UFDC Home myUFDC Home  |   Help  |   RSS
<%BANNER%>
HIDE
 Title Page Abstract Introduction Description of area Hydrology Significance of the hydrology of... References Glossary

 Interim report on the hydrologic features of the Green Swamp area in central Florida ( FGS: Information circular 26 )
CITATION SEARCH THUMBNAILS PDF VIEWER PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
 Material Information Title: Interim report on the hydrologic features of the Green Swamp area in central Florida ( FGS: Information circular 26 ) Series Title: ( FGS: Information circular 26 ) Physical Description: 96 p. : illus. ; Language: English Creator: Florida Geological SurveyPride, R. W Publisher: s.n. Place of Publication: Tallahassee Publication Date: 1961
 Subjects Subjects / Keywords: Hydrology -- Florida   ( lcsh )Groundwater -- Florida   ( lcsh )Water-supply -- Florida   ( lcsh )Green Swamp (Fla.)   ( lcsh ) Genre: non-fiction   ( marcgt )
 Notes Statement of Responsibility: by R. W. Pride, F. W. Meyer, and R. N. Cherry. General Note: "References": p. 91-92 Funding: Digitized as a collaborative project with the Florida Geological Survey, Florida Department of Environmental Protection.
 Record Information Source Institution: University of Florida Holding Location: University of Florida Rights Management: The author dedicated the work to the public domain by waiving all of his or her rights to the work worldwide under copyright law and all related or neighboring legal rights he or she had in the work, to the extent allowable by law. Resource Identifier: aleph - 001044241oclc - 01721632notis - AFC7086 System ID: UF00001086:00001

UF00001086 ( PDF )

Title Page
Page i
Page ii
Page iii
Page iv
Page v
Page vi
Abstract
Page 1
Page 2
Introduction
Page 2
Page 3
Page 4
Page 5
Page 6
Page 7
Page 8
Page 9
Page 10
Page 11
Page 12
Page 13
Page 14
Description of area
Page 15
Page 16
Page 17
Page 18
Page 19
Page 20
Page 21
Page 22
Page 23
Page 24
Page 24a
Page 14
Page 25
Page 26
Page 27
Page 28
Page 29
Page 30
Page 31
Page 32
Page 33
Hydrology
Page 34
Page 35
Page 36
Page 36a
Page 37
Page 38
Page 39
Page 40
Page 41
Page 42
Page 43
Page 44
Page 45
Page 46
Page 47
Page 48
Page 49
Page 50
Page 51
Page 52
Page 53
Page 54
Page 55
Page 56
Page 57
Page 58
Page 59
Page 60
Page 61
Page 62
Page 63
Page 64
Page 65
Page 66
Page 67
Page 68
Page 69
Page 70
Page 71
Page 72
Page 73
Page 74
Page 76
Page 75
Page 76
Page 77
Page 78
Page 79
Page 80
Page 81
Page 82
Page 84
Page 83
Page 84
Page 85
Page 86
Page 87
Significance of the hydrology of the area
Page 88
Page 89
Page 90
Page 87
References
Page 91
Page 92
Glossary
Page 93
Page 94
Page 95
Page 96
Main
Full Text

STATE OF FLORIDA
STATE BOARD OF CONSERVATION

FLORIDA GEOLOGICAL SURVEY
Robert O. Vernon, Director

INFORMATION CIRCULAR NO. 26

INTERIM REPORT
ON
THE HYDROLOGIC FEATURES
OF
THE GREEN SWAMP AREA IN CENTRAL FLORIDA

By
R. W. Pride, F. W. Meyer, and R. N. Cherry

Prepared by U. S. Geological Survey
in cooperation with
The Florida Geological Survey
and
Florida Department of Water Resources

Tallahassee, Florida
1961

lteJIV 4 f

AGRI-
CULTURAL
LIBRARY

Page

Abstract.............. ......................... 1
Introduction .................. ................. .. 2
Purpose and scope .......................... 3
Previous investigations ................ ..... 5
Methods of investigation ................ ..... 10
Acknowledgments ........................... 14
Description of the area........................... 14
Location .................................. 15
Topography................................ 17
D rainage ................................... 19
Culture and development .................... 20
Clim ate .................................... 21
Precipitation............................ 21
Temperature............................. 25
G eology .................................... 26
Formations............................. 27
Avon Park limestone ................ 29
Ocala group......................... 29
Suwannee limestone ................. 31
Tampa formation ...... ............. 31
Hawthorn formation ................. 31
Undifferentiated sand, clay and
sandstone........................ 31
Structure .............................. 33
Hydrology ....................................... 34
General ...................................... 34
Surface water..... ......................... 35
Occurrence and movement................. 35
Characteristics of drainage basins ....... 35
Withlacoochee River ................ 36
Little Withlacoochee River .......... 39
Palatlakaha Creek.................. 40
Hillsborough River ................. 43
Reedy Creek ............... ......... 44
Diversions and interconnection of basins. 44
Runoff characteristics ......... ......... 47
Seasonal distribution. .............. 47
Areal distribution and basin funoff.... 51

Rainfall-runoff relation .................. 56
The effects of man-made changes......... 62
Chemical characteristics of surface water. 66
Ground water .................. ............ 71
General................................. 71
Nonartesian ground water ................ 72
Occurrence ........................ 72
Fluctuations of the water table ....... 74
Recharge and discharge ............. 78
Artesian ground water ................... 79
Occurrence ................... ...... 79
Fluctuations of thepiezometric surface 79
Shape of the piezometric surface ..... 82
Chemical characteristics of ground water 85
Significance of the hydrology of the area .......... 87
Eastern part ....................... ........... 88
Western part....................... .......... 88
Theoretical effects of increased drainage of the
Green Swamp area ...................... 89
Theoretical effects of water conservation in the
Green Swamp area....................... 90
References..................................... 91
Glossary...................................... 93

ILLUSTRATIONS

Figure
1 Map of Florida showing location of Green
Swamp area............................. 16
2 Average monthly rainfall of Green Swamp
area ................................... 23
3 Average annual rainfall of Green Swamp
area ................................... 23
4 Map of Green Swamp area showing isohyetals
for 1959 water year and principal streams
draining the area ..................... facing 24
5 Relation of annual water loss in humid areas
to temperature.......................... 25
6 Generalized geologic cross sections across
the Green Swamp area and their locations 28

7 Surface-water drainage features in Green
Swamp area of central Florida ......... facing 36
8 Flow diagram of the upper Oklawaha River 41
9 Annual discharge of Withlacoochee River
at Trilby, Florida....................... 49
10 Average monthly discharge of Withlacoochee
River at Trilby, Florida. ................. 49
11 Flow-duration curve of Withlacoochee River
at Trilby, Florida................ ....... 50
12 Relation of effective annual rainfall and
annual water loss, Withlacoochee River at
Trilby, Florida, 1931-59................. 59
13 Relation of effective annual rainfall and
annual runoff, Withlacoochee River at Trilby,
Florida, 1931 59 ....................... 59
14 Relation of effective annual rainfall and
annual water loss, Palatlakaha Creek above
Mascotte, Florida, 1946-59 .............. 61
15 Relation of effective annual rainfall and
annual runoff, Palatlakaha Creek above
Mascotte, Florida, 1946-5. .............. 61
16 Double-mass curves of measured runoff
versus computed runoff, Withlacoochee
River and Palatlakaha Creek basins....... 64
17 Mineral content in the Withlacoochee River 68
18 Generalized hydrologic cross section along
line B-B', figure 6....................... 73
19 Map of Green Swamp area showing ground-
water data-collection points ............ facing 74
20 Hydrographs of long-term water-level re-
cords from wells near the Green Swamp
area ................................... 75
21 Hydrographs of water-level and rainfall re-
cords at wells in the Green Swamp area... 77
22 Map of Green Swamp area showingthe shape
of the piezometric surface of the Floridan
aquifer .............................. facing 82

Table
1 Surface-water data-collection points in the
Green Swamp area and vicinity ........... 6
2 Test-well data in the Green Swamp area.. 13
3 Outflow from Green Swamp area, water
year 1958-59............................. 52

INTERIM REPORT ON THE
HYDROLOGIC FEATURES OF THE GREEN SWAMP
AREA IN CENTRAL FLORIDA

By
R. W. Pride, F. W. Meyer, and R. N. Cherry

ABSTRACT

The Green Swamp area as usedin this report is located
near the center of the Florida Peninsula. It covers an area
of almost 900 square miles of swampy flatlands and sandy
ridges. The elevation of the land surface varies from about
200 feet above mean sea level in the eastern part to about
75 feet in the river valleys in the western part of the area.

About 720 square miles of the Green Swamp area is
drained by the Withlacoochee River. Streams that drain
into the St. Johns River, Hillsborough River, Kissimmee
River, and Peace River basins also originate in or near
the area.

The drainage divides of these basins are broken in sev-
eral places by swamp channels and gaps in the surrounding
ridges. Water may flow through these gaps from one basin
to another, the direction often not definitely established but
depending onthe relative elevation of the water level in each
basin. These interconnections have a significant influence
on the surface drainage pattern of the area.

During the water year ending September 30, 1959, the
rainfall over the Green Swamp area was about 72. 5 inches.

FLORIDA GEOLOGICAL SURVEY

This was 20 inches above normal. The runoff from the area
during this period was 24. 26 inches.

Drainage operations in recent years have not signifi-
cantly changed the amount of annual runoff from the total
area. However, the distribution of the drainage has been
changed by canals that divert some of the flow fromthe upper
Palatlakaha Creek into the Withlacoochee River.

The Floridan aquifer underlies all of the Green Swamp
area. It is composed of porous marine limestones. The
aquifer crops out inthe western part of the area and occurs
at depths of more than 200 feet below land surface in the
eastern part. The Floridan aquifer is overlain by a non-
artesian aquifer which consists primarily of sand and clay.
The nonartesian aquifer is thin or absent inthe western part
of the area and ranges from 50 to 100 feet in thickness in
the eastern part. The principal source of recharge of ground
water in the nonartesian aquifer is local rainfall.

Piezometric levels in the Floridan aquifer occur at an
elevation of about 130 feet above mean sea level inthe south-
eastern part of the area. Recharge to the Floridan aquifer
occurs along the eastern side of the area.

The maximum mineral content found in the surface water
was 122 parts per million and the maximum in ground water
was 350 parts per million. Generally water is considered
to be usable if the mineral content is less than 400 to 500
parts per million. Surface water was highly colored but
ground water was relatively clear.

INTRODUCTION

To satisfy the demands of a rapidly increasing popula-
tion, many acres of land in Florida are converted each year
to residential and industrial uses. Most of this land previ-
ously had been devoted to agricultural uses. Urbanization
of these areas and the demand for increasing the food supply
thus require that man search for new areas to develop for
agricultural uses. This search, in many instances, has led
to the development of marginal lands.

FLORIDA GEOLOGICAL SURVEY

This was 20 inches above normal. The runoff from the area
during this period was 24. 26 inches.

Drainage operations in recent years have not signifi-
cantly changed the amount of annual runoff from the total
area. However, the distribution of the drainage has been
changed by canals that divert some of the flow fromthe upper
Palatlakaha Creek into the Withlacoochee River.

The Floridan aquifer underlies all of the Green Swamp
area. It is composed of porous marine limestones. The
aquifer crops out inthe western part of the area and occurs
at depths of more than 200 feet below land surface in the
eastern part. The Floridan aquifer is overlain by a non-
artesian aquifer which consists primarily of sand and clay.
The nonartesian aquifer is thin or absent inthe western part
of the area and ranges from 50 to 100 feet in thickness in
the eastern part. The principal source of recharge of ground
water in the nonartesian aquifer is local rainfall.

Piezometric levels in the Floridan aquifer occur at an
elevation of about 130 feet above mean sea level inthe south-
eastern part of the area. Recharge to the Floridan aquifer
occurs along the eastern side of the area.

The maximum mineral content found in the surface water
was 122 parts per million and the maximum in ground water
was 350 parts per million. Generally water is considered
to be usable if the mineral content is less than 400 to 500
parts per million. Surface water was highly colored but
ground water was relatively clear.

INTRODUCTION

To satisfy the demands of a rapidly increasing popula-
tion, many acres of land in Florida are converted each year
to residential and industrial uses. Most of this land previ-
ously had been devoted to agricultural uses. Urbanization
of these areas and the demand for increasing the food supply
thus require that man search for new areas to develop for
agricultural uses. This search, in many instances, has led
to the development of marginal lands.

INFORMATION CIRCULAR NO. 26

The most productive lands for manytypes of agriculture
have been found in river flood plains and in swamp bottoms
where for centuries deposits of fertile organic soil have been
built up. The reclamation of these areas for agriculture has
required extensive drainage developments. In many places,
drainage canals and ditches have been constructed without
any provisionfor water control and the inevitable result has
been overdrainage.

The development of the Everglades of southern Florida
about half a century ago was in the interest of reclamation
and the drastic effects of overdrainage were not foreseen.
The construction of drainage canals lowered the average
water level several feet, not only in the Everglades but also
in the coastal ridge. This lowering of the fresh water level
resulted in salt water moving inland along the coastal area
to pollute the ground water and to threaten the water supply
of the heavily populated area along the lower east coast. As
the organic soils of the Everglades dried up, they began to
shrink and the land surface began to subside. The mucky
soils in many areas caught fire and almost completely dis-
appeared. These are some of the results of overdrainage
of the Everglades. Present plans for redeveloping the Ever-
glades include conservation areas and water-control struc-
tures, as well as more effective drainage canals.

Purpose and Scope

The Green Swamp area in central Florida is another
area where man is developing agricultural land from mar-
ginal land. Though the area is by no means as extensive as
that of the Everglades, the present efforts for its develop-
ment are similar to the early efforts for developing the Ever-
glades in that many miles of canals and ditches have been
constructed to improve the drainage.

Lest the early mistakes of the Everglades be repeated,
the Florida Department of Water Resources considered that
an appraisal of the physical and hydrologic features of the
area was needed to determine the broad effects of draining
and developing the swamp.

FLORIDA GEOLOGICAL SURVEY

The future water supplies in many localities may de-
pend, in part, upon the management of the waters of this
one area. The purpose of the investigation that preceded
this report was to define the areal hydrology and its effect
upon that of other areas and thus tofacilitatebetter manage-
ment.

A reconnaissance of the general hydrology of the Green
Swamp area was made by the U..S. Geological Survey in co-
operation with the Florida Geological Survey and the Florida
Department of Water Resources. The investigation covered
a 2-year period beginning July 1, 1958. During the first
part of this period, water records were collected; the drain-
age characteristics of the area were determined by field
reconnaissance; and many of the facts concerning the phys-
ical and geologic features and their influence on the hydro-
logy were established. The report of this investigation was
prepared during the latter part of the 2-year period.

Even prior to this investigation, the Green Swamp area
was knowntobe the source of several large streams in cen-
tral Florida. Also the area was believed to be an important
recharge area for ground water in central Florida. Other
than this general information, very little was known of the
hydrology of the area.

The scope of this reconnaissance does not permit a com-
prehensive appraisal of the water resources of the area.
However, the reconnaissance does provide information re-
quired by the State of Florida for determining its respon-
sibility and policy in regard to the Green Swamp area and
for formulating future plans for water management of the
area.

Some of the features that have been determined are:
the amount of rainfall on the area; the pattern of surface-
water drainage; the amount and direction of surface-water
runoff; the direction of ground-water movement; the inter-
relationship of rainfall, surface water, and ground water;
the effects of improved drainage facilities'; and the effects
of the hydrologic environment on the chemical quality of
water of the area.

INFORMATION CIRCULAR NO. 26

A generalappraisal of the hydrology of the Green Swamp
area and its significance to central Florida has been made
on the basis of the findings of this investigation. This re-
connaissance will be the foundation for a more comprehen-
sive investigation.

Previous Investigations

Only minor investigations of the water resources and
geology of the Green Swamp area were made before the re-
connaissance described in this report was started. A few
long-term records of streamflow and ground-water levels
had been collected in the vicinity, however, as part of the
statewide data-collection programs.

Records of streamflow have been collected for several
years at sites near the boundaries of the area. Table 1
lists the gaging stations that have long-term records and
the points where surface-water data were collected for the
present investigation. Locations of these points are shown
in figure 7. The longest streamflow record listed is that
for the Withlacoochee River at Trilby (station 39 in table 1)
where continuous records have been collected since 1930.
Streamflow records for the gaging stations listed in table 1
have been published annually in water-supply papers of the
U. S. Geological Survey.

Ground-water levels in three wells near the area of in-
vestigation have been recorded and are published in the an-
nual water-supply papers of the U. S. Geological Survey.
They are as follows: well 810-136-1 (Polk 44), which taps
the Floridan aquifer, is located about 1. 5 miles northeast
of Davenport, Polk County, and continuous records have
been collected since 1946; well 810-136-1 (Polk 47), which
taps the nonartesian aquifer, is also located about 1.5 miles
northeast of Davenport, Polk County, and continuous re-
cords have been collected since 1948; and well 816-211-1
(Pasco 16), which taps the Floridan aquifer, is located about
1. 5 miles north of Zephyrhills, Pasco County, and records
have been collected since 1936.

Table 1, Surface-Water Data Collection Points in the Green Swamp Area and Vicinity

Station Drainage area
No. Location (square miles )) Type of data Period ofrecord

1

Lowery Lake near Haines City, Florida

Cypress swamp in sec. 3, T. 27 S.,
R. 26 E., near Haines City, Florida

Cypress swamp in sec. 34, T. 25 S.,
R. 26 E., near Haines City, Florida

Big Creek near Clermont, Florida

Little Creek near Clermont, Florida

Lake Louisa near Clermont, Florida

Lake Minnehaha at Clermont, Florida

Palatlakaha Creek at Cherry Lake
outlet near Oroveland, Florida

Palatlakaha Creek near Mascotte,
Florida

Lake Apopka at Winter Garden, Florida

Shingle Creek at airport near
Kiseimmee, Florida

Cypress Creek near Vineland, Florida

Chemical analysis of I water sample

Chemical analysis of I water sample

Chemical analysis of 1 water sample

Occasional discharge measurement
Daily record of stage and discharge
Chemical analyses of 4 water samples

Occasional discharge measurement
Discharge measurement at monthly
interval
Chemical analyses of 4 water samples

Daily record of stage
Chemical analysis of 1 water sample

Daily record of stage

Daily record of stage and discharge

Daily record of stage and discharge

Monthly observation of stage
Daily record of stage

Daily record of stage and discharge

Daily record of stage and discharge

I- ~ -

Nov. 13, 1959

May 5, 1959

May 5, 1959

1945-47, 1952-56.
July 1958 to date.
Aug. 19, 1958; Mar.17, 23; Nov. 19, 1959

1945-41 1952-56.

July 1958 to date.
Aug. 19, 1958; Mar. 17, 23; Nov. 19, 1959

March 1957 to date.
Nov. 13, 1959

June 1945 to date.

March 1957 to date.

May 1945 to March 1956.

December 1935 to August 1942.
September 1942 to date.

October 1958 to date.

August 1945 to date.

Table 1. (Continued)

Station Drainage area
No. Location (square miles) Type of data Period of record

Reedy Creek near Loughman, Florida

Horse Creek near Davenport, Florida

Peace Creek drainage canal near
Dundee, Florida

Crystal Lake at Lakeland, Florida

Withlacoochee-Hillaborough overflow
near Richland, Florida

Hillaborough River at State Highway 39
near Zephyrhills, Florida

Crystal Springs near Zephyrhills,
Florida

Blackwater Creek near Knights,
Florida

Hilleborough River near Zephyrhills,
Florida

Cypress swamp in sec. 15, T.25 S.,
R. 25 E., near Polk City, Florida

Withlacoochee River near Eva, Florida

Cypress swamp in sec. 4, T.26 S.,
R. 25 E., near Polk City, Florida

- '"'117 ''

"'117

Daily record of stage and discharge

Chemical analysis of 1 water sample

Daily record of stage and discharge

Chemical analysis of 1 water sample

Daily record of stage and discharge
Discharge measurement at monthly
interval
Chemical analyses of 3 water samples

Chemical analysis of 1 water sample

Discharge measurement at hi-monthly
interval
Chemical analyses of 2 water samples

Daily record of stage and discharge

Daily record of stage and discharge

Chemical analysis of 1 water sample

Daily record of stage and discharge
Chemical analyses of 6 water samples

Chemical analysis of 1 water sample

May 4, 1959.

July 1958 to date.
August 19, Sept. 11,1958; Feb. 27, Mar. 18,
Mar. 24, Noy.13, 1959.

May 4, 1959.

Octobv. 1? 3~to September 1959.

November 13, 1959.

December 1946 to September 1959.

November 9, 1959.

February 1930 to September 1931.

July 1958 to date.
February 25, March 18, November 19,
1959.

November 9, 1959.

October 1934 to date.

November 9, 1959

January 1951 to date.

November 1939 to date.

Table 1. (Continued)

Station Drainage area
No, Location I (square miles) Type of data Period of record

Pony Creek near Berry, Florida
Grass Creek near Rock Ridge, Florida
Little Lake Agnes near Polk City,
Florida
Withlacoochee River near Rock Ridge,
Florida
Gator Creek at Foxtown, Florida
Withlacoochee River at Larkin's
Bridge near Richland, Florida
Florida

Pasco Packing Company canal at

Withlacoochee River at Cummer-
Florida
Cypress swamp in sec. 30, T. 24 S.,
R. 25 E., near Cumnpressco, Florida
Gator Hole Slough in sec. 34. T. 23 S.,
R.23 E., near Slaughter, Florida

Specific conductance, 1 observation
Chemical analysis of I water sample
Chemical analysis of I water sample

Specific conductance, I observation
Chemical analysis of 1 water sample
Specific conductance, 1 observation
Chemical analysis of I water sample

Daily record of stage and discharge
Discharge measurement at monthly
interval
Chemical analyses of 6 water samples

Discharge measurement at monthly
or bi-monthly interval
Chemical analyses of 2 water samples
Chemical analyses of 2 water samples

Chemical analyses of 2 water samples

Chemical analysis of I water sample

Sept. 11, 1958.
June 10, 1959.
Nov. 9, 1959.

Sept. II, 1958.
Nov. 12, 1959.
Feb. 26, 1959.
Nov. 12, 1959.

February 1930 to March 1933.
July 1958 to date.
August 18, Sept. 12, 1958; February 25,
March 19, March 23, November 12, 1959.
February 1957 to date.
Feb. 25, Nov. 12, 1959.
February 25, November 13, 1959.

U

E

Li

*1

ur

'4
Ln

November 18, 1959.

February 26, 1959.

Table I. (Continued)

Station Drainage are
No. ___ qusare miles) Type of data Period of record
36 Gator Hole Slough in sec, 25, T. 23 S., Chemical analysis of water sample February 26, 1959.
R. 22 E., near Slaughter, Florida

37 Tributary of Gator Hole Slough in Chemical analysis of 1 water sample February 26, 1959.
sec. 19, T.23 S., R, 22 E., near
Lacoochee, Florida

38 Withlacoochee River at State Highway Chemical analysis of 1 water sample February 26, 1959.
575, near Lacoochee, Florida

39 Withlacoochee River at Trilby, Florida 620 Daily record of stage and discharge August 1928 to February 1929.
February 1930 to date.
Chemical analyses of 3 water samples August 25, September 12, 1958; Novem-
ber 12, 1959%
40 Bay Lake near Bay Lake, Florida Chemical analysis of 1 water sample November 9, 1959.

41 Bay Root Slough in sec. 14, T. 23 S., Chemical analysis of 1 water sample February 26, 1959.
R. 22 E., near Bay Lake, Florida

42 Little Withlacoochee River at North Chemical analyses of 2 water samples February 26, 1959.

43 Little Withlacoochee River at Rerdell, 160 Daily record of stage and discharge July 1958 to date.
Florida Chemical analyses of 4 water samples August 18, September 12, 1958; March 19,
November 12, 1959.
44 Withlacoochee River at Croom, Florida 880 Daily record of stage and discharge November 1939 to date.
Chemical analysis of I water sample November 13, 1959.

Note: The unclosed dates under period of record indicate that the station was in operation in December 1959.
(a) Drainage basins for Little Creek and Withlacoochee River are interconnected.
(b) Drainage basins for Reedy and Shingle Creeks are interconnected above gaging station.

FLORIDA GEOLOGICAL SURVEY

General descriptions of the geology of the region have
been given by Cooke (1945), Vernon (1951), White (1958),
and Stewart (1959). Stringfield (1936) defined and described
the principal aquifer of Florida.

Analyses of water from surface and ground sources in
the vicinity of the area of investigation are given in reports
by Collins and Howard (1928) and Black and Brown (1951).

Methods of Investigation

Most of the data for this investigation were collected
during the 15-month period from July 1958 to September
1959. For a more complete and comprehensive investiga-
tion, the period of data collection should be long enough to
cover wide range of hydrologic conditions. However, much
can be learned from short-term records.

Surface-water characteristics of the area were deter-
mined by collecting stage, discharge, and chemical-quality
data at gaging stations and miscellaneous sites; by making
field and aerial reconnaissance of the area; and by studying
maps and aerial photographs.

Discharge and stage records were collected at five major
outlets from the Green Swamp area, at other sites within
the area, and in other stream basins adjoining the area of
investigation. At most of these stations, daily discharge
records were collected. At other sites, stage, discharge,
and chemical-quality data were collected at monthly or less
frequent intervals. The conclusions about the runoff char-
acteristics and the environmental factors that influence the
quality of surface water in the area are based on the data
collected at these stations and sites.

The physical features of the land surface of most of the
accessible areas were inspected by automobile travel and
on foot. Inspections were made of thelandforms, the drain-
age development, the interconnections of drainage basins,
the road fills and drainage structures, the type and density
of vegetation, and the surface geology. These features of
the land surface were studied to appraise their relation to
the hydrology.

INFORMATION CIRCULAR NO. 26

An aerial reconnaissance of the entire area from an
altitude of about 1,000 feet enabled the investigators to ob-
serve many features in areas that are not easily accessi-
ble by land travel. Stereophotographs in color of many of
the important features were made from the air to provide a
closer examination of these features.

Recent aerial photographs of the entire area and ad-
vance copies of topographic maps for much of the area were
used to define the drainage divides and to locate places where
the basins are interconnected.

Information onthe quality of surface water obtained dur-
ing high, intermediate, and low flows indicates the general
chemical characteristics and the extremes in mineral con-
tent during the period of study. Because of the limited scope
of the investigation the data on quality of water were col-
lected by reconnaissance of the area. Data were collected
overthe entire area generallywithin a period of 1 to 3 days.
Most of the data-collection points were visited at least twice
during the study. Information obtained in this manner was
used to determine the quality of water prevalent in the area
at a given time and also was used to considerable advantage
in understanding the interrelationships between water above
and below the land surface.

Test drilling was started in February 1959 to obtain in-
formation on the occurrence of artesian and nonartesian
ground water in the Green Swamp area. Five test wells
were drilled into the Floridan aquifer (principal artesian
aquifer) and were equipped with water-level recording in-
struments to record the fluctuations of the piezometric sur-
face. A shallow well was drilled intothe nonartesian aquifer
beside each of four of the deep wells. Water-level recorders
were installed on these wells to record the fluctuation of
the water table. Four of these installations were equipped
with standard 8-inch rain gages and tipping-bucket attach-
ments to record rainfall.

A second test-drilling program was started in July 1959
to provide additional geologic and hydrologic information.
The five existing test wells were deepened and five new test
wells were drilled into the Floridan aquifer.

FLORIDA GEOLOGICAL SURVEY

The wells are numbered on the basis of a state-wide
grid of 1-minute parallels of latitude and 1-minute merid-
ians of longitude. A well number is a composite of three
parts separated by hyphens. The first part of the number
is composed of the last digit of the degree and two digits of
the minute that identifies the latitude on the south side of
the 1-minute quadrangle in which the well is located. The
second part of the number is composed of the last -digit of
the degree and two digits of the minute that identifies the
longitude on the east side of the same 1-minute quadrangle.
The third part of the number indicates the order in which
the wells were inventoried in that quadrangle.

Test-well data are shown in table 2.

The wells were drilled by a churn drill, and during the
drilling a log was made and samples of rock cuttings were
collected. The well logs show changes in lithology with
depth, the occurrence of representative fossils in the vari-
ous formations, relative hardness of the rocks, drilling
speeds, and significant changes in water levels in the well.
Identification of the geologic formations was made in the
field and is considered to be tentative until verified by mi-
croscopic examinations of the well cuttings.

An inventory of selected wells was made to obtain in-
formation on the depth of the well, the amount of casing,
and the depth to static water level. Nearly all of the inven-
toried wells penetrated the Floridan aquifer. The approxi-
mate elevations above mean sea level at most wells were
determined by use of an altimeter. During the period October
to December 1959, water-level measurements were made
in wells to determine the elevation of the piezometric sur-
face. These measurements were used to prepare a piezo-
metric map (figure 22) which shows the direction and relative
movement of water in the Floridan aquifer.

Samples of ground water were collected during the test
drilling and analyzed to determine the chemical quality of
the water in various formations. Water samples were col-
lected periodically from the deep and the shallow observa-
tion wells to determine progressive changes in quality.

Table Z. Test Well Dta la the Green Swamp Are.

Totl depth I C.aIn 0 ol
below land Depth bolow kge
Date eurfae Diamer I and surface Diameter in depth
Well Number drilled (fee) (Inee)l (fee.t) Lche. (feet) Aquoer Rem-ark

UM Coot,
32-154-1 Feb. 1959 73 6 63 5 63-73 nlorlde
July 1969 160 t 73-160 Florida Deepened for geologic c trol.
319-14-2 Feb. 1959 21 6 1(6 t 16-a Ionatneellua Gravel pekedl (Umeatone pebblel.

82.-149-1 Feb. 19I9 95 6 I 59 5 9-95 Florid.
Juy 1959 190 6 100 5 95-292 FlorItdn Deepened for geologic control.

822-149-2 Feb. 1959 3 6 1i 5 18-21 NonurtestL Grvel pckejd (limstone pebbles).

Pol Coutar
81.157.2 July 1959 168 3 S2 2a 110-168 Florldan bditleg U.S. .S. weUdeepened for geologic co rol. B1S-19-3 Feb. 1959 90 6 7?s 5 7-90 Florldau July 1959 217 ft 90-17 Florida. Deepened tor eolQotL coulred 813-149-2 Feb. 1959 V7 6 10 St 20-b7 Nonarteialn Gravel mpaked (llietone pebbles). 81i-201-1 Ju3l 1959 20! 3 40 2i 40-255 Floridan 810.144.1 July 1959 349 6 101 )ot 101-49 Florida 816-.06-1 July 1959 200 3 41 a) 41.200 Florlda Inltor Ceni, 27.198-1 July 1959 175 3 99 It 99-175 Florida 821-156-1 July 1959 53 3 3 53 at Floridan Cler1 at 53 feet lmpletrable. Wall destroyed. I21.15--2 July 1959 49 3 49 21 o FIrlrid Cbert at 49 9et impanetrable. Beotom of caingl blhsted pep for uae s observatiom well. 821 02-3 r.eb. 1959 29 6 20 5f 20.29 Flridas July 9 99 143 2 29-143 Florldan Deepened for geolel eaOrol. 81a-202-4 Feb. 1939 Li 6 5 )f 5-12 Florildn Gravel packed (Umetone pebbles). 1182138-1 Feb. 1959 114 6 103 10 0-114 Floridn 12.-111-2 Fb. 1959 10 6 13 Si 13-30 Nourteela OG vel pald (limeaten pebble.l. FLORIDA GEOLOGICAL SURVEY Acknowledgments This report was prepared by the Water Resources Divi- sion of the U. S. Geological Survey in cooperation with the Florida Geological Survey and the Florida Department of Water Resources. The investigation was conducted and the report prepared by R. W. Pride, hydraulic engineer of the Branch of Surface Water and project leader; F. W. Meyer, geophysicist of the Branch of Ground Water; and R. N. Cherry, chemist of the Branch of Quality of Water. The authors wish to express their appreciation for the cooperation of the many residents andpublic officials of the area for information given during the well inventory and re- connaissance of the area. Special acknowledgment is due the Cummer Company, the Florida Forest Service, and the Florida State Road De- partment for granting permission to drill test wells. The following agencies made financial contributions for the col- lecting of data used in this report: Hillsborough County, Marion County, Pasco County, Polk County, Sumter County, Lake Apopka Recreation and Water Conservation Control Authority, Oklawaha Basin Recreation and Water Conserva- tion and Control Authority, Tsala Apopka Basin Recreation and Water Conservation Control Authority. The work on this project was done under the supervi- sion of the Florida Water Resources Division Council com- prised of A. O. Patterson, district engineer of the Branch of Surface Water M. I. Rorabaugh, district engineer of the Branch of Ground Water and J. W. Geurin, district chemist of the Branch of Quality of Water. DESCRIPTION OF THE AREA One of the most prominent topographic features in the central part of the Florida Peninsula is an extensive area of flatland and swamp at a relatively high elevation that is INFORMATION CIRCULAR NO. 26 calledGreenSwamp. Five major drainage systems originate in or near the GreenSwamp area and flow in several direc- tions to the sea. Figure 1 shows the location of the Green Swamp area and its relation totheheadwaters of streams in the central part of the Florida Peninsula. The area is the headwaters of the Oklawaha River, which flows generally northward to become the largest tributary of the St. Johns River; the Kissimmee and Peace rivers that flow southward; the Hillsborough River that flows southwestward; and the Withlacoochee River that flows northwestward. Green Swamp has many swamps and marshes, some of which are interconnected but many of which are separated by ridges, hills, and upland plains. Location The Green Swamp area is west of the center of central Florida on a high sandy ridge that forms the major axis of the peninsula. The boundaries of the area are not well es- tablished and may vary according to individual interpreta- tion. The small drainage basin that is generally known as Green Swamp Run lies in the headwaters of the Big Creek watershed in southern Lake County and northeastern Polk County. However, the boundaries of the Green Swamp area, as designated for this investigation, have been extended to encompass a much larger area. The project area (fig. 1, 7) includes the southern parts of Lake and Sumter counties, the northern part of Polk County, and the eastern parts of Pasco and Hernando counties. The eastern boundary of the Green Swamp area is U. S. Highway 27, from Clermont south-southeastward to Haines City. Mostly, this highway follows the top of a relatively high ridge that forms the divide for surface drainage be- tween the Big Creek and Reedy Creek basins. The southern and southwestern boundaries of the area are generally along the divides that separate drainage to the north into the Big Creek and Withlacoochee River basins from that drainage to the south into the Peace River and Hillsborough River basins. This boundary follows a mean- dering line from Haines City westward to Providence and then northwestward to Dade City. FLORIDA GEOLOGICAL SURVEY Map of Florida showing location of Green Swamp area. Figure 1. INFORMATION CIRCULAR NO. 26 The western boundary of the area is U. S. Highway 301 from Dade City to. St. Catherine. The Withlacoochee and Little Withlacoochee rivers, which drain the greater part of the GreenSwamp area, cross this boundary and converge into one channel about 3 miles to the west. The northernboundary extends from St. Catherine east- ward along the Little Withlacoochee River basin divide to State Highway 50 and along State Highway 50 eastward to Clermont. The boundaries described enclose an area of 880 square miles. Topography The Green Swamp area is in the Central Highlands to- pographic region as defined by Cooke (1945, p. 8). The area is almost surrounded by high sandy ridges, forming a basin that is roughly quadrilateral in shape, breached onthe north by the Palatlakaha Creek, Little Withlacoochee River, and Withlacoochee River channels and on the southwest by the diversionary channel from the Withlacoochee River to the Hillsborough River basin. The area is bordered on the eastern side by the Lake Wales Ridge, on the southern side by the northern terminus of the Winter Haven and Lakeland ridges, and on the western side by the Brooksville Ridge (White, 1958, p. 9-11). These ridges form natural surface-water drainage divides. Although designated the Green Swamp area, the entire area is not a continuous expanse of swamp but is a compos- ite of many swamp, sloughs, and sinkholes that are distri- buted fairly uniformly within the area. Ifiterspersed among the swamps are ridges, hills, and flatlands that are several feet higher than the surrounding marsh. Several large and many small lakes rim the area. These are most numerous in the southeastern and northeastern parts of the area. The elevation of the land surface varies from about 200 feet above mean sea level in the eastern part to about 75 FLORIDA GEOLOGICAL SURVEY feet in the river valleys in the western part. The interior of the area is a broad flat marsh with only slight relief. State Highway 33 crosses the area in a northerly di- rection from Polk City to Groveland. A few miles to the west, the Seaboard Air Line Railroad extends in a north- northwesterly direction through the interior of the area. This railroad divides the area into two approximately equal parts. The alignment of the uplands swamp in the eastern part differs from that of the western part. A prominent topographic feature affecting the drainage of the eastern part of the area is the alternate pattern of low ridges and swales that extends in a generally N. 15* W. direction from the southern boundary to the Polk-Lake county line. The ridges parallel the major axis of the Florida Pen- insula and their configuration suggests a shoreline of a form- er marine terrace, although the swales between the ridges could have been formed by stream erosion. Aerial photo- graphs of the area between U. S. Highway 27 and the Sea- board Air Line Railroad show five of these long narrow ridges with intervening swales. The uniform pattern of elongated ridges and swamps in the eastern area is brokenby saddles inthe ridges and cross connection of the swamps in many places. Some of these saddles may have been formed by partial dismemberment of the ridges due to the collapse of underground solution channels or by erosion. Many of the saddles have been lowered to the level of the swamps between the ridges and connect the adjacent drainage basins. In the western part of the Green Swamp area there is little evidence of the elongated ridges, and the main land- surface features are large swamps, flatlands, and rolling hills. There are many small swamps in patches and strips, generally less than half a mile across, in the area. Most of these swamps support good growths of cypress trees while in the uplands pine and scrub oak trees grow abundantly. In the fringe areas pine and cypress growths intermingle in an irregular fashion. The largest continuous expanse of swampland lies within the valley of the Withlacoochee River INFORMATION CIRCULAR NO. 26 and is more than a mile wide at places. Exposures of lime- stone are found from Rock Ridge in Polk County, northwest- ward to Richloam in Hernando County. Drainage Because of the flat land and poorly developed stream channels, the surface drainage of the Green Swamp area is sluggish. Following heavy rainfall the water stands in large shallow sheets over much of the area, which delays the time of concentration in streams. The water is removed from the land surface by the combined effect of four processes. Part is removed by evap- oration from the increased surface area of water retained temporarily in storage on the land surface, and part is lost through transpiration by the cypress trees and other luxu- riant vegetation. Another part of the water seeps downward into the sandy soil or percolates into the underlying porous limestone. The part of the water that remains on the land surface eventually collects inthe stream channels and drains from the area. Of the four processes by which water is removed from the land surface, the only one that can be measured directly is the removal of water by streamflow. Streamflow measured at a gaging station is the surface flow from the basin above that point together with ground-water inflow to the stream. Surface drainage from most of the Green Swamp area is generally toward the north and west. In contrast to this pattern, however, the headwaters of the Peace River basin originate along the southern limits of the area and flow is generally southward. Alongthe easternboundary of the area drainage is toward the east and southeast into the headwaters of Reedy Creek, a tributary of the Kissimmee River. Other drainage from the Green Swamp area is toward the south- west into the Hillsborough River through a natural diver- sionary channel through the watershed boundary in eastern Pasco County. FLORIDA GEOLOGICAL SURVEY Culture and Development The GreenSwamp area is sparsely populated except for the few small towns and communities on the ridges along the border and along State Highway 33. There are also a few ranchand farmhomes in rather isolated locations scat- tered through the area. Most of the land is owned in large tracts by private in- dividuals or corporations. The only large area of public land in the area is the Withlacoochee State Forest, part of which is within the boundaries of the Green Swamp area in Sumter, Hernando, and Pasco counties. The principal industry is agriculture. Much of the up- land area has been cleared and planted in citrus groves. Other upland areas have been cleared, planted in pasture, and are used for cattle raising. Very little of the land is cultivated. Because of poor drainage the low swampland is unsuitable for agriculture. Drainage of the swamps has been attempted by the digging of ditches and canals connect- ing many of the isolated swamps and sinkholes to the natural stream systems. In spite of the many miles of these ditches, water still stands in the low pockets. Even in the cleared areas that are suitable for agricultural uses, no attempts have been made to reclaim the many small, round cypress swamps that dot the area. Lumbering is an important industry in the western part of the area, particularly in the Withlacoochee River Swamp where there are extensive growths of cypress trees. The first access roads to penetrate the interior of the swamp were trails and tram roads built for cypress lumbering op- erations. Considerable lumber is produced also by pine flat- woods which are interspersed withthe swamps and marshes. Other native vegetation of the area consists of saw palmetto, scrub oak, and occasional hammocks of cabbage palm trees. The southern boundary of the Green Swamp area lies adjacent to extensive phosphate deposit in Polk County. Some phosphate is mined within the area but the amount is only a small percentage of that produced in southern Polk and western Hillsborough counties. INFORMATION CIRCULAR NO. 26 Deposits of sand, suitable for building uses, are found in many places on the eastern ridges, and the quarrying of this sand for the commercial market contributes to the local economy. Climate The Green Swamp area is located near the center of the Florida Peninsula a few miles nearer the Gulf of Mexico than the Atlantic Ocean. This geographical position, well south in the Temperate Zone, and the proximity to large bodies of subtropical water produce a warm subhumid climate. Precipitation and temperature, the principal climatic elements that influence the hydrology of the Green Swamp area, are described separately. Precipitation The study of precipitation in central Florida can be restricted to rainfall only, because snow and hail are vir- tually unknown. Most of the records of rainfall collected in the Green Swamp area are published monthly by the U. S. Weather Bureau in "Climatological Data, Florida. " The normal or long-term average annual rainfall of the Green Swamp area is 52. 6 inches. This normal is based on the unweighted mean of the annual rainfall record for 34 or more years at each of the following U. S. Weather Bureau stations: Brooksville Chinsegut Hill, Clermont 6 miles south, Isleworth, Kissimmee, Lake Alfred Experiment Sta- tion, Lakeland, and St. Leo (fig. 4). These precipitation stations are located uniformly around the boundary of the Green Swamp area except for the station 6 miles south of Clermont, which is the only one within the area. The un- weighted mean of the rainfall values at these seven stations is assumed to represent the average rainfall on the project area. The rainfall records were checked for consistency by plotting the cumulative rainfall at each station against the mean cumulative rainfall for the seven stations. The FLORIDA GEOLOGICAL SURVEY tests showed that the records are consistent and that there has been no change in catch because of moving of rain gages or other reason. The average rainfall for the stations at Brooksville and St. Leo, both west of the area, is slightly higher than that for the other five stations which are located farther inland. The average rainfall at the seven stations ranges from a minimum of 49.9 inches at the Clermont station to a maxi- mum of 56. 6 inches at the Brooksville station. In view of the small deviation of these extreme values from the mean, the average figure of 52.6 inches of rainfall for the area of investigation appears to be reasonably accurate. About 60 percent of the annual total rainfall occurs dur- ing the wet season from June through September. In the spring and early summer, thunderstorms of local high in- tensity and short duration sweep over the area. Showers occur almost daily, or perhaps several times a day, during June and July. Heavier and more prolonged rainfalls occur generally from August through September and are often in- tensified by tropical storms that occasionally reach hurri- cane proportions. Onthe other hand, therehavebeen periods of a month or more with little or no rainfall. These periods of below average rainfall usually occur during the winter season from November to February. The average monthly rainfall of the Green Swamp area, based on long-term rec- ords of the seven U. S. Weather Bureau stations, is shown by bar graph in figure 2. The amount of rainfall in the area varies yearly as well as seasonally. During wet years the annual rainfall aver- ages about one and one-half times that of the dry years. The average annualrainfall of the seven representative sta- tions during 1931-59 is shown by bar graph in figure 3. During the period of this investigation, July 1958 to September 1959, daily records were collected at 11 addi- tional rain gages (fig. 4), some of which were located in the interior of the project area. The denser network of rain gages closer to the project area gives a better indica- tion of the amount of rainfall on the area during the period of investigationthan would the seven stations used for deter- mining the long-term average. INFORMATION CIRCULAR NO. 26 10 9 8 7 .E O4 2 0 SOURCE, U.S. WEATHER BUREAU DATA FOR STATION AT BROOKSVILLE, CLERMONT, ISLEWORTH, KISSIMMEE, LAKE ALFRED, LAKELAND AND ST. LED. Average monthly rainfall of Green Swamp area. m O CALENDAR YEAR Figure 3. Average annual rainfall SOURCE'. u. war 1UCR URAu OATA FOR rTATmIOS AT SIRORIs.L, CLLAUORT ILEWIATH, MKISIlr. LAXR AlFRED, LALUARO AM of LEO.Green Swamp area of Green Swamp area. Based on records 1931-59 Fe-1 Average 4.4"- 4"---~J Ja F'MrA MIJ a A S6 0N D Figure 2. FLORIDA GEOLOGICAL SURVEY The total rainfall computed for the 1959 water year for each of the 18 rain gages in or near the area was used to prepare the isohyetal map shown in figure 4. The locations of the rain gages and the yearly total rainfall, in inches, at each of these gages are shown on the map. Lines of equal rainfall were drawn by interpolating between the measured rainfall at each station. Visual inspection of the isohyetal map in figure 4 indicates a total rainfall of the GreenSwamp area of about 72. 5 inches for the year, distributed fairly uniformly over the area. Thus during the 1959 water year the rainfall was 20 inches above normal. The rainfall records used for the preparation of the isohyetal map were from standard 8-inch nonrecording gages or fromtipping-bucket attachments to water-stage recorders at stream-gaging stations. Most of these rainfall records were collected by the U. S. Weather Bureau. However, at three sites within the area (Big Creek near Clermont, With- lacoochee River near Eva, and Dunham Ranch) rainfall rec- ords were collected by the U. S. Geological Survey as part of the basic data for this investigation. Standard 8-inch gages were used at these project stations so that the records are comparable with those collected at the official U. S. Weather Bureau stations. Rainfall data obtained from graduated glass test-tube gages were collected at each of the Florida Forest Service firetowers and at the Cummer and Townsend ranches within the area. The cumulative yearly rainfall totals as deter- mined from each of these gages were higher thanthose from the standard stations by amounts ranging from 20 to 28 inches. The difference is too great to be attributed to the vagaries of distribution as some of the tube gages are lo- cated only a few miles from standard stations. On the basis of this comparison and because the records from the test- tube gages are consistently higher than those from the offi- cial standard gages, they are assumed to be in error and the yearly amounts have been disregarded in drawing the isohyetal lines shown in figure 4. 820' 20' 2900' I I a C* c0 *0 I O , mi' r. 0n (D 3 ? (1 a! Id 82*00' 81*20 M 29,W 20 10' BZ'00' 50 40' LEGEND SCALE -*-* Boendlry of Gr1.m Stomp aral 9 0 5 Il Roinlol reasurIig ,laion .,h 10cor.t* record o flofZill 0o0ainq t0lolw inth stl-th'm wrWd orist1l Ilrw thu Inmitiw~ to ?i2. Annual rmflroil at uai., n r hes _ 73- hIabyf141 Ie, .a inchli 28600' 2800' FLORIDA GEOLOGICAL SURVEY Acknowledgments This report was prepared by the Water Resources Divi- sion of the U. S. Geological Survey in cooperation with the Florida Geological Survey and the Florida Department of Water Resources. The investigation was conducted and the report prepared by R. W. Pride, hydraulic engineer of the Branch of Surface Water and project leader; F. W. Meyer, geophysicist of the Branch of Ground Water; and R. N. Cherry, chemist of the Branch of Quality of Water. The authors wish to express their appreciation for the cooperation of the many residents andpublic officials of the area for information given during the well inventory and re- connaissance of the area. Special acknowledgment is due the Cummer Company, the Florida Forest Service, and the Florida State Road De- partment for granting permission to drill test wells. The following agencies made financial contributions for the col- lecting of data used in this report: Hillsborough County, Marion County, Pasco County, Polk County, Sumter County, Lake Apopka Recreation and Water Conservation Control Authority, Oklawaha Basin Recreation and Water Conserva- tion and Control Authority, Tsala Apopka Basin Recreation and Water Conservation Control Authority. The work on this project was done under the supervi- sion of the Florida Water Resources Division Council com- prised of A. O. Patterson, district engineer of the Branch of Surface Water M. I. Rorabaugh, district engineer of the Branch of Ground Water and J. W. Geurin, district chemist of the Branch of Quality of Water. DESCRIPTION OF THE AREA One of the most prominent topographic features in the central part of the Florida Peninsula is an extensive area of flatland and swamp at a relatively high elevation that is INFORMATION CIRCULAR NO. 26 Temperature Knowledge of temperature variations in central Florida is pertinent to a study of its water resources because of the dominant influence of temperature on rates of water losses by evaporation and transpiration. Water loss from a drainage basin is the difference be- tweenthe average rainfall over the basin and the runoff from the basin for a given period (Williams, 1940, p. 3). In humid regions, where generally there is sufficient water to satisfy the demands of vegetation, the mean annual water loss is principally a function of temperature (Langbein, 1949, p. 7). The relation between mean annual temperature and mean annual water loss under such conditions is shown in figure 5, which is taken from U. S. Geological Survey Cir- cular 52. For the Green Swamp area where the mean annual temperature is 7Z2F. the annual water loss would be 48 inches according to this figure. 80 70 S 60 E 50 0 a S40 5 o 4 30 0 Natural water loss, Figure 5. 30 40 U50 U in inches Relation of annual water loss in humid areas to temperature. __ (From Longbein, W. B, and others 1949) ___ ^~~~~ ______ FLORIDA GEOLOGICAL SURVEY The mean monthly temperature in the Green Swamp area ranges from 61 F. for January to 820 F. for August. The lowest temperature recorded during the 66-year period of record at the Clermont station was 18* F. and the high- est was 104'F. Daily temperatures recorded at the U. S. Weather Bureau stations show that all parts of the area are essentially at the same temperature, varying no more than Z to 30 F. Killing frost occurs infrequently in this area, and damage to vegetation, although severe from the standpoint of agri- culture, seldom is great enough to affect the hydrologic factors pertinent to water supplies. Geology1 The Green Swamp area is underlain by several hundred feet of marine sediments, composed primarily of limestone, that have been periodically exposed to erosion. The lime- stone is mantled with a varying thickness of plastic deposits, chiefly sand and clay, that were deposited in fluctuating shallow seas. No attempt has been made to separate for- mations within the plastic material because of lack of data. Topographically, the surface of the area resembled a structural basin, or trough, opening to the north. However, test-drilling data show that the Green Swamp area overlies part of an eroded, faulted anticline. The oldest formations are exposed along the axis of the fold and eroded remnants of younger formations rim the flanks of the uplift, present- ing a basin-like shape. The classification and nomenclature of the rock units conform to the usage of the Florida Geological Survey and also, except for the Tampa formation, and the Ocala group and its subdivisions, with those of the U. S. Geological Survey which regards the Tampa as the'Tampa limestone and the Ocala group as two formations, the Ocala limestone and the Inglis limestone. INFORMATION CIRCULAR NO. 26 The porous limestone formations underlying the Green Swamp area are a part of the principal artesian aquifer in the State. The principal artesian aquifer was first described by Stringfield (1936) and later named the Floridan aquifer by Parker (1955, p. 189). According to Parker, the Floridan aquifer includes those limestone formations ranging in age from the middle Eocene (Lake City limestone) through early Miocene (Tampa formation, and the permeable zones of the middle Miocene (Hawthorn formation) that are in hydraulic contact with the rest of the aquifer. Test drilling in the Green Swamp area penetrated the following formations of the Floridan aquifer (from youngest to oldest); the Suwannee limestone of Oligocene age; the Ocala group which includes (from youngest to oldest) the Crystal River, Williston, and Inglis formations, and the Avon Park limestone of Eocene age. The Floridan aquifer is overlain by a varying thickness of plastic deposits. Generally, these plastic sediments are thickest under the topographic highs. The upper part of the plastic deposits forms a distinct hydrologic unit, commonly referred to as the nonartesian (unconfined) aquifer. The basal portion of the plastic deposits is composed mostly of clay and is less permeable than both the overlying, satu- rated, clayey sands and the underlying Floridan aquifer. Where present, the clay retards the rate of water movement between the two aquifers. The lithology was described and the formations were identified during the drilling of the test wells. Drilling logs of the test holes were supplemented by electric and gamma- ray logs. of wells 828-154-2 and 825-151-1 (Lake County), driller's logs of wells 834-153-1 and 822-140-1 (Lake Coun- ty), and a log by E. R. Applin (from Florida Geological Survey file of well 813-210-1 (Pasco County). Formations Generalized cross sections (fig. 6) were constructed between wells along line A-A' from southwest to northeast, along line B-B' from west to east, and along line C-C' from south to north across the area (see sketch map in fig. 6). FLORIDA GEOLOGICAL SURVEY Figure 6. Generalized geologic cross sections across the Green Swamp area and their locations. INFORMATION CIRCULAR NO. 26 The contacts of the formations have been projected between widely separated wells. Additional wells would improve the definition of the features. Two fault zones have been infer- red on the cross sections. They probably represent only a few of those that exist on the major structure (anticline) within the Green Swamp area. The inferred regional dips on cross sections C-C' and B-B' show that faults probably occur between wells 822-149-1 and 825-151-1; and between wells 822-140-1 and 822-138-1. The fault between the latter two wells is located in the vi- cinity of the Kissimmee faulted flexure (Vernon, 1951, p. 56). The fault between the former two wells is highly interpre- tive but, considering thefracturing characteristics of lime- stone, it seems more probable that the beds are faulted rather than bent. The existence of many rectangular drain- age features tends to indicate fracturing. Faulting here is further supported by the occurrence of fractures and faults in Citrus and Levy counties, a few miles northwest of the Green Swamp area. Avon Park Limestone: The Avon Park limestone (Applin and Applin, 1944,p. 1686) of late middle Eocene age was the deepest formation penetrated bytestwells. Theformation is near the surface on an upthrown block in southwest Orange County and on the crest of the anticline in eastern Pasco County and in southern Lake and Sumter counties (see cross section B-B' in fig. 6). The formation is at considerable depth in the area and southwest of Green Swamp. The top of the formation was identified in the field by a distinct color change from tan to brown limestone and by abundant "cone- type" foraminifer.s. However, a more recent observation indicates that perhaps the top of the formation is 20 to 30 feet below the first "cone" zone at an unconformity probably at the occurrence of a coral reef. The formation is characteristically a brown dolomitic, fossiliferous limestone. It is highly permeable and is the best source of water for most of the high-capacity wells in the area. Ocala Group: The Ocala group (Puri, 1957) consists of limestone formations of late Eocene age which have been FLORIDA GEOLOGICAL SURVEY separated on the basis of lithology and fossils. Test drill- ing indicated that localunconformities may occur within the group inthe GreenSwamp area, but the general stratigraphic sequence was recognized. The subdivisions of the Ocala group are (in ascending order) the Inglis, Williston, and Crystal River formations. The Williston formation was not easily recognized by field examination of the well cuttings; therefore, it was included with the Crystal River formation on the geologic sections (fig. 6). The Inglis formation is a hard, white to tan fossilifer- ous limestone containing small amounts of clay probably as fill. The texture of the formation is finer than that of the CrystalRiver and Williston formations because of dolomiti- zation. The formation contains fragments of an Eocene echinoid, Periarchus lyelli. The formation is about 50 feet thick except on the fault block located on the east side of the Lake Wales ridge where it is thin (see section B-B' in fig. 6). There are some indications that the Inglis is thin- ner over the central part of the area (see section A-A' in fig. 6). The Williston formation is a medium to hard, tan to cream limestone containing abundant fossils. The rock cuttings are slightly coarser than those from theInglis for- mation but lack the large foraminifer Lepidocyclina ocalana that is characteristic of the Crystal River formation. The Williston ranges from 10 and 20 feet in thickness in some wells but is thin or absent in others. The Crystal River formation consists primarily of a coquina of large foraminifers with some clay fill. The Crystal River contains local zones of chert where it is ex- posed or occurs near the surface in an area extending from Rock Ridge in northern Polk County, through Cumpressco at the southern end of Sumter County, to Slaughter in Pasco County. The formation is generally 50 to 100 feet thick but is absent near southwestern Orange County. The Crystal River formation is overlain unconformably by the Suwannee limestone, Tampa formation, or undiffer- entiated young plastic deposits. The Crystal River contains INFORMATION CIRCULAR NO. Z6 cavity fill, composed largely of the basal part of the over- lying undifferentiated plastic deposits, in the area from State Highway 33 to U. S. Highway 27. Suwannee Limestone: The Suwannee limestone of Oligo- cene age (Cooke, 1945, p. 86) is a white, dense, fossiliferous limestone that is present inthe southern andwestern parts of the Green Swamp area and absent in the remainder. The Suwannee limestone crops out along the Withlacoochee River near the junction of Polk, Pasco, and Sumter counties. Southwestward in Pasco County the formation thickens rap- idly. Apparently many of the springs along the upper Hills- borough River flow from exposures of Suwannee limestone or the overlying Tampa formation. The Suwannee limestone is unconformably overlain by either undifferentiated plastic deposits or the Tampa formation. Tampa Formation: The Tampa formation of early Mio- ceneage (Cooke, 1945, p. 11) is a white, sandy, fossiliferous limestone containing abundant mollusks that reportedly occur at or near the surface in southeastern Pasco County and along State Highway 50 in southern Sumter County. The Tampa formation was not recognizedin any of thetest holes drilled in the area. The Tampa formation is overlain un- conformably by undifferentiated plastic deposits. Hawthorn Formation: The Hawthorn formation (Cooke, 1945, p. 144) of middle Miocene age consists chiefly of inter- bedded sandy to silty, montmorillonitic clay and white to gray, sandy phosphatic limestone. About 35 feet of light green to gray clay with interbedded phosphatic sandstone was present in well 810-144-1 in Polk County. It did not occur in any of the other test holes. Where present, the Hawthorn forma- tion generally retards the movement of water between the nonartesian and Floridan aquifers. The Hawthorn formation as used in this report is not differentiated from the other plastic deposits. Undifferentiated Sand, Clay, and Sandstone: Undiffer- entiated plastic deposits, ranging from Miocene to Recent in age, cover the entire Green Swamp area except in those areas where the Crystal River formation, Suwannee lime- stone, or Tampa formation is exposed. FLORIDA GEOLOGICAL SURVEY The plastic deposits consist primarily of quartz sand with varying amounts of clay, phosphate sand, and calcareous sandstone. The following general stratigraphic sequence was indicated (oldest to youngest): (1) dark gray to green to blue, phosphoritic clay; (2) light tan to gray to blue, mont- morillonitic clay; (3) light green to gray clay and interbed- ded phosphatic sandstone (Hawthorn formation); (4) fine to coarse, quartz sand and varying amounts of white kaolinitic clay; and (5) fine quartz sandwith varying amounts of organic detritus or clay. The dark phosphoritic clay ranges froml0 to 25 feet in thickness. It was generally present in the area east and south of the limestone plain at Rock Ridge in Polk County. Similar clay containing more sand occurs as cavity fill within the Crystal River formation. The dark clay may be part of the Tampa formation of early Miocene age, or may repre- sent the basal part of the Hawthorn formation of middle Miocene age. A dark blue clay was present below the Haw- thorn formation in well 810-144-1 in Polk County. The clay retards movement of water between the nonartesian aquifer and the Floridan aquifer. Clay, resembling the clays in the Hawthorn formation, occurs above the Suwannee limestone and Crystal River for- mations in the western part of the area but becomes pro- gressively more sandy toward the eastern part. Its age and stratigraphic position is questionable, but the clay is a con- fining bed in the western part of the area. Sand containing some variegated white to red to tan, kaolinitic clay underlies the entire Green Swamp area except where the Crystal River formation, Suwannee limestone, or Tampa formation is exposed. The deposit is about 100 feet thick under the ridgesthat rimthe area andis thin or absent over the limestone plain in the vicinity of Rock Ridge. The sand has a higher permeability than that of the underlying clay and constitutes part of the nonartesian aquifer. The geologic age and origin of the sand deposits are questionable. Surficial sand ranging from 0 to 10 feet in thickness overlies the clayey sand deposit and is the upper part of the nonartesian aquifer. INFORMATION CIRCULAR NO. 26 Structure The hydrology, stratigraphy, and topography of the Green Swamp area are directly related to structure formed by the Ocala uplift and indirectly to the Peninsular Arch. The Peninsular Arch (Applin, 1951, p. 3), a buried anticlinal structure of Paleozoic sediments, trends generally north- northwestward and its crest is located east of the Green Swamp area. Aflexure that probably developed on the west- ern flank of the Peninsular Arch in the Tertiary limestones is called the "Ocala uplift. The Green Swamp area is lo- cated at the southern end of the Ocala uplift. Vernon (1951, p. 54) dated the deformation of the Ocala uplift as post-Oligocene in age. However, the structure in- ferred in constructingthe cross sections (fig. 6) suggests that uplift and faulting mayhave occurred over a longer period of time, possibly extending from late Eocene to Pleistocene. The occurrence of apparent local unconformities within the Eocene section, such as the unconformity between the Avon Park limestone and the Ocala group, and lithologic changes and unconformities within the Ocala group, might reflect concurrent uplift and deposition. The cross sections indi- cate a slight thinning of the Inglis over the limestone plain near Cumpressco in Sumter County. This change in thick- ness could be caused either by error in selecting the top or bottom of the formation or by thinning of the Inglis over the area during active uplift. A general thinning of the Ocala group occurs northward and the Suwannee limestone was entirely absent. This sug- gests erosion or nondeposition over a structural high. Faults complicate the definition of the geologic history. No conclusions have been reached as to the number or age of the faults that occur in the Green Swamp area. The faults shown on the cross sections have been inferred from a gen- eral interpretation of the structure. The linear topographic features and drainage patterns show that the area probably is highly fractured. The vertical displacement along the faults probably ranges from a few feet to about 100 feet. FLORIDA GEOLOGICAL SURVEY Structure probably could affect the hydrology of the Green Swamp in the following ways: (1) The joints or fault zones have been widened by solution, causing zones of high perme- ability within the Floridan aquifer. These zones appear as troughs in the piezometric surface of the Floridan aquifer. (2) The displacement along the faults could posi- tion formations of different lithology (hence permeability) one against the other, break- ing the hydraulic continuity. (3) Faults cutting confining beds increase or decrease ground-water circulation between aquifers, depending upon the displacement. HYDROLOGY General The endless circulation of water by evaporation, move- ment through the atmosphere, precipitation, and movement to the sea by surface and underground routes is known as the hydrologic cycle. Because of this natural circulation, water may be regarded as a transient but renewable re- source. Although the amount of water involved in the hy- drologic cycle remains about constant for the earth as a whole, the amount available at a particular place varies with precipitation. The water supply of the earth, whether it is on the sur- face or below the ground, has its origin in precipitation. Of the precipitation that reaches the ground, a portion is returned to the atmosphere by evapotranspiration, a por- tion remains above ground and is stored temporarily in lakes, ponds, and swamps, or moves to the sea as streamflow, and a portion filters into the ground, some to replenish the soil moisture and some to enter the zone of saturation and INFORMATION CIRCULAR NO. 26 recharge the ground-water supply. Ground water moves laterally under the influence of gravity, toward areas of discharge such as wells, springs, streams, lakes, or the ocean. Much of the dry-weather flow of streams is derived from ground-water discharge. Surface Water Occurrence and Movement The occurrence and availability of water on the land surface are determined largely by topography, climate, and geology. As there is no inflow of water from outside of the Green Swamp area, the only source of water is from rain- fall. Inthe previous discussion of precipitation, it was stated that the average rainfall for the area is 52. 6 inches. Much of this total rainfall is stored temporarily on the land sur- face inthe many natural reservoirs such as swamps, lakes, and other topographic depressions. This temporary storage of water during wet seasons increases the area of water surface and, thus, increases the opportunity for evaporation. The retention of water ontheland surface likewise increases the opportunity for percolation to recharge the aquifer. In addition, a considerable amount of water is transpired by the cypress trees and other aquatic plants which grow lux- uriantly in the area. Thus, evaporation, transpiration, and percolation remove much of the water while it is stored temporarily on the land surface. The topographic, climatic, and geologic influences that contribute to this loss of surface water also tend to reduce the magnitude of the flood peaks and to sustain the base flow of streams during dry periods. Characteristics of Drainage Basins The headwaters of five stream systems lie in the Green Swamp area. These streams, listed in order of the propor- tion of the area drained, are: Withlacoochee River, Little Withlacoochee River, Palatlakaha Creek, Hillsborough River, FLORIDA GEOLOGICAL SURVEY and Reedy Creek. Palatlakaha Creek drains to the St. Johns River and Reedy Creek to the Kissimmee River. Other streams that head near the boundaries of the Green Swamp area are: Davenport and Horse Creek in the Kissimmee River basin, Peace Creek drainage canal and Saddle Creek in the Peace River basin, Fox Branch in the Hillsborough River basin, and Jumper Creek and a major canal that head northwest of Mascotte in the Withlacoochee River basin. The drainage system of the Green Swamp area and vi- cinity is shown on the map in figure 7. Of the total area of 880 square miles, 720 square miles are drained by the Withlacoochee River and its tributaries. Withlacoochee River: The headwaters of the Withla- coochee River are a group of lakes and swamps in the north- central part of Polk County in the vicinity of the towns of Polk City and Lake Alfred. Lakes Van and Juliana, the uppermost of these headwater lakes, drain into Lake Mattie. Surface drainage from Lake Mattie spills through a wide shallow marsh along the northeastern shoreline and flows northward through a series of interconnected shallow swamps to the northern boundary of Polk County, where it enters Withlacoochee River, which at that point is poorly defined. Other headwater tributaries of the Withlacoochee River originate in the marshes between Lake Mattie and Lake Lowery and flow generally northward between the confining sand ridges. These sand ridges form elongated north-south drainage basins in the eastern part of the Withlacoochee River basin, the upper Palatlakaha Creek basin, and in other areas of the central peninsula. The valleys between the ridges are not deeply incised by natural processes but their effectiveness as drainage channels has been improved by the many miles of canals and ditches that have been dug. Some parallel drainage basins are interconnected in several places by gaps or saddles through the ridges. Through these gaps water may flow from one stream valley into another. The amount and direction of flow depend on the relative elevation of water levels in the adjoining basins and the.hydraulic con- veyance characteristics of the connecting channels. D.-p do la, C U lmp s dimn i l tur tlhItIUm ^il__ flfai.qo i htod Sfar ftti

S uon. m Ofr lim. pow

S t.mr. _0n .e60r 11W-i0
T- CC-C W"kp 4.1

A SI'..ll. .~ ON" H T lbd I

ci.tnwM iy V te toi, I Stwt, mSC
Pf tfHN fi I l fl t*W/ Ldo t:
Pr".np kn FridD T-Wf fuMI

I______ _______ ___ __~ __ _

INFORMATION CIRCULAR NO. 26

West of the Seaboard Air Line Railroad the tributaries
of the Withlacoochee River are not confined by the ridges
that are prominent in the area east of the railroad. These
tributaries have developed more generally fan- shaped basins
than those in the eastern part. The Withlacoochee River
follows a meandering course, generally in a southwesterly
direction, to the point of diffluence to the Hillsborough
River. Between the railroad and this point of diffluence,
the main tributaries of the Withlacoochee head in the south-
ern part of the area and flow in a northwesterly direction.

Many miles of canals and ditches have been dug within
the basins of these tributaries in an effort to improve the
effectiveness of the drainage system. These canals and
ditches, for the most part, have been dug tofollow the natu-
ral drainage courses throughthe shallow swamps. However,
in some places, probably to provide firm footing for the
excavation equipment andto avoid clearing through the dense
growth of cypress trees, the ditches have been dug along
the rim of the large swamps rather than through the interi-
or. Also to provide better alignment in some places the
ditches have been cut through ridges to connect the adjacent
swamps. These shortcuts havebypassed the circuitous nat-
ural drainage routes through swamps and, in general, have
straightened and shortened the entire course of the waterway.

The first of the larger tributaries west of the Seaboard
State Highway 33 near Polk City. This stream flows in a
northwesterly direction to the Withlacoochee River.

Grass Creek is the next large tributary and empties
into the Withlacoochee River about 1 miles downstream
from Pony Creek. Grass Creek heads in a group of small
lakes in the vicinity of Polk City. These lakes are Little
Lake Agnes, Lake Agnes, Clearwater Lake, and Mud Lake.
Little Lake Agnes is connected to Lake Agnes by a marsh.
The outlet from Lake Agnes is a ditch leading from the
northern end of the lake and connecting with the network of
canals and ditches that carry the water through the swamp
in a northwesterly direction. During high water Clearwater
Lakehas a surface outlet on its southern shorethat connects

FLORIDA GEOLOGICAL SURVEY

with Mud Lake. The outlet from Mud Lake is a ditch from
the northeastern end of the lake, connecting with the main
ditch from Lake Agnes about 1 mile downstream. Several
other tributaries flow into Grass Creek as it crosses the
swamp.

Gator Creek empties into the Withlacoochee River at
the Polk-Pasco county line. This is the largest tributary
upstream from the diffluence of the Withlacoochee River to
the Hillsborough River. Gator Creek heads in several small
swamps northeast of Lakeland and flows northwestward
through a network of swamp channels and ditches. The total
drainage area of the Gator Creek basin is 92 square miles.

From the point of diffluence to the Hillsborough River,
the channel of the Withlacoochee River turns abruptly to the
north in a wide sweeping curve and continues in a north-
westerly direction to the western boundary of the Green
Swamp area at U.S. Highway 301.

About 12 miles downstream from the point of diffluence,
a major canal draining several lakes and swamps east of
Dade City empties into the river from the west bank. This
canal receives also the drainage from an area of hills and
lakes west of Dade City and the effluent from citrus concen-
trate plants at Dade City. The water supply for the plants
is obtained from wells.

One of the larger tributaries entering the Withlacoochee
River from the east is formed by the confluence of Devil
Creek and Gator Hole Slough. Devil Creek heads in a swamp
about 2- miles east of the abandoned lumbering camp of
Cumpressco. The drainage divide between the headwaters
of Devil Creek and the Withlacoochee River is a low ridge
about 3 miles north of, and roughly paralleling, the river.
At higher stages some water from the Withlacoochee River
is diverted naturally into Devil Creek through a gap in the
low ridge. This water is returned to the Withlacoochee
Riverfarther downstream. Unlike many of the other swamp
channels in the Green Swamp area, there have been few, if
any, improvements in the Devil Creek channel, and drainage
in this basin is poor.

INFORMATION CIRCULAR NO. 26

Gator Hole Slough heads just east of the Seaboard Air
Line Railroad and flows generally westward through an un-
improved swamp channel, entering the eastern boundary of
the Withlacoochee State Forest about 3 miles west of the
railroad. It continues within the boundaries of the State
Forest to the confluence with Devil Creek which empties
into the Withlacoochee River 2- miles farther west.

There are no other large tributaries of the Withlacoochee
River between this point and the western boundary of the
GreenSwamp area, 4 miles downstream. The drainage area
of Withlacoochee River at Trilby (station 39) at the western
boundary is 620 square miles. All of this drainage basin is
within the project area except for 60 square miles of lakes
and hills west of U. S. Highway 301 and south of U. S. High-

Little Withlacoochee River: The Little Withlacoochee
River is the largest tributary of the Withlacoochee River
within the Green Swamp area. It heads near State Highway
33 in Lake County and flows in a westerly direction. Much
of the basin east of the Seaboard Air Line Railroad consists
of rolling hills, lakes, and swampy flatlands. In the western
part of the basin the slopes are considerably flatter. The
elevation of the land surface varies from about 125 feet near
State Highway 33 to about 75 feet at the western boundary of
the Green Swamp area.

Bay Root Slough is the headwater tributary of the Little
Withlacoochee River. This slough collects the drainage from
several lakes and swamps east of the Seaboard Air Line
Railroad and flows northwestward to the Lake-Sumter county
line where it enters the Withlacoochee State Forest. Down-
stream from the eastern boundary of the State Forest, Bay
Root Slough forms the Little Withlacoochee River.

The river channel within the State Forest is wide and
shallow and contains dense growths of. cypress trees. The
channel has been allowed to remain in its natural swampy
condition to store as much water as possible, rather than
to remove the waterbyimproved drainage, as a precaution-
ary measure against fire damages to the valuable cypress
and pine trees in the State Forest.

FLORIDA GEOLOGICAL SURVEY

The Little Withlacoochee River emerges from the State
Forest reservation near the Sumter-Hernando county line,
where it is joined on the north by a major canal. This canal
drains a swampy area between the State Forest reservation
andState Highway 50. The river continues throughthe swamp
in a westerly direction to the crossing of State Highway 50
where it turns and flows in a northwesterly directiontoward
U. S. Highway 301. Another canal joins the river about a
quarter of a mile upstream from U. S. Highway 301. This
then turns westward to Big Gant Lake and then to the Little
Withlacoochee River.

The Little Withlacoochee River continues westward from
U. S. Highway 301 and empties into the Withlacoochee River
3 miles downstream. The total drainage area of Little
Withlacoochee River at Rerdell (station43) is approximately
160 square miles.

Palatlakaha Creek: Palatlakaha Creek drains along nar-
row strip along the eastern part of the Green Swamp area.
This creek forms the headwaters of the Oklawaha River
which is the largest tributary of the St. Johns River.

From its headwaters to the point at which this drainage
system empties into the St. Johns River, the many tribu-
taries of the stream bear different names. To aid in under-
standing the relationship of these tributaries to the entire
system, the flow diagram shown in figure 8 has been pre-
pared.

Lowery Lake, the largest of a group of lakes located
near Haines City, is the headwaters of Palatlakaha Creek.
Most of the drainage from Lowery Lake is to the north into
Green Swamp Run through a culvert in the old Haines City -
inundated.

The Palatlakaha Creek basin is confined by sand'ridges
that extend from Lowery Lake northward almost to Lake
Louisa. Between Lowery Lake and the Polk-Lake county
line the drainage course is called Green Swamp Run. The
stream channels inthis water course are not deeply.incised,
and drainage occurs through wide shallow swamps.

INFORMATION CIRCULAR NO. Z6

Green Swamp Run

Big Creek

Lake Louisa Little Creek

Lake Minnehaha

Lake Minneola
SPalatlakaha Creek

Cherry Lake connects these lakes.

Lake Lucy
Lake L a Johns Lake
Lake Emma
Lake Apopka
Palatlakaha Creek
and many small lakes Apopka-Beauclair Canal

Lake Harris Lake Beauclair

Lake Eustis Dora Canal

Haines Creek Lake Yale

Lake Griffin -- Lake Yale Canal-

Oklawaha River

St. Johns River

Figure. 8. Flow diagram of the upper Oklawaha River.

FLORIDA GEOLOGICAL SURVEY

Big Creek and Little Creek drain the basin between the
Polk-Lake county line and Lake Louisa. Big Creek is a
continuation of Green Swamp Run. The stream channels for
both Big Creek and Little Creek have more definitely incised
valleys and the flood plain swamps are not as wide as those
for Green Swamp Run.

The Big Creek basin is confined along its eastern bound-
ary by a high ridge. However, along the western boundary,
the ridge is broken by swamps in several places and the
Big Creek and Little Creek basins are interconnected. Big
Creek, including Green Swamp Run, drains an area of about
70 square miles. The basin, from Haines City to Lake
Louisa, is about 25 miles long and from 2 to 4 miles wide.
The elongated shape of this basin, in addition to the flat
slopes and the dense vegetation inthe swamp channel, result
in an inefficient drainage system.

Little Creek drains the area west of Big Creekand emp-
ties into Lake Louisa. The western boundary of the Little
Creek basin is fairly well defined by low ridges. However,
in a few places the ridges are broken by saddles. The
exchange of surface drainage between Little Creek and
Withlacoochee River through these saddles appears to be
negligible.

The southern boundary of the Little Creek basin is not
well defined. The probable boundary is along an old road
that extended from State Highway 33 to U. S. Highway 27
about a mile or two north of the Lake-Polk county line. This

Much of the drainage from the area that was formerly
drained by Little Creek has been diverted into the Withla-
coochee River by interceptor canals. These canals are lo-
cated near the Polk-Lake countyline. However, some water
from its former basin still drains into Little Creek through
natural swamp channels thatwere not closedwhenthe inter-
ceptor canals were dug.

The present (1960) drainage area for Little Creek, as
outlined in figure 7, is about 15 square miles during dry

INFORMATION CIRCULAR NO. 26

periods. During wet periods water flows into the basin
through the openings in the road along the southern bound-
ary of Lake County.

Lake Louisa is the uppermost of a chain of large lakes
in the upper Palatlakaha Creek system. Lake Minnehaha,
Lake Minneola, and Cherry Lake are next in order below
Lake Louisa. These lakes are connected by the wide, deep
channel of Palatlakaha Creek. In addition to draining these
lakes, Palatlakaha Creek also drains an area of smaller
lakes and upland marshes westward to State Highway 33.
This area affords storage facilities for large quantities of
water.

During the latter part of 1956 an earthen dam with two
radial gates was built at the outlet of Cherry Lake to main-
tain the stages of the waterway and lakes upstream during
prolonged periods of dry weather. The water surface from
the upper pool at this dam to Lake Louisa is essentially
level except during periods of high discharge. During the
period of maximum discharge in 1959, the stage of Lake
Louisa was about 0. 7 foot higher than that of the upper pool
at Cherry Lake outlet. The fall between Lake Louisa and
Lake Minnehaha was about 0. 3 foot during this period.

The channel below Cherry Lake has been improved by
a canal leading into Lake Lucy and Lake Emma. Palatlakaha
Creek follows a more definite channelwith steeper gradient
from Lake Emma to its mouth at Lake Harris. In this reach

Hillsborough River: One of the major drainage outlets
fromnthe area of investigation is a wide shallow swamp chan-
nel where part of the flood waters from the Withlacoochee
River overflows into the Hillsborough River basin. This
outlet is located in southeastern Pasco County. The outlet
is about a mile wide and extends 3- miles southward from
the Withlacoochee River to Fox Branch. The confluence of
the outlet and Fox Branch forms the Hillsborough River.

White (1958, p. 19-24) presents considerable evidence
to support an assumption thatthe Withlacoochee-Hillsborough
overflow channel was formerly the main outlet of the upper

FLORIDA GEOLOGICAL SURVEY

Withlacoochee River. The drainage pattern, as shown in
figure 7, could certainly lead to this conclusion.

The valley of the Withlacoochee-Hillsborough overflow
is crossed by the road fill and bridge of U. S. Highway 98.
The channel at the highway crossing has been narrowed to

The overflow channel is crossed also by State Highway
35A about 1 mile downstream from U. S. Highway 98. Two
bridges cross the channel at State Highway 35A.

Reedy Creek: Reedy Creek is one of the headwater trib-
utaries of the Kissimmee River. The drainage from 5 square
miles of the GreenSwamp area in southeastern Lake County
flows eastward past U. S. Highway 27 into Sawgrass Lake
and into marshes in the headwaters of Reedy Creek. The
hydraulic conveyance of the waterway east of the highway
has been improved by dredging, and the capacity of the chan-
nel is adequate for satisfactory drainage. The upstream
channel west of the highway is unimproved.

Diversions and Interconnection of Basins

Although surface drainage from the Green Swamp area
follows rather definite routes and although the drainage di-
vides are generally determined by the topographic features,
there are severalplaces where the basins are interconnected
and water is diverted from one basin to another. Many of
thesepoints of diversion have been mentionedunder thefore-
going discussion of the drainage basins. The hydrologic im-
portance of these interconnections, which are integralparts
of the drainage systems, is shown inthe following discussion.

The arrows on the map in figure 7 locate and show the
direction of flow through many of the saddles inthe drainage
divides. The interconnections that are shown on the map
are the most important ones disclosed bythe investigation,
but they by no means include all such points in the small
subbasins where there is no definite drainage divide.

INFORMATION CIRCULAR NO. 26

One of the major diversionary channels is the Withla-
coochee-Hillsborough overflow in southeastern Pasco County.
This diversion was discussed in detail under the section de-
scribing the Hillsborough River basin. This overflow chan-
nel is at a higher elevation than the Withlacoochee River
channel and receives no discharge when the Withlacoochee
River is at low stages. However, at high stages more than
a fourth of the flow from the Withlacoochee River is diverted
through this channel into the Hillsborough River basin.

Another major area where the basins are interconnected
is near the Polk-Lake county line in the eastern part of the
Green Swamp area. The sand ridges in this area are dis-
membered by a transverse network of swamps that connect
the Withlacoochee River and Little Creek basins. The align-
ment of the swamps and the relative widths of the flood
plains shown on aerial photographs indicate that, in the for-
mer natural state, water carried to this areafromthe south
was discharged by either of three different routes Big
Creek, Little Creek, or Withlacoochee River. The greater
evidence indicates that most of the drainagefromthe south-
eastern area ran off via Big Creek and Little Creek.

In recent years, probably since 1950, extensive opera-
tions by the landowners have considerably altered the pat-
tern of drainage in the eastern area. Many miles of canals,
ditches, and drains have been dug; levees with resulting
borrow ditches have been constructed; and road fills with
drainage structures have been built. These physical changes,
whichwere made for the development of the area, apparently
changed the proportion of the water that drained by the three
routes. Based entirely on the present pattern of drainage
canals and without any factual data on the streamflow from
the upper basins prior to the development of the area, it
appears that the most significant change has been a decrease
in the area drained by Little Creek and an increase in the
area drained by the Withlacoochee River.

Major canals, extending in a northwesterly direction,
near the Polk-Lake county line appear to have intercepted
the greater part of the flow from an area of about 60 square
miles that was formerly the headwaters of the Little Creek
basin. This area is roughly 18 miles long and 3 to 4 miles

FLORIDA GEOLOGICAL SURVEY

wide. It extends from the present southern divide of the
Little Creek basin southward almost to the town of Lake
Alfred. The greater part of the water from this area now
drains to the Withlacoochee River. However, as discussed
in preceding paragraphs, some flow still enters the Little
Creek basin from its former upper basin. Then, too, before
the interceptor canals were dug, some flow escaped from
the Little Creek basin through the natural swamp channels
into the Withlacoochee River. Because the drainage basins
were unconfined before development, and still are uncon-
fined even with the present drainage canals, the change in
proportion of drainage between the two basins and the in-
creased effectiveness of the system in draining the areas
maybe determined from streamflow records only for broad
areas rather thanfor the local areas where the major changes
have occurred. The runoff under present conditions, as
compared with the runoff that occurred during the earlier
years for which streamflow records are available, is dis-
cussed in a following section of this report.

A levee now fills a saddle in the drainage divide be-
tween Green Swamp Run and the Withlacoochee River in
sec. 5, T. 25 S. R. 26 E. south from the Polk-Lake county
line (fig. 7). Prior to the construction of the levee, drain-
age from Green Swamp Run divided into flow westward into
the Little Creek basin (now the Withlacoochee River basin)
and flow northward into Big Creek basin. The levee closed
through which natural diversion occurred.

Lowery Lake and swamps in the upper Withlacoochee
River basin are connected by a natural saddle in the confin-
ing ridge along the northwest shoreline of the lake. This
saddle is 200 to 300 feet wide and is one of the points at
which flow is diverted between the Palatlakaha Creek and
Withlacoochee River basins. The two basins are intercon-
nected at this point only at high stages. Water may flow
through this saddle in either direction, depending on the
distribution of rainfall and the relativewater levels in the
basins.

There are three interconnections between Big Creek
and Little Creek. These openings, all in Lake County, are

INFORMATION CIRCULAR NO. 26

small and their net exchange of water is probably negligible
in comparison with the total flow from the basin.

Other places, shown on the map in figure 7, where
basins are interconnected are: between the Withlacoochee
Lake Mattie and Pony Creek, between Pony Creek and Grass
Creek, and between the Withlacoochee River and Devil Creek.
Many of these interconnections act as equalizing channels
through which water mayflow in either direction, depending
on the relative water levels in connecting basins.

Runoff Characteristics

Seasonal Distribution: Streamflow records for Withla-
coochee River at Trilby (station 39) are available for the
periods August 1928 to February 1929, and February 1930
to December 1959. The drainage basin above this gaging
station occupies two-thirds of the Green Swamp area. The
records for this station are a good index for showing the
variations of surface runoff from the entire area.

The yearly and seasonal runoff from the basin follow
the same pattern as that for the corresponding rainfall.
Figure 9 shows a bar graph of the annual discharge for
Withlacoochee River at Trilby for each complete year of
record, 1931-59. The relationship between rainfall and run-
off is shownby comparing the runoff pattern infigure 9 with
the rainfall pattern in figure 3. Rainfall and runoff records
show that the year 1959 was the wettest of record. The
total rainfall over the Green Swamp area for the 1959 calen-
dar year was 73. 6 inches. The average discharge at the
Trilby station for that year was 1, 152 cubic feet per second.
Runoff in inches, adjusted for the flow diverted to the Hills-
borough River, for that year was 29. 2 inches.

The drought of 1954-56 was probably the most severe
dry period of record, considering its 3-year duration and
yearly deficiencies. Average rainfall over the area for
1954-56 was 38. 5, 40. 3, and 46.6 inches peryear, respec-
tively. The prolonged period of low rainfall resulted in low
discharges at the Trilby station for each of the three years;

FLORIDA GEOLOGICAL SURVEY

the lowest yearly mean discharge at the Trilby station, how-
ever, was 75.4 cubic feet per second for 1932, a year in
which the annual rainfall amounted to 42. 4 inches. Distri-
bution and intensity of the rainfall and the effluent from cit-
rus concentrate plants, derived from ground-water sources,
probably helped account for the fact that annual discharges
for 1954-56 were higher than those for 1932.

The average monthly flow for the Withlacoochee River
at Trilby is shown by bar graphs in figure 10. These aver-
ages indicate that runoff from the basin is lowest for the
months of November through June, with May being the month
of lowest flows. The season of highest flows is the 4-month
period, July throughOctober. During these months, 58 per-
cent of the average annual flowfrom the basin has occurred.

The variation of runoff, as shown in figure 10, is due
to seasonal rainfall. The runoff and rainfall patterns are
similar except that increases in discharge lag the increases
in rainfall by 1 to 2 months. The seasonal distribution of
rainfall is shown in figure 2. The time lag is an indication
of the effects of natural storage capacity of the drainage
basin.

At the beginning of the wet season in June and July,
much of the water is temporarily stored in the swamps and
the streamflow does not increase in the same proportion as
the rainfall. After the surface and ground-water reservoirs
are nearly filled there is less storage capacity in the area
and runoff from subsequent heavy rains reaches the streams
more quickly. September has been the month of highest
runoff. Although the average rainfall decreases sharply in
October, the runoff remains high as the result of water
draining into the stream channels from storage.

The duration curve of daily flow for Withlacoochee River
at Trilby for the 29-year period, 1931-59, is shown in fig-
ure 11. It indicates the percentage of time that a specified
discharge was equaled or exceeded during the period of
record. In a strict sense the flow-duration curve applies,
only to the period for which data were used to develop the
curve. However, if the flow-duration curve represents a
long-term period of flow of the stream, the curve may be

INFORMATION CIRCULAR NO. 26

in 0 U
0101 0)

.0 0)
.0 )fl
0) 0)

CALENDAR YEAR
Annual discharge of Withlacoochee River at
Trilby, Florida.

2

0
a
V

C
c
u
J)

1000

Based on
900 records 1928-

1929, 1930-1959
800

700

600 -

500
Average (1931-59)
383 cfs

300

S200 -----

In00 --------

0
J FMA MJ J A S D ND

Figure 10. Average monthly discharge of Withlacoochee
River at Trilby, Florida.

1200

1100

1000
iooo

S900
Soo
800

700
Uv

600

500

400

300

200

100

Figure 9.

'* nAVERAGE (1931-59)
383 cfs

--------- -
-i---
m'

too
n r"

FLORIDA GEOLOGICAL SURVEY

PERCENT OF TIME DISCHARGE EQUALED OR EXCEEDED THAT SHOWN

Figure 11. Flow-duration curve of Withlacoochee River
at Trilby, Florida.

10,000
apoo
6.000

4.000

2.000

INFORMATION CIRCULAR NO. 26

considered a probability curve and used to estimate the
percent of time a specified discharge will be equaled or
exceeded in the future. The use of flow-duration curves to
indicate the future pattern of flow from a basin is valid only
if the climatic conditions remain the same and amount and
distribution of runoff from the basin is not significantly
changed by man-made changes in drainage systems, storage
capacities, land use, and diversions.

The flow duration curve for Withlacoochee River shown
in figure 11 may be only an approximate representation of
duration of future low flows because of the progressive in-
creases in ground-water pumpage above the gaging station.
Since about 1941 or 1942, the effluent from the citrus con-
centrate plants at Dade City has been draining into the With-
lacoochee River above the Trilby gaging station. The water
used by these plants is obtained from wells. The effluent,
measured at station 32 during 1958-59, ranged from 5 cubic
feet per second, whentheplants were at minimum operation,
to about 66 cubic feet per second at peak operation during
the citrus packing season. This inflowto the river produces
higher discharges at the gaging station during dry periods
than would be derived from the natural yield of the basin.

Areal Distribution and Basin Runoff: The total runoff
from 862 square miles, most of which is in the area of in-
vestigation, was measured by gaging stations at each of the
five major outlets. Table 3 summarizes the mean discharges
determined at each of these outlets for the water year begin-
ning October 1, 1958, and ending September 30, 1959.

This table shows the distribution of the discharge from
month to month as well as by streams draining from the
area. During the 1959 water year, which was the wettest
of record, the Withlacoochee River carried 62 percent of
the total runoff. Little Withlacoochee River carried 20
percent, Bit Creek carried 6 percent, Little Creek carried
4 percent, and about 8 percent of the flow was discharged
into the Hillsborough River. The proportional amounts of
the runoff carried by each of these outlets varywith amount
and distribution of rainfall because of the increasing rate of
discharge diverted betweenthe interconnected basins during

Table 3. Outflow from Oreen Swamp Area, Water Year 1958-59

Mean dlccharge in cubic fet per second
Sta. No. 4 Sta. No. Sta. No. 17 Sta. No. 39 Sta. No. 4
Big Creek Little Creek Withlacoochee- Withlacoochee Little Withlacoochee Total of
near Clermont near Clermont Hillsborough River at Trilby River at Rerdell all cutlets Cubic feet
(drainage area, (drainage area, overflow near (Drainage area. (Drainage area. (Drainage area per second per Runoff in
Month 67.0 sq. mi.) 15 sq. mi.) Richland 620 sq. mi.) 160 sq. mi.) 862 sq. ml.) square mile inches
October 1958 20.7 2 0 55.1 12.5 90.3 0.105 0.12

November 33.2 a 10 0 98.9 35. 4 177. .206 .23

December 11.8 9 *10 75.3 21.1 127.2 .148 .17

January 1959 62.9 43 *90 600 182 977.9 1.13 1.30

February 31.6 20 4 384 99.9 539.5 .626 .65

March 128 *107 *350 1,457 760 2,802 3.25 3.5

April 168 *115 *250 1.742 459 2,734 3.17 3.54

May 72.2 42 30 790 122 1,056.2 1.23 1.42

June 63.8 74 220 915 178 1,450.8 1.68 1.87

July 205 *108 *270 1,921 488 2,992 3.47 4.00

August 15z 93 *160 1,461 622 2,488 2.89 3.33

September 133 *117 *210 1,942 701 3,103 3.60 3.88

Water year 90.5 61.8 134 956 308 1,545 1.79 24.26

Percent
oftotal 6 4 8 62 20 100

Estimated on basis of discharge measurements made at monthly Intervals and records for other stations

INFORMATION CIRCULAR NO. 26

high water. During prolonged periods of little or no rain-
fall the combined flows of Big Creek, Little Creek, and
Withlacoochee-Hillsborough overflow are negligible.

Continuous records of stage and discharge for Big Creek
near Clermont (station 4) are available from July 17, 1958,
to September 30, 1959. Maximum discharge for this period
was 283 cubic feet per second on July 12, 1959. The mini-
mumdischarge was 0. 4 cubic foot per second on September 7,
1958. Since May 1945 occasional discharge measurements
have been made at this gaging station. The maximum dis-
charge measured was 428 cubic feet per second on Septem-
ber 23, 1947. No flow was observed on May 31, 1945.

The discharge of Little Creek was measured at a site
(station 5) about half a mile upstream from Lake Louisa.
Discharge measurements were made at monthly intervals
during the period of the investigation. The maximum dis-
charge measured was 210 cubic feet per second onMarch 23,
1959. The minimum discharge measured was 0. 6 cubic foot
per second on October 15, 1958. Since May 1945 occasional
discharge measurements have been made on the same days
that the flow of Big Creek was gaged. The maximum dis-
charge measured during the earlier years was 359 cubic
feet per second on September 23, 1947. No flow has been
observed several times.

The 1959 water year mean monthly discharge figures
for Little Creek, shown in table 3, were estimated on the
basis of monthly discharge measurements and daily records
for nearby stations. Little Creek carries only a small per-
centage'of the total discharge fromthe area of investigation.
Although the estimated figures of discharge for Little Creek
maybe subject to considerable error, the compositefigures
of runoff fromthe total area of investigation should be fairly
reliable.

Daily records of discharge fromthe upper Withlacoochee
River basin are available from July 22, 158, to September 30,
1959. These records were collected at the crossing of State
Highway 33 near Eva (station 23). The drainage area above
this gaging station is approximately 130 square miles.

FLORIDA GEOLOGICAL SURVEY

The maximum discharge at this stationfor the period of
record was 836 cubic feet per second on March 21, 1959.
No flow occurred October 3, 1958. For several days in
August, September, and October 1958, the flow was less
than 1 cubic foot per second. The mean annual discharge
for the 1959 water year was 208 cubic feet per second.

During the period February 1930 to September 1931
daily discharge records of the Withlacoochee-Hillsborough
overflow were collected. This gaging station was located at
the crossing of State Highway 35A which is about 1 mile
downstream from U. S. Highway 98. For the period July
1958 to September 1959 discharge measurements were made
at monthly intervals at U. S. Highway 98 (station 17). The
water stage was determined by measuring from a reference
point on the bridge. A rating curve of stage versus dis-
charge was plotted. The stage of each significant rise was
determined from floodmarks on a crest-stage indicator and
the corresponding discharge was computed from the rating
curve.

The maximum discharge, that occurred during the period
July 1958 to September 1959, was 870 cubic feet per second
in March 1959. At low stages of the Withlacoochee River,
no flow occurs in this overflow channel. Correlation of
discharge in this overflow channel with those for the gaging
station on Withlacoochee River at Trilby (station 39) indicates
that little or noflow occurs in the overflow channel when the
discharge at Trilby is less than 200 cubic feet per second,
which is about 50 percent of the time according to the flow-
duration curve for the Trilby station, shown in figure 11.
During the period of the investigation, noflow occurred from
July to the latter part of December 1958.

The monthly discharges, shown in table 3, for the
Withlacoochee-Hillsborough overflow were estimated from
discharge measurements, recorded peak stages, and daily
records for nearby stations.

Many of the characteristics of streamflow for the With-,
lacoochee River at Trilby have been discussed in the pre-
ceding section relating to the seasonal distribution of flow.

INFORMATION CIRCULAR NO. 26

The discharge at the Trilby gaging station does not
represent the natural runoff from the Withlacoochee River
basinbecause of the high-water flow diverted fromthe basin
by the Withlacoochee-Hillsborough overflow channel and the
effluent into the river from the citrus concentrate plants at
Dade City. Above the stage at whichthe Withlacoochee River
flows through the diversionary channel, the amount of
discharge diverted into the Hillsborough River increases in
direct proportion to the discharge of the Withlacoochee
River. Studies of basin runoff for either the Withlacoochee
or the Hillsborough rivers must be adjusted for the amount
of discharge diverted from one basin to the other. Per-
centagewise, the plant effluent into the basin is small except
when the discharge in the Withlacoochee River is extremely
low and the plant is at peak operation.

The average discharge for the Withlacoochee River at
Trilby for 29 years of record, 1931-59, is 383 cubic feet
per second. The maximum discharge of record was 8, 840
cubic feet per second on June 21, 1934. Flood-frequency
studies by Pride (1958, p. 104) indicate that this flood was
an event of rare occurrence. Historical data collected after
the flood of 1934 showed that itwas the highest flood peak in
at least 75 years. The minimum discharge for the period
of record at the Trilby station was 8. 6 cubic feet per second
June 9-17, 1945.

The average discharge for the Little Withlacoochee
River at Rerdell (station 43) for the 1959 water year was
308 cubic feet per second. The maximum discharge that
occurred during the period of the investigation was 1, 940
cubic feet per second on March 22, 1959. The minimum
for this period was 2. 9 cubic feet per second onOctober 17,
18, 1958.

The runoff characteristics of Little Withlacoochee River
basin and those for the Withlacoochee River basin follow the
same generalpatternbut differ in some respects. The Little
Withlacoochee River basin has a higher runoff. The runoff
at the Rerdell gaging station was 1. 92 cubic feet per second
per square mile (26. 1 inches) for the 1959 water year. The
concurrent runoff at the Trilby station, adjusted for the flow
diverted to the Hillsborough River basin, was 1.76 cubic

FLORIDA GEOLOGICAL SURVEY

feet per second per square mile (23.9 inches). Also,
comparison of discharge records for stations at Trilby and at
Rerdell show that the flood peaks of the Little Withlacoochee
River generally occur about a week before those of the
Withlacoochee River.

Rainfall-Runoff Relation

Many attempts have been made to express in equational
form a relation between precipitation and streamflow. An
approximate relation for a particular drainage basin may be
determined using yearly figures of precipitation and runoff
from that basin, but this relation would not necessarily be
correct for other drainage basins.

The three general factors that affect the relation between
rainfall and runoff are: (1) climatic factors, the most
important of which are rainfall and temperature; (2) drainage
basin characteristics, which include size, shape, surface
slope, the amount of water area, the character of the surface
and subsurface geology, and the condition and type of
vegetative cover; and (3) storage underground and in natural
lakes, ponds, swamps, and artificial reservoirs.

Runoff measured at a gaging station is the total surface
flow from the basin including ground water that has seeped
into the stream above the station. When the flow is con-
verted into runoff in inches over the drainage basin, it can
be compared with the average precipitation over the basin
which is measured also in inches. The ratio of rainfall to
runoff is better defined when average values for long periods
are used. Thus, by using yearly averages the effects of
storage are minimized as part of the water that is tempo-
rarily stored on the surface and underground during the wet
season is eventually removed as streamflow.

The amount of precipitation that fell the previous year
is one of the factors that affect the relation between precip-
itation and runoff and cause points on a graph of annual;
precipitation plotted against annual runoff to scatter. Gen-
erallythe scatter of the points can be reduced by plotting an
effective precipitation (Searcy and Hardison, 1960) instead
of the observed precipitation.

INFORMATION CIRCULAR NO. 26

Use of an effective precipitation is one way of making
allowance for the variable amount of water carried over
from year to year as ground-water storage in the basin.
The effective precipitation (Pe) commonly used is that
proportion of the current year's precipitation (P ) and the
proportion of the preceding year's precipitation (P1) that
furnishes the current year's runoff, or

Pe = aPo + bP1.

The coefficients a and b can be determined by statistical
correlation. The effective annual precipitation thus deter-
mined for the Withlacoochee River basin above Trilby and
for the Palatlakaha Creek basin above Mascotte is

Pe = 0. 3Po + 0. 7P.

Runoff is the residual of precipitation after all of
nature's demands have been met. These demands taken
collectively are called water loss. A simple definition for
water loss is: Water loss equals precipitation minus runoff
adjusted for change in storage and for seepage into and out
of the basin. The basic concept is that water loss is equal
to evapotranspiration, that is, water that returns to the
atmosphere and thus is no longer available for use. How-
ever, as used in this report, theterm applies to differences
between measured inflow and outflow even where part of the
difference may be seepage.

The equation for water loss is:

L= Pe R -S Se

Where, L = water loss
Pe = effective precipitation
R = surface runoff

AS = increase in storage both
surface and underground
Se = net seepage out of the basin
(seepage into basin equals
negative out-seepage)

FLORIDA GEOLOGICAL SURVEY

The annual water-loss curve for Withlacoochee River
at Trilby is shown in figure 12. The P = L line (dashed
line) in figure 12 represents the theoretical limit of water
loss which would occur if the loss equaled the precipitation
and none ran off as streamflow. The average water-loss
curve is shownbythe solid line which was drawn to average
the annual figures of effective precipitation and loss (P R)
,e
for the basin. The departures of the yearly data from the
average curve may be caused in part by storage and seepage
changes and in part by differences inthe way precipitation is
and Se in the water loss equation for Withlacoochee River at
Trilby and they thus add to the apparent evapotranspiration.

The effective annual precipitation of about 30 inches,
indicated by the point where the downward extension of the
curve coincides with the Pe = L relation, is the probable
yearly amount below which no runoff would occur. Under
some conditions of intensity and distribution of precipitation,
there could be runoff with less than the 30 inches of precip-
itation.

As shown by the curve in figure 12, the average water
loss increases with the precipitation until it becomes a
constant for higher values of precipitation. This is the
maximum loss that would occur regardless of the amount of
precipitation and is called the potential natural water loss
for the basin. The potential natural water loss for the
Withlacoochee River at Trilby is shown to be 45 inches.
This figure compares favorably with the 48 inches of average
water loss shown for 72 in figure 5.

Figure 13 shows the plotted yearly figures of rainfall
and runoff for the Trilby station and an average curve. The
average curve was determined by using the curve in figure 12
and plotting the departures of the potential water-loss curve
from the limiting P = L line.

The runoff from the upper Palatlakaha Creek basin has
been measured since 1945. Streamflow records were col-.
lected near Mascotte (station 9) for the period 1945-56 and
at Cherry Lake outlet near Groveland (station 8), about 6
miles upstream, for the period 1956-59. The drainage area

INFORMATION CIRCULAR NO. 26

F-
..41

z34

Figure 12.

Figure IZ.

0 10 20 30 40 50 60
Water los, La PI-R, in Inches
Relation of effective annual rainfall and annual
water loss, Withlacoochee River at Trilby,
Florida, 1931-59.

Annual runoff, in inches

Figure 13.

Relation of effective annual rainfall and annual
runoff, Withlacoochee River at Trilby, Florida,
1931-59.

FLORIDA GEOLOGICAL SURVEY

at the lower station is approximately 180 square miles and
at the upper station, 160 square miles. Figures of runoff
per square mile at the two stations are assumed tobe equiv-
alent for studying the characteristics of the upper basin.

The water-loss curve for the upper Palatlakaha Creek
basin is shown in figure 14. Annual changes in storage in
the many lakes and swamps above the gaging station were
computed by measuring the water-surface area from maps
and using the year-end changes in stage as recorded for
Lake Minnehaha. No allowance was made for change in
underground storage or for seepage into or out of the basin.

The water-loss curves shown in figures 13 and 14
indicate that for a year in which rainfall was 30 inches or
less, the natural losses would equal the rainfall and no
runoff would occur from either the Withlacoochee River or
the Palatlakaha Creek basins, both draining from the Green
Swamp area.

The potential natural water loss for the upper Palatlakaha
Creek basin, as shown in figure 14, is 50 inches which also
compares favorably with the 48 inches of annual water loss
shown for 72* in figure 5.

The potential natural water loss in the Palatlakaha
Creek basin appears to be 5 inches more than that for the
Withlacoochee River basin. Increased evaporation losses
from the open-water surface of the many lakes in the
Palatlakaha Creek basin and ground-water movement out of
the basin could account for this difference, but as explained
in the next section, man-made changes may be a more
logical explanation for part of the difference.

Figure 15 shows the relationship of the effective annual
rainfall and runoff for the upper Palatlakaha Creek basin.
The average curve was determined by using the curve in
figure 14 and plotting the departures of the potential water-
loss curve from the limiting Pe = L line. The figures of
annual runoff have been adjusted for changes in storage as
explained in a preceding paragraph but not for diversions
from the basin.

INFORMATION CIRCULAR NO. 26

10 20 30 40
Water loss, L P--R-AS, in inches

Figure 14.

Relation of effective annual rainfall and annual
water loss, Palatlakaha Creek above Mascotte,
Florida, 1946-59.

TO

1957 194
1948

50 954 -- Nolt- Total runoff ad-
o justed for change
195/ in storage in lakes
/ and swamps above
1955 gaging station
40 i
1 1946-53
I I 1954-59
I

20
30 --- ----------o------

0

10 15
Annual runoff,

20
in inches

Figure 15.

Relation of effective annual rainfall and annual
runoff, Palatlakaha Creek above Mascotte,
Florida, 1946-59.

25 30

FLORIDA GEOLOGICAL SURVEY

Many of the physical changes that have been made on
the land surface through man's efforts have already been
described. The most extensive developments of the area
have occurred in recent years, but the first changes in the
hydrologic characteristics undoubtedly occurred several
years ago when logging trails and tramroads were built and
much of the native timber was cleared from the area. The
early developments of the area cannot be evaluated as they
predate the period of data collection, but they probably had
only minor effects on the hydrology.

The present network of canals, ditches, roadfills,
borrow canals, and dikes was constructed over a period of
several years but the most extensive developments probably
occurred after 1950. The canals and ditches were con-
structed to improve the effectiveness of the drainage system.

Some of the results of the improved drainage in small
local areas are obvious from inspections of the area.
Improved pasturelands now occupy areas where the water
once stood for long periods in shallow sheets over flatlands.
Citrus groves have been planted in many flatland areas
because flood hazards have been reduced by improved drain-
age. The water levels in many small landlocked sinkholes,
cypress swamps, and saucer-like depressions have been
lowered by ditches connecting them with swamps at lower
elevations. These are some of the obvious results of
improved drainage in small areas but give no indication of
overall changes in drainage from the general area.

Changes in the drainage characteristics of the area of
investigation can be detected by comparing the hydrologic
datafor early years before drainage developments occurred
with the data collected since the major developments have
occurred. Since February 1930 continuous daily-discharge
records have been collected for the Withlacoochee River at
Trilby, the major drainage outlet of the area. Rainfall
records at network stations in central Florida began at an
earlier date than did the records of discharge. Some of the
rainfall records extend back about half a century. Daily-
discharge records have been collected since 1945 in the

INFORMATION CIRCULAR NO. 26

upper Palatlakaha Creek basin. A longer record would be
more useful for detecting changes ortrends inthe pattern of
discharge from the upper Palatlakaha Creek basin, but the
available records cover the periodwhen most of the changes
occurred and have been used in this study.

Double-mass curves of cumulative measured runoff and
cumulative computed runoff have been plotted to provide a
means of examining the records of streamflowfromthe area
of investigation to detect changes that may have occurred
(Searcy and Hardison, 1960). A double-mass curve is
obtained by plotting cumulative totals of one variable against
cumulative totals of a second variable for the same period
of time. Any significant change in the relationship of these
variables is identified by a break or change in slope of the
straight line averaging the points. The variables used in
preparing the curves shown in figure 16 are the values of
cumulative computed runoff, taken from the precipitation-
runoff relations in figures 13 and 15, and cumulative meas-
ured runoff at each of the two gaging stations.

The rainfall pattern is not affected by the progressive
changes to the drainage system in the Green Swamp area.
Likewise, it canbe assumed that the theoretical or computed
runoff has not changed as it is taken from an average curve
for several years of record. Therefore, any change in slope
in the double-mass curves of figure 16 would be caused by
changes in actual or measured runoff.

Curve A in figure 16 is the double-mass curve for the
Withlacoochee River basin above the Trilby gaging station.
Straight lines are drawnto average severalpoints that show
definite overall trends. Changes in slope of these lines
occur between 1934 and 1935 and between 1954 and 1955.
The change in slope between 1934 and 1935 is an indication
of a change in the runoff pattern but the authors have no
knowledge of the cause of such a change. Between 1935 and
1954 the yearly values, in general, are averaged by a single
straight line, thus indicating no significant change in runoff
pattern from the basin between these dates. Minor devia-
tions of the plotted yearly values of runoff are probably
caused by variations of rainfall distribution and intensity
during the year and are not indications of changes in the

(II WihlIn.h lir I t. Trllb, FIP I1 19l1-1

AII PalriM s .rHt enm ktui*t, Fr., 1 44--

, -

*t~- --- -- -

^ :L I L_ != _j-------

umInl iAllAT Mn7, Ii vl* elloNES
101 o-9M WIlth 591* 9 WI. r MItWk Cml
bina, 49L4

Figure 16.

Double-mass curves of measured runoff versus computed runoff,

Withlacoochee River and Palatlakaha Creek basins.

14D

INFORMATION CIRCULAR NO. 26

long-termtrends. Yearlyvalues of runoff for 1955-59 define
an average line with a flatter slope than that for 1935-54.
This change in slope indicates that a higher rate of runoff
from the basin occurred during 1955-59 than that indicated
from the same rainfall pattern of previous years.

Curve B in figure 16 is the double-mass curve for the
upper Palatlakaha Creek basin. The figures of annual runoff
were adjusted for storage changes as described in the
previous discussion of the relation between rainfall and
runoff. For the period 1945-53, curve B takes the general
direction as shown by the straight line. However, after
1953, a definite break occurs in the slope of the average
line of curve B indicatingthat less runoff occurred from the
area. The decrease in runoff since 1953 as indicated by the
difference in slope of the line averaged about 3 inches per
year. Three possible explanations of the decreased runoff
from the upper Palatlakaha Creek basin since 1953 could be:
(1) increased seepage and evapotranspiration losses since
the Cherry Lake controlwas constructed in 1956, (Z) lowered
ground-water levels during the dry years of 1954-56, which
allowed more infiltration to the aquifers, and (3) the diver-
sion of the headwaters of Little Creek intothe Withlacoochee
River. (Much of the drainage work in the area was started
during the dry years of 1954-56.)

The minimum stage of the chain of lakes above Cherry
Lake was stabilized by the construction of the control in
1956. The average increase in stage at this control is
small, ranging from 2. 5 feet at low flows to none at high
flows. The water-surface area above the Cherry Lake
control was increased also by a small amount. Seepage and
evapotranspiration losses may have been slightly increased
by the low-head pool but the amount is negligible, percent-
agewise. The increased runoff during 1959, an extremely
wet year, partly compensated for the effect of the dry years
of 1954-56. Additional years of record will be required to
establish the slope of the curvethat will average the effects
of wet and dry years since 1953 and to evaluate further
changes in diversions mentioned in (3) above.

FLORIDA GEOLOGICAL SURVEY

Curve C in figure 16 has been plotted to show the
cumulative runoff from the combined Withlacoochee River
and Palatlakaha Creek basins. The average line defining
curve C has the same slope for the entire period, 1945-59.
This indicates that there has been no significant loss from
the combined basins.

The only remaining explanation for the significant de-
crease in runoff from the Palatlakaha Creek basin is a
decrease in the size of the drainage area. Such a change in
the headwaters of Little Creek, a tributary to Palatlakaha
Creek, has been previously discussed. This change has
resulted in the diversion of part of the flow from the Little
Creek basin into the Withlacoochee River basin. This also
explains the increased runoff from the Withlacoochee River
basin for 1955-56 as indicated by curve A in figure 16.
However, the gain to this basin is not as obvious as the loss
from the Palatlakaha Creek basin because of the difference
in size of the drainage basins.

The effects of the diversions from the Palatlakaha
Creek basin to the Withlacoochee River basin are not
clearly indicated at present by the plotting of data for the
precipitation-water loss relations (figs. 12, 14) and the
precipitation-runoff relations (figs. 13, 15). The fact that
such diversions have occurred is more apparent when the
cumulative amounts are magnified by the double-mass curves
in figure 16.

Chemical Characteristics of Surface Water

The chemical characteristics of a water depend on the
environment through which it has passed. For instance,
rain water usually has a very low mineral content as there
is a very low concentration of minerals in the atmosphere.
However, rain water does contain some dissolved gases
such as carbon dioxide and oxygen. As the rain water
contacts the ground some minerals are dissolved, depending
mostly on the solubility of the minerals, the time of contact,
and the acidity of the rain water.

INFORMATION CIRCULAR NO. 26

Because of prolonged leaching, the minerals at the
surface of the ground in most of the Green Swamp area are
of low solubility; therefore, surface water is generally low in
mineral content. The more soluble minerals are generally
found below the surface of the ground. Also, water under
the ground has a greater area of contact with these soluble
minerals and the time of contact of the water with the
soluble minerals is usually longer than water onthe surface.
Therefore, water that has been stored in the ground and
returned to the surface usually contains a higher mineral
content than that remaining on the surface.

Water has been typed according to the concentrations
of the individual mineral constituents for convenience in the
discussion that follows. For example, a chloride-type water
is a water that contains more chloride in parts per million
than any other mineral constituent. The concentration of
chloride could be less than 10 parts per million and still be
a chloride-type water. Mineral content that is less than
100 parts per million is generally considered to be low. A
water that contains more than 100 units of color is considered
to be highly colored.

The surface water of the Green Swamp area is low in
mineral content. The content ranged from 18 to 122 parts
per million. Generally water is considered to be usable if
it contains less than 400 to 500 parts per million of mineral
content. However, water that contains less than 400 to 500
parts per million may be suited for one industrial or
municipal concern but may be completely unusable for
another. The most undesirable characteristic of the surface
water in the Green Swamp area would probably be its color.
Most of the water is highly colored and color values as high
as 600 were noted. None of the surface water was hard,
that is, the water had a hardness of less than 120 parts per
million.

The effects of environment are discussed by comparing
the characteristics of water at various points in the river
basins. The characteristics are determined for both high
and low flows.

68 FLORIDA GEOLOGICAL SURVEY

During the period of relatively high flow in November
1959, the water in the Withlacoochee River between Eva
(station 23) and Dade City (station 31) was found to be
chloride in type, low in mineral content, and highly colored.
Between Dade City and Croom (station 44) the water changed
to a carbonate type and mineral content increased from 30
to 56 parts per million (fig. 17). This increase was greater
between Dade City and the bridge on Cummer Cypress
Company road (station 33) than between other points along
the river within the area of investigation. Part of this
increase in mineral content could be due to industrial and
municipal disposals into a drainage canal at Dade City.
This canal empties into the Withlacoochee River between

Between station 33 and Trilby (station 39), there was a
slight decrease in mineral content. This decrease probably
was caused by inflow water from Gator Hole Slough and
Devil Creek. This water is low in mineral content.

120- o
110-
100 -

90-
70a -

30-
0o-0 5 t1 s1 20 25 30 35 40 45 80
Relative downstream distance in miles

Figure 17. Mineral content in the Withlacoochee River.

INFORMATION CIRCULAR NO. 26

Between Trilby and Croom (station 44) there is an
indication of ground-water inflow to the river. The mineral
content of the possible total inflows was estimated on the
basis of data collected on November 12-13, 1959, from
three stations:

Discharge Mineral
in cubic content in
feet per parts per
Station second million

(39) Withlacoochee River at Trilby 950 46
(43) Little Withlacoochee River at
Rerdell ................... 120 46
(44) Withlacoochee River at Croom 1,230 56

By using the equation, 0OC1 C Q2C2 + Q3C3 = Q4C4
(Hem, 1959): where,

Q is the discharge in cubic feet per second
C is the mineral content in parts per million
Q1C1 is the instantaneous load at Trilby
02C2 is the instantaneous load at Rerdell
Q3C3 is the instantaneous load between the stations
04C4 is the instantaneous load at Croom

the instantaneous load, Q3C3, was computed. The increase
in discharge (Q3) was determined to be 160 cubic feet per
second by subtracting the sum of discharges at Trilby and
Rerdell from that at Croom. The mineral content (C ) thus
computed is about 120 parts per million. The mineral
content in the principal aquifer in this area is usually
greater than 250 parts per million. The computed mineral
content indicates that the inflow between the stations was
probably a composite of surface-water and ground-water
inflows.

At the time of low flow in September 1958, the water in
the Withlacoochee River near Eva was chloride in type, of
low mineral content, and highly colored. Between Eva and
Dade City, the water changed to a carbonate type and in-
creased in mineral content from 32 to 59 parts per million.
From Dade City to Trilby, the mineral content increased

FLORIDA GEOLOGICAL SURVEY

and undoubtedly part of this increase was due to waste
ground-water inflow probably occurs along this reach of the
river, but the data are insufficient to advance any conclusions
concerning the quantity and chemical quality.

The Little Withlacoochee River was investigated at
waters were chloride in type and of low mineral content,
ranging from 30 to 86 parts per million. The color inten-
sities ranged from 80 to 150 units. At Rerdell the water
was carbonate in type for both high and low flows. This
indicates ground-water inflow to the stream between the

Samples were obtained at the Withlacoochee-Hillsborough
overflow at U. S. Highway 98 (station 17) in February, March,
and November 1959, and at Hillsborough River at State
Highway 39 near Zephyrhills (station 19) in November 1959.
These data indicate the possibility of ground-water inflowsto
the stream above these points. The water was carbonate in
type, low in mineral content, and contained some color. The
mineral content was 44 parts per million in Withlacoochee-
Hillsborough overflow and was 48 parts per million in
Hillsborough River at State Highway 39. This mineral
content is higher than that along the upper reaches of the
Withlacoochee River and most other areas within the Green
Swamp. The mineral content in the upper Withlacoochee
River was less than 30 parts per million.

The waters in Big Creek and Little Creek near Clermont
(stations 4 and 5) were chloride in type, of low mineral
content, and highly colored. These streams are similar in
chemical characteristics and show little change in charac-
teristics between high and low flows. In Big Creek, the
mineral content from four samples ranged from 21 to 31
parts per million when discharge ranged from 5. 9 to 279
cubic feet per second. Little Creek was sampled four times
and the mineral content ranged from 18 to 23 parts per
million and discharge ranged from 4. 7 to 210 cubic feet per
second. The color intensities inthese streams ranged from
130 to 300 units.

INFORMATION CIRCULAR NO. 26

Horse Creek near Davenport (station 14) and Reedy
Creek near Loughman (station 13) were sampled once in
November 1959 when most of the streams in the area were
above base flows. Chemical analysis of the water sample
from Horse Creek shows that the water was carbonate in
type, of low mineral content (58 ppm), and colored (72 units).
The mineral content was similar to those collected in the
Hillsborough River above State Highway 39 (station 18).
The water from Reedy Creek was chloride in type, of low
mineral content (33 ppm), and colored (80 units).

Water samples were collected for analysis inSeptember
1959 from Lake Louisa near Clermont (station 6) and in
November 1959 from Bay Lake near Mascotte (station 40),
Little Lake Agnes near Polk City (station 27), Crystal Lake
at Lakeland (station 16), and Lowery Lake near Haines City
(station 1). The mineral content in waters of these lakes
was less than 70 parts per million. Based on these five
analyses, there does not appear to be any significant dif-
ferences in the mineral content in the waters of the lakes.

Ground Water

General

Ground water is the subsurface water in the zone of
saturation that zone in which all pore spaces are filled
with water under atmospheric or greater pressure. Ground
water can be divided into two classes related to the geology
of an area: (1) ground water that occurs under nonartesian
(unconfined) conditions, and (2) ground water that occurs
under artesian (confined) conditions.

In an unconfined aquifer the upper surface of the zone
of saturation is freeto rise and fall. In a confined or semi-
confined aquifer the deposits are completely saturated and the
water is under sufficient pressure (greater than atmospheric)
to rise above the top of the aquifer in wells that tap it. The
level to which artesian water will rise'intightly cased wells
that penetrate the artesian aquifer is called the piezometric
surface.

FLORIDA GEOLOGICAL SURVEY

Ground-water movement depends on the permeability
of the aquifer and the hydraulic gradient. Generally, clays
have low permeability or low water-transmitting capacity,
and coarse gravels have high permeability.

Ground water may move through a confining bed when
there is a difference between the head of the water across
the bed. When the water table is higher thanthe piezometric
surface of the confined water, the potential leakage is down-
ward (recharge to the artesian aquifer). If the piezometric
surface of the confinedwater is higher than the water table,
the potential leakage is upward (discharge from the artesian
aquifer). The rate of ground-water movement through the
confining bed depends on (1) its vertical permeability and
divided by the thickness of the bed).

The ground water of the Green Swamp has been divided
into two general classes based on its occurrence: (1) non-
artesian ground water, which occurs in the undifferentiated
plastic deposits, and (2) artesian groundwater, which occurs
in the porous limestones of the Floridan aquifer (fig. 18).

Nonartesian Ground Water

Occurrence: In the Green Swamp area the nonartesian
ground water occurs primarily in an unconfined aquifer
composed of undifferentiated plastic deposits, mainly sand
and clayey sand. The nonartesian aquifer ranges from 0 to
more than 100 feet in thickness. The base of the aquifer is
marked by either a dark silty clay which ranges from 0 to
more than 50 feet in thickness, or a light green silty
montmorillonitic clay which ranges from 0 to more than 30
feet in thickness. The major portion of the nonartesian
aquifer consists of fine to coarse grained quartz sand with
varying amounts of multi-colored kaolinitic clay. The non-
artesian aquifer is thick in the eastern part of the area and
thin or absent over the limestone plain (exposures of the
Floridan aquifer) in the western part.

S100

zc-i
Ja

i..-i

w 50
-I
0

>W
wr
-1
'a

0 I 2 4 6 8 10 miles
i i

Figure 18. Generalized hydrologic cross section along line B-B', figure 6.

FLORIDA GEOLOGICAL SURVEY

Data obtained during test drilling showed that the clay
at the base of the nonartesian aquifer is of sufficiently low
permeability to confine water inthe Floridan aquifer (fig. 18).
The dark clay in the eastern part of the area does rot extend
across to the western part of the area. A light green clay
forms the confining bed in the western part but appears to
interfinger with the clayey sands of the nonartesian aquifer
to the east.

Fluctuations of the Water Table: The water table fluc-
tuates with changes in ground-water storage in the same
manner as the water surface of a reservoir or lake that
rises when water is added (recharge) and- falls when water
is withdrawn (discharge). When the rates of recharge and
discharge are equal, the water surface does not fluctuate
but when they are unequal, the water surface rises or falls.
Fluctuations of the water table were recorded in several
shallow wells within the area but the records are of insuf-
ficient duration to study the correlation between the water
table, surface-water flow, and artesian water levels. The
only well inthe area with a long-term record of water-table
fluctuations is well 810-136-Z (Polk County 47). The well is
located north of Davenport and east of the Green Swamp
area (fig. 19) in an area where flowing artesian wells are
found. The hydrograph of well 810-136-2 (fig. 20) shows
that the water-table fluctuations ranged about 6 feet during
the period 1948-59. Comparison of the hydrograph of well
810-136-2 (nonartesian) with that of well 810-136-1 which
taps the Floridan aquifer shows that the water table fluctu-
ated through about 1 foot greater range than the piezometric
surface did. The general correlation between the two
hydrographs seems to indicate that a relatively good hydraulic
connection between the nonartesian aquifer and the Floridan
(artesian) aquifer occurs nearby. Stringfield (1936, p. 148)
noted the apparent lack of confining beds between the non-
artesian and artesian aquifers in this area.

Water-table fluctuations were recorded in four wells
in the Green Swamp area during 1959. They are: wells
832-154-2 and 822-149-2 in Lake County, well 822-138-2
in Orange County; and well 813-149-2 in Polk County (see
fig. 19 for locations). The hydrographs of the water table
at wells 813-149-2, 822-149-2, and 822-138-2 showed ap-
proximately the same range of fluctuation and -the same

UNITED STATES DEPARTMENT OF THE INTERIOR
GEOLOGICAL SURVEY

ah Ir Li.e
a ~APOPKIA

1~ ~take

S MASCOTTE \ Apopho

-835 W WINTER
le Mkmeo CLERMONT GARDEN
833- 'r. ~ 324-
03154 GROVELANDI S, L_4
R~oer ,4 (1 take
i "Umnehoho,
S832-154-2
0- C'r/hn
83 8 -0-1 -I ains
/.8 2 9 2 0 -- 1 9 5 1, h2 14 6 S S P
82 2-1 oke
I O Y Sp.82BS 154-I as ->
S C 0 NT~n i oryg
821209- COI'NTY
-s2o-2 827-158-I
Lannonhee- -I -~, N t
T,rlIby 0 S o' Cree5
8,6_811- IA Se, AIOt

82p 8.5-151-1 I i-i
so \c 824-142-1, OQ
82220 S',
.2 o1 201 I- r S149 2,,
-2 8 -138-1
DA D CUMPRCSSCJ2 R, 821-158-2822
-21202-1 S,
CDUII I ORANGE COUNIl

YoLKR < D OUN SCELA COUNI)

29 Ridge e
81POLIO \ N
Rlchlon X aiode~
816-2-1 2Pb'Q 0E L A

-2G,92~

Bi 13149-i i. I 036

L '0- COUNTY S ~ ~ ~ ,P~ 0-144-1 Davenport
813-210-1 810-1512 t r; \ Sp

SI-LOB-155-1

*~~~~ HAINESO -J ff
~CITY rCITY
22

PLANT 24'LAKE
AeLA ALPNEO %AINES 000
C TyC

J Jl

ft

LS~ AI.Ai- I

924-200-i
Well (.eIlloomer I5 lehe fiel
IllileOP nei cobler' III 050fl41

-be, .oSedIroe .0-i Pr q

5hallo obse- loon ell f-l. )l-
lie nOnofllesOl 04,1,e

P Soe -,oll
S S .iq.
D ~la~!:ou~i;r .0-aS YI~

SDUedortl 01 Seen Snoop 0100

COYt or town

0 1 2 3 4 Sniles

Figure 19. Map of Green Swamp area showing ground-water
data-collection points.

~

INFORMATION CIRCULAR NO. 26

GROUND WATER LEVELS FROM SELECTED WELLS

Woatr lvel flucluotlon l

Well 810-136-2 (NonorlHin oqufr)

4 115 h l
i iJ ttI
.. /! ,.

o 10 1165
. ...1 ....*. :

I 8 21 I Well 810-1F3-1 (FLoden qo

Akt J A I I I 1

W.9 816-2 1)-1Floridan Mquller)

I
1

E

X
c-

54 -- -- - -. - - - - 15 .

-_ I A_ _- A I5
rT5
-'

2 I1

6 35

267 -- --1--r -- t--75f-

Figure 20. Hydrographs of long-term water-level records
from wells near the Green Swamp area.

FLORIDA GEOLOGICAL SURVEY

head relationship to the piezometric surface; therefore, the
hydrograph of well 813 -149-2 was chosen as representative
of the other wells.

Figure 21 shows a comparison of recorded fluctuations
of the water table and piezometric surface with recorded rain-
fall at wells 832-154-1 and2 (Lake County), 821-202-3 (Sumter
County, and 813-149-1 and 2 (Polk County). All hydrographs
show rapid rises following heavy rains. The hydrograph
of well 813-149-2 shows that the water table remained near
the surface and that downward leakage to the Floridan
aquifer was maintained during 1959. The hydrograph of
well 832-154-2 (nonartesian) shows a slightly greater range
in fluctuation of the water table than in well 813-149-2
(nonartesian). This probably is due either to pumping from
a drainage canal near well 832-154-2 or the proximity of
the well to an abrupt change in the land surface. The hydro-
graphs of 832-154-2 and 832-154-1 (artesian) show that the
water table fluctuates above and below the piezometric
surface. However, the indicated direction of ground-water
movement is predominantly from the artesian to the non-
artesian aquifer. A conductivity recorder installed in well
832-154-1 showed a decrease in conductance in the water
from the Floridan aquifer within 2 or 3 days following a rain.
The conductance indicates the mineral concentration in water.
Therefore, decreased conductance in this well implies
dilution of groundwater in the Floridan aquifer by flow from
the nonartesian aquifer which normally contains less highly
mineralized water. Apparently the zone that separates the
two aquifers is sufficiently permeable to permit significant
amounts of ground water to move between the aquifers.

Comparison of the water-level fluctuations in well
832-154-2 with those of well 821-202-3 shows that the range
of fluctuation and rates of recession of the water table and
piezometric surface of the Floridan aquifer are similar
because the Floridan aquifer is exposed in the vicinity of
well 821-202-3, and locally the aquifer is unconfined. The
thin, saturated zone of the nonartesian aquifer that occurs
locally above the unsaturated portion of the Floridan aquifer
is called a "perched" aquifer.

INFORMATION CIRCULAR NO. 26

0 orrlaon auifer) 1 04

82 ,0.,j ,o2
, ,,,, 5 1-14-1 ,___ i 103

4 832-154-2 100
5 _(Nonartesion aquifer) 99
Located about 3 miles south of Moscotle
4

0 ,132
813-149-2 (Nonorfe8siu rqu3r) 13

S13- 149-I 12
-cS s

S 4 --- -- --- -- -= 1 B1-149-1 --- -- --. ~'128
(Floridon oquifer)27
5L 127
Located about 25 miles north of Polk City *
3

o ,
0a

Z 3 3i 9

682 9
21-202-3 (Floridn aquiferin unconfined are)
1. Located at Cumpressco

I I
._U 2

JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC
1959

13
12

i)
0
A4

Figure 21. Hydrographs of water-level and rainfall records

at wells in the Green Swamp area.

FLORIDA GEOLOGICAL SURVEY

Recharge and Discharge: Ground water in the non-
artesian aquifer within the Green Swamp area is replenished
principally by local rainfall. It is discharged by: (1) flow
into surface-water bodies, (2) evapotranspiration, (3) down-
ward leakage into the Floridan aquifer, (4) subsurface
discharge through the nonartesian aquifer, and (5) draft by
a few wells.

In the eastern part of the Green Swamp area the non-
artesian aquifer is thick and has a relatively large storage
capacity, and the water table stands at high elevation. In
the western part of the area the aquifer is thin or absent
and accordingly has a small storage capacity.

Movement of nonartesian ground water within the aquifer
is from areas of high head to areas of low head. Conse-
quently, ground water moves from areas where the water
table is high in the eastern and southern parts of the Green
Swamp area to areas where the water table is low in the
western part. Ground-water flow from the east tends to
maintain a shallow water table in the west so that part of
potential recharge from rainfall in the western part is
rejected. The direction of ground-water movement indicates
that the eastern part is a recharge area.

The topographically higher sand ridges in the eastern
and southern parts of the area have a large ground-water
storage capacity because of their thickness. However,
most of the high sand ridges (such as the Lake Wales Ridge)
do not store large quantities of ground water. The water
table beneath the ridges was found to occur at approximately
the same elevation as the regional water table onthe eastern
and western sides of the ridge. Therefore, as much as 100
to 200 feet of the deposits are unsaturated beneaththesehigh
sand ridges.

The water table ranges in elevation from about 100 to
130 feet above mean sea level in the eastern and southern
parts of the area and from 75 to 100 feet above mean sea
level in the western part. The slope'of the water table
conforms generally to the topography. Surface-water bodies
in sandy areas generally indicate places where the land
surface intersects the water table and ground-water divides

INFORMATION CIRCULAR NO. 26

coincide generally with surface-water divides. See figure 7
for drainage divides. During dry periods some of the
ground-water divides may shiftposition, or even disappear,
because of changes in direction of ground-water movement.

Artesian Ground Water

Occurrence: The Floridan aquifer, the principal source
of artesian ground water in Florida, underlies all of the
GreenSwamp area. Thetop of the aquifer is at land surface
in the western part of the area and occurs at depths of more
than 200 feet below land surface in the eastern part. The
actual thickness of the aquifer is not knownbecause the test
wells have penetrated only 100 to 200 feet of the aquifer.
However, a relatively impermeable zone of undetermined
extent occurs in other wells at a depth of about 1,000 feet,
perhaps forming the base of the aquifer. Thus, the thick-
ness is probably on the order of 700 to 900 feet.

Fluctuations of the Piezometric Surface: Ground water
in the Floridan aquifer in the Green Swamp area is replen-
ished (recharged) principally by local rainfall that perco-
lates downward from the surface of the ground through the
nonartesian aquifer and by direct percolation into the exposed
portions of the aquifer. Groundwater in the artesian aquifer
is discharged by (1) subsurface outflow to areas of lower
piezometric head beyond the limits of the Green Swamp
area, (2) seepage and spring flow into the surface-drainage
systems, (3) leakage to the nonartesian aquifer, (4) evapo-
transpiration, and (5) pumping. The piezometric surface
rises when the rate of recharge exceeds discharge and
declines when that of recharge is less than discharge. The
fluctuations of the piezometric surface were observed by
continuous water-level recording instruments and by periodic
water-level measurements in selected wells. No long-term
records of fluctuations of the piezometric surface are avail-
able for wells within the Green Swamp area, but long-term
records are available for nearby wells 810-136-1 (Polk
County 44) and 816-211-1 (Pasco County 16). Hydrographs
of the water level in these wells are presented in figure 20.

FLORIDA GEOLOGICAL SURVEY

The two hydrographs indicate that the range of fluctuation
of the piezometric surface was greater in western Pasco
County than in northeastern Polk County during the period
1947-59. The piezometric surface at well 816-211-1 fluc-
tuated from about 56 to 74 feet below land surface and at
well 810-136-1 it fluctuated from about 0. 5 to 5 feet below
land surface. The large fluctuation in well 816-211-1 is
caused by its proximity to local recharge and discharge
areas near the Hillsborough and Withlacoochee rivers. Also,
well 810-136-1 is located in an area which has a relatively
different type geology from that of well 816-211-1 because
the top of the aquifer is deeper. The hydrographs of well
810-136-1 and a nearby water-table (nonartesian) well
810-136-2 show that the piezometric surface is usually
higher than the water table (fig. 20). Therefore, well
810-136-1 is in a potential artesian leakage area.

Short-term water-level records are available for wells
832-154-1 and 822-149-1, in Lake County; well 822-138-1,
in Orange County; well 821-202-3, in Sumter County; and
well 813-149-1, in Polk County (fig. 19). The water-level
records of well 813-149-1 show the same general range of
fluctuations of the piezometric surface as wells 822-149-1
and 822-138-1. The wells are located in areas where the
water table is above the piezometric surface and the Floridan
aquifer occurs at about the same depth. Therefore, the
water level in well 813-149-1 is probably representative of
artesian water levels in the eastern part of the Green Swamp
area.

Hydrographs comparing the fluctuations of the piezo-
metric surface-of the confined water in wells 832-154-1,
821-202-3, and 813-149-1 with fluctuations of the water
table and rainfall are presented in figure 21.

The fluctuations of the piezometric surface in the three
wells showed the same general trend, inthat rainfall caused
a rise in water level and lack of rainfall caused a decline in
water level. Differences in the range, of fluctuation, rates
of recession, and rate of responseto recharge and discharge
are apparent if the hydrographs are compared from east to
west.

INFORMATION CIRCULAR NO. 26

Well 813-149-1 is located on the piezometric high in
the southeastern part of the Green Swamp area. The water-
level rise caused by rainfall continues for about 2 days
after the rain indicating slow seepage. The subsequent
water-level recession occurs at a rate of about 1 foot in 30
days. The maximum range in fluctuation is about 2 feet;
however, an increase in recession during December 1959,
which is indicated by steepening of slope, was probably
caused by increased pumping of ground water from the
aquifer during the citrus season.

Well 832-154-1 is between the recharge areas in the
eastern part and discharge areas in the western part of the
Green Swamp area. The peaks caused by rainfall occur
about 1 day after the rain. The peaks are sharper than
those of well 813-149-1 and the subsequent recession occurs
at a rate of about 1 foot in 20 to 30 days. However, the
initial slope of recession is greater for well 832-154-1;
this may be due partly to greater intensity of rainfall and
partly to the proximity of the well to local recharge and
discharge areas. The flow of ground water increases and
decreases in the same manner as water flowing in a stream
with steep gradients and a well defined channel. The range
of fluctuation of the water level in well 832-154-1 is about 2
feet and it is about the same as the range inwell 813-149-1.
Another difference between the hydrographs of these two
wells is that the water level in well 813-149-1 had a steeper
recession curve than well 832-154-1 during December 1959.
This is probably related to effects of local pumping.

Well 821-202-3 is located in the western part of the
Green Swamp area. The water-level peaks caused by rain-
fall are steeper thanthose inthe wells in the eastern part of
the area and they occur about 1 to 2 days after the rainfall.
The subsequent recession rate is about 1 foot in 10 to 15
days. On November 21, 1959, no rainfall was recorded at
well 821-202-3 but the piezometric surface responded to
rainfall in the same manner as the water level in wells
813-149-1 and 832-154-1. This indicates that there is
probably good hydraulic continuity in the aquifer from east
aquifer. The range of fluctuation of thepiezometric surface

FLORIDA GEOLOGICAL SURVEY

at well 821-202-3 is about 3 feet. The greater rate of
recession and range of fluctuation at well 821-202-3 could
be due to the combined effects of (1) nonartesian conditions
in the Floridan aquifer while artesian conditions occur at
the other two wells (discussed in nonartesian section),
(2) direct entry of rainfall through exposed portions of the
aquifer, and (3) the proximity of the well to surface-water
drainage.

The fluctuations of the piezometric surface in well
813-149-1 and the water table in well 813-149-2 with rain-
fall show the same general trend, but the position of the
water table is above the piezometric surface. This suggests
that the wells tap two separate but interconnected aquifers.
The direction of ground-water movement is from the non-
artesian aquifer to the Floridan aquifer.

The fluctuation of the piezometric surface in well
832-154-1 and of the water table in well 832-154-2 with
rainfall do not coincide, but show the same general trend.
The hydrographs of the piezometric surface and the water
table cross several times, and the rate of water-table
recession is much greater than that of the piezometric
surface. This indicates that the wells penetrate separate
aquifers and the water level fluctuates in responseto chang-
ing recharge and discharge conditions in each aquifer. The
direction of ground-water movement between the aquifers
changes, but the predominant direction is from the Floridan
to the nonartesian aquifer. The greater rate of recession
of the water table may be due partly to a nearby drainage
ditch which is pumped intermittently.

Fluctuations of the water level in the nonartesian
aquifer were not recorded near well 821-202-3, therefore
no comparison can be made. The relatively thin section. of
nonartesian aquifer (sand and clay) in this area could store
little water. The low permeability of the clay confining bed
is probably one of the factors that causes increased surface
runoff in the western part of the Green Swamp area.

Shape of the Piegometric Surface: The piezometric map
(fig. 22) shows the shape of the piezornetric surface of the
Floridan aquifer in the Green Swamp and nearby areas.

F \PL 11\I.ON

h\ell
95
Num.,e n Ihe watrl le>et,
,1 feel oDoe mean see
level ep -Oct 1959

o8---
Co-nf ,,he represen's tle
'' of the perneletrec
iuloce, in fee obove
neon sea level Doshed
Cor LOur represent m-n
fered posilien of co-
i ou

Contou intlerval 10feel

EC 3
City or town

0 1 2 3 4 5 mtles

Figure 22. Map of Green Swamp area showing the shape of

the piezometric surface of the Floridan aquifer.

INFORMATION CIRCULAR NO. 26

The map was constructed from water-level measurements
made, during the period September through October 1959,
in selectedwells that were cased into the aquifer. The water
level in eachwell was referred to mean sea leveldatum and
contour lines were drawn to connect points of equal piezo-
metric head. The direction of ground-water movement in
the aquifer is perpendicular to the contour line from areas
(sometimes referred to as mounds) usually indicate principal
areas of recharge to the aquifer. Piezometric lows (some-
times referred to as troughs) usually indicate areas of dis-
charge from the aquifer. The distance between the contour
lines indicates the hydraulic gradient of the piezometric
surface. The hydraulic gradient may vary because of unequal
amounts of recharge or discharge, changes in permeability
of the aquifer, or changes in thickness of the aquifer.

Figure 18 shows an east to west hydrologic cross
section through the central part of the Green Swamp area.
On the eastern side, the top of the aquifer occurs between
100 to 200 feet in depth beneath a relatively thick, non-
artesian aquifer. The thickness and type of material com-
prising the confining beds and the position of the piezometric
surface of the confined water andwater table indicate that a
good hydraulic connection exists between the two aquifers.

The general direction of ground-water movement in the
Floridan aquifer is outward in all directions from an elon-
gated piezometric high that extends approximately from
central Lake County to southern Polk County at least. The
apex of the high (130-foot contour) occurs within the south-
eastern quarter of the Green Swamp area. The northern
extent of the elongated mound lies within the eastern half of
Green Swamp (fig. 22). The southern boundary of the Green
Swamp area extends across the piezometric high.

The direction of ground-water movement in the Green
Swamp is eastward to discharge areas in Orange, Osceola,
and eastern Polk counties, westward to local discharge
areas in the western half of the Green Swamp area, south-
ward toward areas of lower piezometric head in west-central
Polk County and northeastern Hillsborough County, and
northward into southern Lake County.

FLORIDA GEOLOGICAL SURVEY

The 130-foot contour line (fig. 22) encloses an area of
high piezometric head in the Floridan aquifer. The mound
near Polk City is in a moderately flat, poorly drained area.
The land surface ranges in elevation from about 130 to 150
feet above mean sea level and there are numerous cypress
swamps and lakes. The Floridan aquifer is overlain by
about 100 feet of sand and clay of which the basal 10 to 50
feet is considered to act as a confining bed. Recharge to
the Floridan aquifer within the high is believed to occur by
downward percolation through the confining bed and also
through the many cypress swamps and lakes which appar-
ently mark sinkholes in the underlying limestone.

The hydraulic gradient, as indicated by the distance be-
tween contour lines, is greater toward the east than toward
the west, thereby inferring a greater rate of ground-water
movement eastward. If it is assumed that the permeability
and thickness of the aquifer are uniform, thenthe discharge
to the east would be two to three times greater than the
discharge to the west.

The difference in hydraulic gradient could be caused by
(1) a decrease in permeability in the aquifer eastward, (2) a
decrease inthethickness of the aquifer eastward, (3)water-
table conditions occurring within the Floridan aquifer in the
western part of the area, (4) barrier effects of structure,
(5) pumping to the east, or (6) by combinations of these
conditions.

The piezometric map shows numerous small recharge
mounds in the western part of the Green Swamp. These are
probably caused by rainfall that occurred immediately prior
to the period of water-level measurements and will dissipate
after a short time. Local runoff and recharge causes large
fluctuations of the piezometric surface in the western area.

Figure 22 indicates that there is considerable ground-
water discharge from the Floridan aquifer into the Withla-
coochee River downstream (north) from Dade City. The
Floridan aquifer also discharges water into the Hillsborough
River downstream from the diffluence of the Hillsborough
and Withlacoochee rivers, and some ground water is dis-
charged into the Withlacoochee River upstream from the

INFORMATION CIRCULAR NO. 26

diffluence. There is also evidence that ground-water flow,
as well as surface-water flow, is diverted from the Withla-
coochee River through the Withlacoochee-Hillsborough over-
flow.

Pumping in the area north of Lakeland is indicated by
cones of depression in the piezometric surface. There are
trough-like lows developed on the piezometric surface near
the Hillsborough River that probably indicate rapid move-
ment of ground water to the discharge areas by means of
solution-widened fracture zones. Discharge of groundwater
from the Floridan aquifer occurs by upward leakage to the
Reedy Creek and Lake Marion Creek areas on the eastern
flank of the Green Swamp. Some ground water from the
northern part of Green Swamp area moves northward and
is discharged into tne Palatlakaha Creek basin.

Chemical Characteristics of Ground Water

Information concerning the quality of ground water was
obtained by field observations and from the analyses of
water samples collected from wells within the Green Swamp
area.

The conclusions concerning the quality of water are
based for the most part on the data collected during the
reconnaissance of November 1959. During this reconnais-
sance all of the deep observation wells were sampled by
pumping. All samples that contained suspended matter were
filtered in the field. These data were supplemented by data
collected during the period of investigation and during pre-
vious investigations.

The quality is discussed by comparing the water char-
acteristics inthe area of the piezometric high (southeastern
area) with those for the remaining area.

In general, the mineral content of water in the Floridan
aquifer along the southern boundary of the area was less
than 150 parts per million and that for the remaining area
was about 300 parts per million. Figure 22 indicates that
water enters the ground and the Floridan aquifer in the

FLORIDA GEOLOGICAL SURVEY

southeastern part of the area; that is, the ground water in
this area is being replenished by surface water. This could
account for lower mineral content. The water in the Floridan
aquifer adjacent to the lakes was generally lower in mineral
content than that more distant. This could indicate downward
movement of the surface waterfromthe lakes to the aquifer.

The mineral content in the nonartesian aquifer generally
was higher than the mineral content inthe water on the land
surface and lower than the mineral content of the Floridan
aquifer. This could have resulted from prolonged leaching
of the shallow sediments, increase in contact time, and
increase in area of contact of the water with soluble minerals
under the ground. Generally the mineral content in the
water in the nonartesian aquifer was less than 125 parts per
million.

In the southeastern area, where the nonartesian aquifer
overlies the Floridan aquifer, it contains beds of calcareous
sandstone. This results in water of variable mineral content
within this area; that is, two wells may be drilled to the
same depth and within the same general area, yet the water
in one well may have been in contact with a calcareous bed,
or other soluble material, while the water in the other well
may not have been. Also, two wells may penetrate similar
formations and one of these could receive more water
directlyfrom surface sources thanthe other and their water
characteristics would be dissimilar.

In the western area, the mineral content inthe artesian
aquifer is more uniform than in the southeastern area. The
mineral content during the November reconnaissance ranged
from 269 to 350 parts per million.

In the western area the Floridan aquifer occurs at or
near the land surface. Water at shallow depths in this area
is more highly mineralized than water at depths of several
hundred feet in the southeastern area.

Previous investigations indicate that the sinkhole lakes in
the Green Swamp, particularly inthe southeastern area, may
be contributing recharge to the Floridan aquifer (Stringfield,
p. 148). A few samples collected from wells in this area

INFORMATION CIRCULAR NO. 26

tend to bear this out. One sample collected from a well that
was 200 feet deep had a mineral content of only 15 parts per
million. This low mineral content is not much greater than
that in rain water.

Low mineral content occurs in the waters of the non-
artesian aquifers. This low content may not be a result of
the rapid movement of water but possibly may be due to the
absence of soluble material.

SIGNIFICANCE OF THE HYDROLOGY OF THE AREA

The findings of this reconnaissance maybe summarized
by appraising the hydrology of the area and bytheorizing on
some of the changes in the hydrology that could result from
increased drainage or from water conservation in the area.

Much of the project area consists of rolling hills and
flatlands at relatively high elevation. These areas are
presently developed for agricultural uses. Marsh areas
have limited or no agricultural use because of the high
water table and poor surface drainage. Therefore, any plan
of proposed water management should consider the use of
the land of the area.

Information collected thus far indicates that thehighest
piezometric levels of the Floridan aquifer in central Florida
occur in the southeastern part of the Green Swamp area.
This is an area of high topography and consequently the
origin of most streams that drain the area. Thus the south-
eastern area may be considered as the headwaters both for
the drainage systems and for the principal ground-water
aquifers in much of central Florida.

Because the hydrology of the area east of the Seaboard
Air Line Railroad is somewhat different from that of the
area west of the railroad, the important hydrologic features
of the two areas are discussed separately in the following
sections.

88 FLORIDA GEOLOGICAL SURVEY

Eastern Part

Drainage systems east of the Seaboard Air Line Railroad
are poorly defined and the surface drainage is inefficient.
Most of the rainfallis disposed of by evapotranspiration and
seepage intothe ground. The remainder runs off as stream-
flow.

The annual potential natural water loss for the upper
Palatlakaha Creek basin is 50 inches. Little or no runoff
would occur from the basin when the annual rainfall is 30
inches or less, assuming normal intensity and distribution.

Ground-water recharge in the eastern part of the area
is fromthe surfacetothe nonartesian aquifer to theFloridan
aquifer. Both surface and ground water are low in mineral
content. The predominant directions of movement of ground
water in the Floridan aquifer are toward the north and the
east. Ground-water gradients are steeper east of the piezo-
metric high than those on the west, indicating that more
ground water apparently moves eastward than westward.

The drainage systems have been improved by many
miles of canals and ditches. These improvements have
caused minor changes in the distribution of runoff by the
diversion of water from the Palatlakaha Creek basin to the
Withlacoochee River headwater tributaries which carry it
to the western area. Breaks in slope of the double-mass
curves in figure 16 indicate that significant diversions
between the two basins started about 1954.

Western Part

West of the Seaboard Air Line Railroad, in the Green
Swamp area, the drainage systems are better defined than
those in the eastern part and drainage is somewhat more
efficient. Many miles of canals and ditches have improved
the drainage in local areas but have not significantly changed
the total runoff from the western area.