WRIR 01-4275


You can VIEW THIS REPORT as an html document on the Statistical Overview of Select USGS Water-level Monitoring Sites web site
Or you can DOWNLOAD THIS REPORT (6,344 KB) in Portable Document Format (PDF)
The Adobe PDF Reader program is available for free from Adobe.

Prinos, Scott T., Lietz, A.C., and Irvin, R.B., 2002, Design of a Real-Time Ground-Water Level Monitoring Network and Portrayal of Hydrologic Data in Southern Florida: Water Resources Investigations Report 01-4275, 108 p.

ABSTRACT:

Ground-water resources in southern Florida are under increasing stress caused by a rapid growth in population. As a result of increased demands on aquifers, water managers need more timely and accurate assessments of ground-water conditions in order to avoid or reduce adverse effects such as saltwater intrusion, loss of pumpage in residential water-supply wells, land-surface subsidence, and aquifer compaction.

Hydrologic data were analyzed from three aquifer systems in southern Florida: the surficial aquifer system, which includes the Biscayne aquifer; the intermediate aquifer system, which includes the sandstone and mid-Hawthorn aquifers; and the Florida aquifer system represented by the lower Hawthorn producing zone. Long-term water-level trends were analyzed using the Seasonal Kendall trend test in 83 monitoring wells with a daily-value record spanning 26 years (1974-99). The majority of the wells with data for this period were in the Biscayne aquifer in southeastern Florida. Only 14 wells in southwestern Florida aquifers and 9 in the surficial aquifer system of Martin and Palm Beach Counties had data for the full period. Because many monitoring wells did not have data for this full period, several shorter periods were evaluated as well. The trend tests revealed small but statistically significant upward trends in most aquifers, but large and localized downward trends in the sandstone and mid-Hawthorn aquifers.

Monthly means of maximum daily water levels from 246 wells were compared to monthly rainfall totals from rainfall stations in southwestern and southeastern Florida in order to determine which monitoring wells most clearly indicated decreases in water levels that corresponded to prolonged rainfall shortages. Of this total, 104 wells had periods of record over 20 years (after considering missing record) and could be compared against several drought periods. After factors such as lag, seasonal cyclicity, and cumulative functions were considered, the timing of minimum values of water level from 15 ground-water monitoring wells and average minimum rainfall values agreed 57 to 62 percent of the time over a 20 to 26 year period. On average, the timing of water-level minimums and rainfall minimums agreed about 52 percent of the time, and in some cases only agreed 29 percent of the time.

A regression analysis was used to evaluate daily water levels from 203 monitoring wells that are currently, or recently had been, part of the network to determine which wells were most representative of each aquifer. The regression also was used to determine which wells provided data that could be used to provide estimations of water levels at other wells in the aquifer with a coefficient of determination (R2 value) from the regression of 0.64 or greater. In all, the regression analysis alone indicated that 35 wells, generally with 10 years or more of data, could be used to directly monitor water levels or to estimate water levels at 180 of 203 wells (89 percent of the network). Ultimately, factors such as existing instrumentation, well construction, long-term water-level trends, and variations of water level and chloride concentration were considered together with the R2 results in designing the final network.

The Seasonal Kendall trend test was used to examine trends in ground-water chloride concentrations in 113 wells. Of these wells, 61 showed statistically significant trends. Fifty-six percent (34 of 61 wells) of the observed trends in chloride concentration were upward and 44 percent (27 of 61 wells) were downward. The relation between water level and chloride concentration in 114 ground-water wells was examined using Spearman's r and Pearson's r correlation coefficients. Statistically significant results showed both positive and negative relations. Based on the results of statistical analyses, period of record, well construction, and existing satellite telemetry, 33 monitoring wells were selected that could be used to assess ground-water conditions in 167 monitoring wells in southern Florida on an interim basis.

A real-time ground-water level monitoring network was designed to provide this information, and a prototype website (http://www.sflorida.er.usgs.gov/ddn_data/index.html) was constructed to provide water managers with daily updates on ground-water conditions in southern Florida. Many of the same analytical tools used to select monitoring wells representative of aquifer conditions are also employed to analyze data for this website. These tools include regression analysis, the Seasonal Kendall trend test, and frequency analysis. The website also includes image maps showing the current conditions for stations in selected geographical areas and aquifers and statistical comparison plots for each station.


CONTENTS:

Introduction
Purpose and Scope
Description of Study Area
Hydrogeologic Setting
Previous Studies
Acknowledgments
Water Use and Precipitation
Population and Water Use
Precipitation
Effects of Water Use
Loss of Pumpage
Aquifer Compaction
Saltwater Contamination
Real-Time Ground-Water Level Monitoring Network Design
Criteria for Selecting Network Wells
Well Construction and Period of Record
Analysis of Long-Term Water-Level Trends in Network Wells
Summary Statistics of Water-Level Data from Candidate Monitoring Wells
Determining Water-Level and Rainfall Correlation
Application of Analytical Technique
Regression Analysis of Network Wells
Analysis of Water-Level and Chloride Data
Chloride Concentration Trends
Relation Between Chloride Concentrations and Water Levels
Selection of Index Wells by Aquifer
Water-Table Aquifer
Water-Level and Chloride Concentration Trend and Correlation Results
Discussion of Well Coverage
Lower Tamiami Aquifer
Water-Level and Chloride Concentration Trend and Correlation Results
Discussion of Well Coverage
Sandstone Aquifer
Water-Level and Chloride Concentration Trend and Correlation Results
Discussion of Well Coverage
Mid-Hawthorn Aquifer
Water-Level and Chloride Concentration Trend and Correlation Results
Discussion of Well Coverage
Surficial Aquifer System
Water-Level and Chloride Concentration Trend and Correlation Results
Discussion of Well Coverage
Biscayne Aquifer
Water-Level and Chloride Concentration Trend Results
Water-Level and Chloride Concentration Correlation Analysis
Discussion of Well Coverage
Portrayal of Real-Time Ground-Water Level Data
Basic Depiction of Data
Regression and Frequency Data Summary and Conclusions
Selected References
Appendix I: Well Construction and Period of Record
Appendix II: Results of Seasonal Kendall Trend Tests of Continuous Water-Level Data
Appendix III: Summary Statistics of Water-Level Data from Candidate Monitoring Wells

FIGURES:
1. Map showing location of study area
2. Column showing comparison of hydrogeologic nomenclature for southern Florida
3. Graphs showing population and ground-water use in the southeastern, southwestern, and northeastern parts of the study area
4-6. Maps showing:

4. Locations of continuous ground-water level monitoring wells that were considered for the real-time ground-water level monitoring network in Collier, Lee, and Hendry Counties
5. Locations of continuous ground-water level monitoring wells that were considered for the real-time ground-water level monitoring network in Palm Beach, Martin, and St. Lucie Counties
6. Locations of continuous ground-water level monitoring wells that were considered for the real-time ground-water level monitoring network in Miami-Dade and Broward Counties
7-9. Graphs showing:
7. Effect of long-term trends on correlation analyses of water level in well and rainfall at well
8. Comparison of smoothed average rainfall deviations and trend adjusted water-level deviations in well G-3264A
9. Steps used in computation of rainfall model
10. Map showing locations of salinity monitoring wells
11. Box plot showing seasonal variation in chloride concentration in well G-1351
12. Map showing network coverage defined by R2 analysis using index wells in the water-table aquifer
13. Graphs showing comparison of smoothed average rainfall deviations and trend adjusted water-level deviations in well C-392
14. Plots showing statistically significant trends in water level at selected wells in the lower Tamiami aquifer
15. Graph showing chloride concentration trends at wells C-489, C-525, and C-528 and L-738
16. Map showing network coverage defined by R2 analysis using index wells in the lower Tamiami aquifer
17. Plots showing statistically significant trends in water level at selected wells in the sandstone aquifer
18-20. Graphs showing:
18. Comparison of water levels at well L-2215 and water levels estimated at well L-2215 using water-level data from well L-729
19. Comparison of water levels at well L-727 and water levels estimated at well L-727 using water-level data from well L-2186
20. Comparison of smoothed average rainfall deviations and trend adjusted water-level deviations in well HE-517
21. Map showing network coverage defined by R2 analysis using index wells in the sandstone aquifer
22. Plots showing statistically significant trends in water level at selected wells in the mid-Hawthorn aquifer
23. Graphs showing chloride concentration trends at wells L-735, L-2820, L-2702, L-1109 and L-2640 in the mid-Hawthorn aquifer
24-25. Maps showing:
24. Network coverage defined by R2 analysis using index wells in the mid-Hawthorn aquifer
25. Network coverage defined by R2 analysis using index wells in the surficial aquifer system
26. Graph showing comparison between water-level data from well STL-175 and estimation of water levels using data from well M-1004
27. Map showing comparison of minimum monthly means of maximum daily water levels and long-term trends in water level and chloride concentration in the Biscayne aquifer
28-31. Graphs showing:
28. Water-level and chloride concentration trends in selected wells near the Hialeah-Miami Springs well field
29. Chloride concentration in wells G-432 and G-901 near the Alexander Orr Well Field
30. Chloride concentration in wells G-1241, G-1435, G-2410, and G-2478 near the Hallandale Well Field
31. Chloride concentration in wells G-854, G-2352, G-1343, G-2125, and G-2130 near the Fort Lauderdale municipal well field
32-33. Maps showing:
32. Network coverage defined by R2 analysis using index wells in the Biscayne aquifer
33. Locations of surface-water monitoring sites shown on the real-time prototype website
34-37. Graphs showing:
34. Long-term chloride concentrations in well L-738, March 24, 1976, to July 11, 2000
35. Water-level data for the previous 30 days, 90 days, and current year and comparison with the long-term data of well L-581 in the mid-Hawthorn aquifer, September 27, 2000
36. Daily maximum water level in well L-581, September 28, 1975, to September 27, 2000
37. Current and historical water levels in well L-581, September 16, 1999, to September 27, 2000
38. Map showing water levels in selected wells in Lee County, September 27, 2000

TABLES:
1. Rainfall stations used for the southeastern and southwestern rainfall models
2. Rainfall and water-level minimum comparison for aquifers in southern Florida
3. Statistically significant Seasonal Kendall trend test results for chloride concentrations in water at selected wells in southern Florida based on two seasons per year
4. Statistically significant correlation between chloride concentrations and instantaneous water levels for selected wells in southern Florida
5. Potential index wells and well groupings based on regression analysis of aquifers in southern Florida
6. Subnetwork regression characteristics by aquifer
7. Basic well construction and statistical information for potential index wells in southern Florida


[an error occurred while processing this directive]