depth) has a much higher mean hydraulic head of 135.35 ft for this same period, likely because of the higher land surface elevations in the area, 143.1 ft (NAVD 88), and the likelihood of a confining unit beneath (table 17). Well construction logs specify 301.5 ft of UCU/Hawthorn Group between depths of 77.5 and 379 ft, which acts as an effective confining unit between the SAS and deeper aquifer units (Decker, 1989a). The mean hydraulic head within the UFA/lower Arcadia Formation (230-ft depth) for 1991–2022 was 16.20 ft (NAVD 88), which is approximately 5 ft higher than heads within the deeper monitored depths (table 18). This would indicate that the UFA/lower Arcadia Formation and UFA/Suwannee Limestone are not as well-connected as the UFA/Suwannee Limestone and the UFA/Avon Park Formation. There are significant increasing trends in hydraulic head within the UFA/lower Arcadia Formation (230-ft depth), UFA/Suwannee Limestone (410-ft depth), and UFA/Avon Park Formation (792-ft depth) and a decreasing trend within the SAS (37-ft depth).

Chloride concentrations were measured within the UFA/lower Arcadia Formation (230-ft depth) and UFA/Avon Park Formation (792-ft depth). Groundwater within each aquifer unit was fresh, having what appeared to be stable, low chloride levels; however, Kendall’s τ and p-values indicate decreasing chloride concentration trends in both datasets (table 19, fig. 53). The mean measured chloride concentration for 1996–2022 within the UFA/lower Arcadia Formation (230-ft depth) was 5 mg/L (table 19). The mean measured chloride concentration for 1993–2022 for the UFA/Avon Park Formation (792-ft depth) was 14 mg/L.

ROMP well 50 has hydraulic heads measured within the SAS (32-ft depth), UFA/Suwannee Limestone (200-ft depth), and UFA/Avon Park Formation (1,393-ft depth) (table 17). The mean hydraulic head within the SAS (32-ft depth) was 41.24 ft (NAVD 88) for 1979–2022, compared to 10.66 and 5.08 ft (NAVD 88) within the UFA/Suwannee Limestone (200-ft depth) and UFA/Avon Park Formation (1,393-ft depth), respectively, for 1985–2022 (table 18). The higher hydraulic heads measured within the SAS are likely due to higher land surface elevations (44-ft depth) and the likelihood of the presence of a confining unit between the SAS and UFA/Suwannee Limestone hydrogeologic/geologic unit designations. Well construction logs describe 83 ft of UCU/Hawthorn Group between depths of 55 and 138 ft, likely resulting in confinement between the SAS and lower aquifer units (SWFWMD, undated e). The hydraulic head data indicate significant increasing trends within the SAS (32-ft depth) and UFA/Suwannee Limestone (200-ft depth) and no significant trends within the UFA/Avon Park Formation (1,393-ft depth) (table 18, fig. 54).

Chloride concentration data from ROMP well 50 indicated fresh groundwater within the SAS (32-ft depth) and UFA/Suwannee Limestone (200-ft depth), having mean chloride concentrations of 54 and 43 mg/L, respectively, and brackish water within the UFA/Avon Park Formation (1,393-ft depth), having a mean chloride concentration of 2,161 mg/L (fig. 54, table 19). The Kendall’s τ and p-values indicate no significant trend in chloride concentration within the UFA/Avon Park Formation (1,393-ft depth) and decreasing trends in the datasets for the shallower depths (SAS and UFA/Suwannee Limestone). There does seem to be an increase in chloride concentration within the UFA/Avon Park Formation (1,393-ft depth) near the beginning of the dataset, despite the lack of a statistically significant long-term trend.

ROMP well 51 has chloride concentrations measured at a single depth within the UFA/Avon Park Formation (220-ft depth); hydraulic heads were not measured (fig. 55, table 19). The mean measured chloride concentration for 1993–2022 was 31 mg/L, and Kendall’s τ and p-values indicate an increasing trend beginning in 2008. However, the absolute values for measured chloride concentration are still low and indicate fresh groundwater at this depth.

ROMP wells TR 8-1, TR 8-2, and 38 are oriented west to east along a transect that extends inland from the shoreline of Tampa Bay (fig. 22). Well TR 8-1 is approximately 2 mi from the bay and has hydraulic heads measured within the SAS (17-ft depth), UFA/upper Arcadia Formation (100-ft depth), UFA/Suwannee Limestone (390-ft depth), UFA/Ocala Limestone (635-ft depth), UFA/Ocala Limestone (900-ft depth), and UFA/Avon Park Formation (1,130-ft depth) (table 17). Hydraulic head data for 1994–2022 have a mean value of 7.37 ft (NAVD 88) within the SAS (17-ft depth) and no significant long-term trends (fig. 56, table 18). There was an upward vertical gradient measured through the first four monitored depths down to the UFA/Ocala Limestone (635-ft depth), indicating artesian pressure and possible confinement between the SAS and lower hydrogeologic/geologic unit designations. Hydraulic head data for 1991–2022 have a mean of 13.56 ft (NAVD 88) within the UFA/upper Arcadia Formation (100-ft depth) and an increasing trend (table 18). The mean hydraulic heads within the UFA/Suwannee Limestone (390-ft depth) and the UFA/Ocala Limestone (635-ft depth) were similar, having values of 16.39 and 16.51 ft (NAVD 88), respectively, suggesting these hydrogeologic/geologic unit designations are well-connected. The vertical gradient below a depth of 635 ft was downward. The mean hydraulic heads within the UFA/Ocala Limestone (900-ft depth) and UFA/Avon Park Formation (1,130-ft depth) for 1992–2022 were 14.23 ft and −6.11 ft (NAVD 88), respectively, suggesting these hydrogeologic/geologic unit designations are not well-connected (UFA/Ocala Limestone and UFA/Avon Park Formation) (fig. 57, table 18). Well construction logs describe 224 ft of the upper part of the UCU/Hawthorn Group between depths of 39 and 263 ft, indicating likely confinement between the SAS and deeper aquifer units (Dewitt, 1992). Well construction logs also describe a likely confining bed separating the upper-middle part of the UFA/Suwannee Limestone from the top of the UFA/Ocala Limestone. Kendall’s τ and p-values indicate a significant increasing trend in hydraulic head data within the UFA/Avon Park Formation (1,130-ft depth) and no significant trend within the UFA/Ocala Limestone at a depth of 900 ft (table 18).

The mean measured chloride concentrations down to a depth of 635 ft indicate fresh groundwater, having values of 82, 101, 136, and 115 mg/L, respectively, within the SAS (17-ft depth), UFA/upper Arcadia Formation (100-ft depth), UFA/Suwannee Limestone (390-ft depth), and UFA/Ocala Limestone (635-ft depth) (table 19). Groundwater at a depth of 900 ft within the UFA/Ocala Limestone had a mean measured chloride concentration of 2,556 mg/L for 1990–2022, indicating brackish water, whereas data from a shorter measurement period (1991–2000) within the UFA/Avon Park Formation (1,130-ft depth) had a mean value of 17,557 mg/L, indicating saline groundwater (fig. 57, table 19). Kendall’s τ and p-values indicate a decreasing trend within the chloride concentration dataset for the UFA/upper Arcadia Formation (100-ft depth), and increasing trends are indicated within the chloride concentration datasets for the UFA/Suwannee Limestone (390-ft depth), UFA/Ocala Limestone (635-ft depth), and UFA/Ocala Limestone (900-ft depth) (fig. 56, table 19).

ROMP well 8-2 is located approximately 4 mi from the bay and has chloride concentrations measured at a single depth of 917 ft within the UFA/Avon Park Formation; hydraulic heads were not measured (fig. 58, table 19). The mean measured chloride concentration for 1999–2022 was 4,157 mg/L, and Kendall’s τ and p-values for the dataset indicate an increasing trend for the entire period of data collection. The chloride concentrations during this period increase from 1,300 to 6,330 mg/L.

ROMP well 38 is located approximately 8 mi from the bay and has hydraulic heads measured within the SAS (2-ft depth), SAS (10-ft depth), UFA/Suwannee Limestone (425-ft depth), and UFA/Avon Park Formation (1,100-ft depth) (fig. 22, table 17). The mean measured hydraulic heads within the two SAS wells (2- and 10-ft depths) are similar, having values of 32.11 and 31.89 ft (NAVD 88) for 2014–22, respectively (fig. 59, table 18). This is not unexpected considering the similar monitoring depths within the SAS. The mean measured hydraulic heads within the UFA/Suwannee Limestone (425-ft depth) and UFA/Avon Park Formation (1,100-ft depth) are also like each other but lower than those in the SAS, having values of 12.35 and 12.24 ft (NAVD 88), respectively, suggesting the UFA/Suwannee Limestone and UFA/Avon Park Formation are well-connected. The higher hydraulic heads within the SAS are likely due to the higher land surface elevation (about 34-ft depth) at the well (table 17) and the SAS and UFA/Suwannee Limestone hydrogeologic/geologic unit designations not being well-connected. Well construction logs describe two confining units between the SAS and UFA/Suwannee Limestone (Mallams, 2023). The first is 67 ft thick and lies between depths of 40 and 107 ft, and the second is 80 ft thick and lies between depths of 260 and 340 ft. Kendall’s τ and p-values indicate significant decreasing trends in hydraulic head within the SAS (2- and 10-ft depths) and increasing trends within the UFA/Suwannee Limestone (425-ft depth) and UFA/Avon Park Formation (1,100-ft depth).

Chloride concentrations are only measured within the UFA/Avon Park Formation (1,100-ft depth), and the mean value of 73 mg/L for 2017–22 indicates fresh groundwater (table 19). There is an increasing trend in the chloride concentration data within the UFA/Avon Park Formation (1,100-ft depth); however, the concentration values are low and these data were only collected for 6 years (fig. 59).

In summary, wells on the eastern coast of Tampa Bay generally had higher hydraulic heads within the shallower groundwater aquifer units than the lower units because of the higher land-surface elevations within the area and the presence of confining units. Trends in hydraulic head were generally upward. This area has a broad cone of depression whose surface elevation has risen from below sea level during 1978–2011 (figs. 17–20) to above sea level in 2019 (fig. 21). Upward trends in hydraulic head generally were observed, particularly at shallower depths. Chloride concentrations were generally in the range of freshwater except at deeper levels within the UFA/Avon Park Formation (monitored as deep as 1,393 ft below land surface), which were in the range of brackish water.

Groundwater Monitoring Wells Along the Southern Coast of Tampa Bay

Six wells within 5.5 mi of the southern coast of Tampa Bay (fig. 22, table 20) were analyzed for trends in hydraulic head and chloride concentration. ROMP wells TR 7-1, TR 7-2, and TR 7-4 are oriented west to east along a transect south of Tampa Bay that extends inland from the shoreline of the Gulf of Mexico (fig. 22). ROMP well TR 7-1 is located less than 0.5 mi from the shoreline and has hydraulic heads measured at a single depth within the UFA/lower Arcadia Formation (320-ft depth). The mean hydraulic head was 17.34 ft (NAVD 88) for 1983–2022, indicating artesian pressure and likely confinement, with a land-surface elevation of 7.6 ft (NAVD 88) at the well site (fig. 60, tables 20 and 21). Well construction logs describe a multizoned artesian system, with the first zone present between depths of 45 and 140 ft and the second zone present between depths of 240 and 380 ft (New, 1981). The top of the first confining unit is specified to begin at a depth range of approximately 20–25 ft, with the second unit described as existing between the two artesian zones. Kendall’s τ and p-values for the dataset indicate no significant trend in hydraulic head (table 21). Chloride concentration data indicate fresh groundwater having a mean value of 88 mg/L for 1981–2022, and Kendall’s τ and p-values indicate no significant trend (table 22).

ROMP well TR 7-2 is located approximately 3 mi from the bay and has hydraulic heads measured within the SAS (12-ft depth), UCU/Hawthorn Group (57-ft depth), UFA/upper Arcadia Formation (198-ft depth), UFA/Tampa Member (358-ft depth), and UFA/Avon Park Formation (957-ft depth) (table 20). The mean hydraulic heads for 1994–2022 within the SAS (12-ft depth), UCU/Hawthorn Group (57-ft depth), and the UFA/upper Arcadia Formation (198-ft depth) have similar values of 14.43, 13.92, and 14.88 ft (NAVD 88), respectively (fig. 61, table 21). Well construction logs describe an upper confining unit and semiconfining beds between depths of 21 and 155 ft and a lower confining unit between depths of 310 and 332 ft (Rappuhn, 1995). Below 332 ft lies the FAS, which in this area, includes the Tampa Member of the Arcadia Formation. Kendall’s τ and p-values for the datasets indicate an increasing trend in hydraulic head within the UFA/upper Arcadia Formation (198-ft depth) and no significant trends within the SAS (12-ft depth) and UCU/Hawthorn Group (57-ft depth) (table 21). Kendall’s τ and p-values for the datasets indicate an increasing trend in hydraulic head within the UFA/Tampa Member (358-ft depth) and UFA/Avon Park Formation (957-ft depth) beginning in approximately 2012. The mean measured hydraulic heads within the UFA/Tampa Member (358-ft depth) and UFA/Avon Park Formation (957-ft depth) were higher than at the shallower depths and similar to each another, with values of 16.38 and 16.67 ft (NAVD 88), respectively (fig. 62, table 21). This similarity suggests the UFA/Tampa Member and UFA/Avon Park Formation are well-connected, whereas the UFA/Tampa Member is not well-connected to the shallower units.

Mean chloride concentrations within well TR 7-2 indicate fresh groundwater within the SAS (12-ft depth), UCU/Hawthorn Group (57-ft depth), UFA/upper Arcadia Formation (198-ft depth), and UFA/Tampa Member, having values of 106, 78, 53, and 127 mg/L, respectively, during 1992–2022 (table 22). Kendall’s τ and p-values for datasets of chloride concentration indicate an increasing trend within the UCU/Hawthorn Group (57-ft depth) and a decreasing trend within the UFA/upper Arcadia Formation (198-ft depth). Chloride concentrations within the UFA/Avon Park Formation (957-ft depth) had a mean value of 2,008 mg/L for all values within the dataset; however, the data included 18 measurements above 5,000 mg/L, including 14 above 10,000 mg/L (table 22). These higher values are more than an order of magnitude above adjacent measurements within the dataset. If these values are excluded from the calculation, the mean chloride concentration is 453 mg/L. Kendall’s τ and p-values indicate a decreasing trend in chloride concentration within UFA/Tampa Member (358-ft depth) and an increasing trend within the UFA/Avon Park Formation (957-ft depth).

ROMP well TR 7-4 is located approximately 5.5 mi east of the Tampa Bay shoreline and has hydraulic heads measured within the SAS (15-ft depth), UFA/lower Arcadia Formation (213-ft depth), UFA/Tampa Member (377-ft depth), UFA/Suwannee Limestone (560-ft depth), and UFA/Avon Park Formation (1,162-ft depth) (table 20). Measured hydraulic heads within the SAS (15-ft depth) had a mean value of 6.12 ft (NAVD 88) for 1988–2022 and a significant increasing trend (fig. 63, table 21). The measured hydraulic heads within the UFA/lower Arcadia Formation (213-ft depth) were lower than those within the SAS, having a mean value of −3.84 ft (NAVD 88) for 1988–2022 and a significant decreasing trend. Prior to 2006, hydraulic heads within the UFA/lower Arcadia Formation (213-ft depth) were generally higher than those within the SAS (15-ft depth). After 2006, the hydraulic heads within the UFA/lower Arcadia Formation were generally lower than those within the SAS (fig. 63, table 21). After 2019, hydraulic heads within the UFA/lower Arcadia Formation were as low as −90 ft (NAVD 88) during some periods. The changes in hydraulic head at well TR 7-4 after 2006 and again after 2019 indicate changes to the stresses on the system. The differences in hydraulic head and responses within the data indicate the SAS (15-ft depth) and UFA/lower Arcadia Formation (213-ft depth) are not well-connected. Well construction logs describe the presence of 356 ft of UCU/Hawthorn Group between depths of 18 and 374 ft, which effectively acts as a confining unit (Decker, 1989b).

Hydraulic heads deeper within the aquifer at well TR 7-4 are generally higher than those within the SAS and UFA/lower Arcadia Formation. The mean measured hydraulic heads were 14.16, 14.11, and 12.51 ft (NAVD 88) within the UFA/Tampa Member (377-ft depth), UFA/Suwannee Limestone (560-ft depth), and the UFA/Avon Park Formation (1,162-ft depth) (fig. 64, table 21). The similarity in heads within the UFA/Tampa Member and UFA/Suwannee Limestone and difference between both and heads within the UFA/Avon Park Formation suggest the UFA/Tampa Member and UFA/Suwannee Limestone are well-connected and the UFA/Suwannee Limestone and deeper UFA/Avon Park Formation are not well-connected. Kendall’s τ and p-values for the hydraulic head data within the UFA/Tampa Member (377-ft depth) and UFA/Suwannee Limestone (560-ft depth) indicate increasing trends that are visible after approximately 2012 (fig. 64, table 21). Kendall’s τ and p-values for the hydraulic head data within the UFA/Avon Park Formation (1,162-ft depth) indicate no significant trend.

Chloride concentration data collected at all but the deepest depth at ROMP well TR 7-4 indicate fresh groundwater having mean values of 32, 32, 28, and 47 mg/L within the SAS (15-ft depth), UFA/lower Arcadia Formation (213-ft depth), UFA/Tampa Member (377-ft depth), and UFA/Suwannee Limestone (560-ft depth), respectively (table 22). Kendall’s τ and p-values indicate a significant decreasing trend in chloride concentration within the UFA/lower Arcadia Formation (213-ft depth), with freshening occurring from 1995 to approximately 2008. Kendall’s τ and p-values for the data indicate an increasing trend in chloride concentration within the UFA/Suwannee Limestone (560-ft depth) and no significant trend within the UFA/Tampa Member (377-ft depth). Kendall’s τ and p-values indicated an increasing trend in chloride concentration within the UFA/Avon Park Formation (1,162-ft depth) with a transition to brackish sometime after 2015 and a mean value of 634 mg/L for 1991–2022 (fig. 64, table 22).

ROMP wells TR SA-1 and TR SA-3 are located farther south along the gulf coast than TR 7-1, TR 7-2, and TR 7-4 (fig. 22). Well TR SA-1 is less than 0.3 mi from the bay and has hydraulic heads measured within the SAS (4-ft depth), UFA/lower Arcadia Formation (328-ft depth), UFA/Suwannee Limestone (708-ft depth), and UFA/Avon Park Formation (995-ft depth). During 1998–2022, mean hydraulic heads had an upward gradient, with values of 1.66, 4.95, 9.78, and 17.04 ft (NAVD 88) within the SAS (4-ft depth), UFA/lower Arcadia Formation (328-ft depth), UFA/Suwannee Limestone (708-ft depth), and UFA/Avon Park Formation (995-ft depth), respectively (fig. 65, table 21). With a land-surface elevation of 6.0 ft at TR-SA1, the mean hydraulic head values indicate artesian pressure within the UFA/Suwannee Limestone and UFA/Avon Park Formation and the likely presence of one or more confining units between the UFA/Suwannee Limestone and the shallower units. Well construction logs describe the presence of approximately 469 ft of UCU or IAS/ICU between depths of 29 and 498 ft (Lee, 1998). Additionally, Lee (1998) noted that the Ocala Limestone is considered to be a semiconfining unit separating the permeable beds of the Suwannee Limestone and Avon Park Formation. Kendall’s τ and p-values for the hydraulic head data within the SAS (4-ft depth) and UFA/Suwannee Limestone (708-ft depth) do not indicate any significant trends for 1998–2022 (table 21). Kendall’s τ and p-values for the measured hydraulic head data within the UFA/Avon Park Formation (995-ft depth) indicate an increasing trend, whereas the Kendall’s τ and p-values for hydraulic head data collected within UFA/lower Arcadia Formation (328-ft depth) indicate a decreasing trend.

Kendall’s τ and p-values for chloride concentration data from ROMP well TR SA-1 indicate increasing trends within the UFA/Suwannee Limestone (708-ft depth) and UFA/Avon Park Formation (995-ft depth), having mean values of 1,105 and 773 mg/L, respectively (tables 22 and 20). The mean value of 773 mg/L within the UFA/Avon Park Formation is skewed by four measurements above 1,000 mg/L, one of which is over 17,000 mg/L. If these outliers are excluded from the calculation of the mean, the mean concentration would be 501 mg/L. The chloride concentrations within the SAS (4-ft depth) and UFA/lower Arcadia Formation (328-ft depth) indicate fresh groundwater at these depths, having mean values of 604 and 182 mg/L, respectively, and Kendall’s τ and p-values indicate no significant trends.

ROMP well TR SA-3 is located approximately 2.5 mi from the Tampa Bay shoreline and has groundwater and chloride concentrations measured at a single depth of 1,096 ft within the UFA/Avon Park Formation (fig. 22, table 20). The mean hydraulic head and chloride concentration for 1998–2022 at this depth were 17.72 ft (NAVD 88) and 264 mg/L, respectively (fig. 66, tables 21 and 22). Kendall’s τ and p-values for the hydraulic head data indicate an increasing trend, with the final yearly average being 5.5 ft higher than the average during the first year (table 21). Kendall’s τ and p-values for the chloride concentration data indicate an increasing trend for the measurement period (table 22).

ROMP well TR 6-1 is located approximately 0.3 mi from the Tampa Bay shoreline and has groundwater and chloride concentrations measured at a single depth of 315 ft within the UFA/lower Arcadia Formation (fig. 22, table 20). Despite not having water levels monitored at multiple depths, four major water-level changes were described in the well construction report for this site (SWFWMD, undated f). Water under artesian pressure was first encountered within the Tamiami Formation at a depth of approximately 40 ft, with hydraulic heads approximately 2.8 ft above the land surface elevation of 4.2 ft (NAVD 88) at the well site. A second artesian zone was encountered within the upper part of the UCU/Hawthorn Group at a depth of approximately 90 ft, with hydraulic heads approximately 2 ft above land-surface elevation. A third artesian zone was encountered within an unspecified formation at a depth of approximately 225 ft with hydraulic heads approximately 4.5 ft above land-surface elevation. A fourth artesian zone was encountered within the UFA/Suwannee Limestone at an approximate depth of 520 ft, with hydraulic heads approximately 13.5 ft above land-surface elevation. The mean hydraulic head and chloride concentration for 1979–2022 at this depth were 4.17 ft (NAVD 88) and 480 mg/L (fig. 67, tables 21 and 22). Kendall’s τ and p-values for the hydraulic head data indicate a decreasing trend that is visible across the entire measurement period (table 21). Kendall’s τ and p-values for the chloride concentration data indicate a significant increasing trend, with a visible period of approximately 10 years, in the beginning of the measurement period followed by a relatively stable period (fig. 67, table 22).

In summary, wells along the southern coast of Tampa Bay have upward water-level trends at deeper depths (TR 7-2, TR 7-4), but have exhibited responses to hydrologic stress where potentiometric levels have dropped to as low as −90 ft (NAVD 88) and recovered since 2019. The causes of the rapid changes have not been identified; however, these changes illustrate the potential sensitivity of karst groundwater systems.