i > I ,
December 1984 Bulletin 840 (technical)
Estimated and Measured
Florida Climate, Crops, and Soils
J. W. Jones, L. H. Allen, S. F. Shih, J. S. Rogers,
L. C. Hammond, A. G. Smajstrala, and J. D. Martsolf
.000 n o o 0 o 00
00 0 0 0, o0ooo o
Agricultural Experiment Stations
Institute of Food and Agricultural Sciences
University of Florida, Gainesville
F. A. Wood, Dean for Research
_ -. and title page: "A. G. Smajstrala" should read "A. G. Smajstrla."
"Title page: Add: "Dr. Hammond is a Professor in the Department of Soil
Science, University of Florida."
Page 3, line 8: "Florida aquifer" should read "Floridan aquifer."
"Iage 3, line 34: "operations" should read "operation."
Last line of first paragraph should read: "essentially no soil E
would be reached."
: 11, lines 12, 14. 15, 16, and 18: "T" should read "7T,,."
1, Equation 4: "(1.8T + 48)" should read 1.8T,,., + 48 ."
i, last paragraph, line 3: Insert "(a = (.23)" at end of sentence.
"Equation 15 is on line 15.
lines 3-5 should read: "Van Bavel and Verlinden (1956) used
a = 0.05 in Equation 10 to obtain E,, and found that k, = 0.7 in
Equation 18 provided a good estimate of ET, for a well-watered
19, line 19: "the difference" should read "the previously described
19, footnote to Table 4. line 3: Insert "(E,,)" after "surface."
age 20, paragraph 2, lines 7-11 should read: "The water budget method
balances water input to an area with water output from the area
during a specified time period. Rainfall (RF) plus irrigation (IR)
is equated to the change in soil water storage (AS) plus water
output in the form of ET, surface runoff (RO), and percolation
from the root zone (PN)."
S20, Equation 19: + AS" should read -AS."
ge 23, fourth line from bottom: "17 mm more" should read "17 mm
30, line 1: Add parentheses as follows: "The ET data presented for
pasture (turfgrass ET data)... ."
30, line 18 should read: "practices significantly reduced ET."
,e 35, Figure 9: "ET,," should read "E,."
Page 36, Figure 10a: Broken lines are k[. values.
Page 41, line 7 should read: "where = relative yield."
Page 43. Figure 14: In the Y value. "328" should read "3.28."
Page 48, Table 11: In the column headings, remove footnote designation
"1" from "Effective rainfall" and add it to "Predicted." After
"Irrigation actually applied," footnote number "1" should read
,last line should read: "of day-to-day water management princi-
ples in humid regions."
5, lines 5, 6 should read: "when soil water does not limit transpira-
ion. .. "
,line 25: "megragram" should read "megagram."
lines 2 and 3 should read: "Pp. 293-296 in Proceedings, Eighth
International Grassland Congress (C. L. Skidmore, ed.). Alden
Press, Oxford, England."
ESTIMATED AND MEASURED
)RIDA CLIMATE, CROPS, AND SOILS
SW. Jones, L. H. Allen, S. F. Shih, J. S. Rogers,
C. Hammond, A. G. Smajstrala, and J. D. Martsolf
Authors, acting as a committee, prepared this bulletin from their
tive research experience on a number of crops in Florida.
Jones and Dr. Shih are Professors and Dr. Smajstrla is an Associ-
rofessor in the Department of Agricultural Engineering, University
)rida. Dr. Rogers is an Adjunct Assistant Professor in the Depart-
of Agricultural Engineering and an Agricultural Engineer with the
Department of Agriculture, Agricultural Research Service. Dr.
i is an Adjunct Associate Professor in the Departments of Agron-
ind Fruit Crops, University of Florida, and a Soil Scientist with the
Department of Agriculture, Agricultural Research Service. Dr.
4 f is a Professor in the Department of Fruit Crops, University of
1.0 Introduction ............................................... 1
2.0 Evapotranspiration Concepts ................................... 4
2.1 Climate Dependence ....................................... 4
2.1.1 Evapotranspiration and the Energy Budget............. 4
2.1.2 Potential Evapotranspiration. .......................... 6
2.2 Surface Dependence of Evapotranspiration .................. 7
2.2.1 Influence of Plant Canopy and Physiological Stage of
Growth on Actual Evapotranspiration .................. 8
2.2.2 Seasonal Evapotranspiration of Crops .................. 9
2.3 Techniques Used to Calculate Potential Evapotranspiration ..... 10
2.3.1 Penman Method ................... .... ............ 10
2.3.2 Pan Evaporation Method. ............................ 14
2.3.3 Thornthwaite Method. ............................ 14
2.3.4 Blaney-Criddle Method ............................... 15
2.3.5 Modified Blaney-Criddle Method using Solar Radiation... 15
2.3.6 Stephens-Stewart Method ............... ............. 17
2.4 Potential Evapotranspiration for Various Regions in Florida .... 17
3.0 Experimental Verification of Evapotranspiration .................. 20
3.1 Basin-Wide Water Budget ................... .............. 21
3.1.1 Annual Water Budget ............................... 21
3.1.2 Monthly Water Budget ........................... 23
3.2 Experimental Plot Water Budget Verification ................. 27
3.2.1 Monthly Water Budget ........... .............. 27
3.2.2 Seasonal Crop Evapotranspiration .................... 30
3.3 Crop Coefficients ..................................... 33
4.0 Application of Evapotranspiration Concepts to Irrigation
4.1 Irrigation to Prevent Crop Yield Reductions .................. 38
4.1.1 Dry Matter Yield ............. ................ 38
4.1.2 Grain Yield ................................... 39
4.2 Effective Rainfall ..................................... 44
4.3 Irrigation Requirement .................. ............... 47
4.4 Example of Irrigation Management Concepts for Humid Regions 49
5.0 Summary............ ............................. 52
Appendix I. Notation ........................................... 55
Appendix II. U.S. Customary to SI Conversion Factors ............... 59
LIST OF TABLES
1 Cloudless sky daily solar radiation received at earth's surface,
averaged for each month for three latitudes ................. 12
2 Pan coefficients for Class A pan for different groundcover, levels
of mean relative humidity, and 24-hour wind ................ 16
3 Application of Penman method to calculate free water evapora-
tion and monthly potential evapotranspiration for three locations
in Florida .............................................. 18
4 Calculated seasonal potential evapotranspiration by Penman
method for three locations in Florida ....................... 19
5 Comparison of several long-term basin-wide water budgets on an
annual basis ................... ........................... 22
6 Average monthly climatological data for the Belle Glade, Florida,
weather station ........................................ 24
7 Monthly evapotranspiration estimated by different methods for
the Everglades Agricultural Area for 19 years ............... 25
8 Evapotranspiration deviation between water budget data and
model predictions for the Everglades Agricultural Area ....... 26
9 Water budget estimates of monthly evapotranspiration for citrus,
pasture, sugarcane, rice, and peaches for various locations in
Florida .............................. ................. 29
10 Average monthly effective rainfall as related to mean monthly
rainfall and average monthly consumptive use ............... 46
11 Example predictions of effective rainfall and irrigation
requirements for the Everglades Agricultural Area, using average
weather conditions from 1924 to 1975 ........................ 48
LIST OF FIGURES
1 Schematic diagram of a general hydrologic cycle .............. 1
2 Typical rainfall distributions for three locations in Florida ...... 2
3 Energy balance at 2:00 p.m. EST, April 29, 1978, for a short
grass pasture near Okeechobee, Florida ...................... 6
4 Five-day averages of potential evapotranspiration for turfgrass
and measured evapotranspiration from Tifway bermudagrass in
30-cm water table lysimeters ........................... 28
5 Evapotranspiration for corn .............................. 31
6 Evapotranspiration for bahiagrass ......................... 32
7 Evapotranspiration for sorghum ............................ 33
8 Evapotranspiration for tomatoes .......................... 34
9 Evapotranspiration for peanuts ........................... 35
10 Crop coefficients for (a) citrus and (b) pasture and turfgrass
reported by USDA-SCS for a modified Blaney-Criddle approach
applied to estimate ET, (kc) and those calculated from data
presented in this report, using Penman estimates of ET, (k'c) .. 36
11 Crop coefficients for corn reported by USDA-SCS for a
modified Blaney-Criddle approach to estimate ET, (kc) and those
calculated from data presented in this report, using Penman
estimates of ET, (k'c) ........................... ..... .. .. 37
12 The relationship between crop dry matter yields and water use 40
13 The relationship between grain yield and estimated ET for corn
grown in deep sandy soil in north Florida .................. .42
14 The relationship between crop yield and estimated ET for
peanuts grown in deep sandy soil in north Florida ............ 43
15 The relationship between crop yield and estimated ET for
soybeans grown in deep sandy soil in north Florida ........... 44
Evaporation equals precipitation in the global hydrological cycle.
However, the pathway of water movement is different for various sur-
faces over the earth. Precipitation exceeds evaporation over lands in
humid climates (such as Florida) since water flows back to the seas
(Figure 1). Therefore the seas, which are sinks for fresh water from land
areas, lose more water by evaporation than they receive from precipita-
The Hydrologic Cycle
r?'-.. .... [T ( Clo.ud Formation
Fig. 1. Schematic diagram of a general hydrologic cycle.
In Florida, part of the rainfall over the land areas may goimmediately
into runoff and part into soil water storage. Much of the water that goes
into the soil is partitioned into evaporation and transpiration. The re-
mainder goes into seepage into streams, percolation into shallow ground-
water, or deep aquifer storage. Even this stored water eventually flows
back to the seas. Figure 1 illustrates this general hydrologic cycle.
Florida has a unique hydrologic cycle among the southeastern states.
First, the average pattern of annual rainfall distribution (Butson and
Prine, 1968), especially of the peninsula, is different (Figure 2). The
300- Milton (M)
3 Lakeland (L)
300- Hialeah (H) --
Fig. 2. Typical rainfall distributions for three locations in Florida (from NOAA,
heaviest rainfall in the peninsula usually occurs during June through
September or October, when rainfall exceeds evapotranspiration. Most
runoff also occurs during this period. The drier months typically include
October or November through May. Much of the rest of the southeast has
a winter wet season of December through March when rainfall exceeds
evapotranspiration and most of the groundwater recharge and runoff
Second, Florida has sandy soils with high infiltration rates. About
one-third of these soils are upland with deep water tables (below approx-
imately 180 cm) and the remaining two-thirds are on flat topography with
shallow water tables (surface to about 180 cm). In the flatwoods topogra-
phy, runoff from storms is low until the soil is saturated. This factor
makes years of low precipitation particularly severe for wetlands and for
lakes such as Lake Okeechobee (Allen et al., 1982). The major rivers of
the Florida peninsula (the Kissimmee, the Caloosahatchee, the St. Johns,
and the Peace) depend in large measure on runoff and dewatering from
high water table soils for their streamflows. Throughout the remainder of
the southeast, where large rivers flow through the Atlantic Coastal Flat-
woods subphysiographic province, the headwaters originate in the Pied-
mont or in the Southern Appalachian Highlands.
Third, Florida has a deep aquifer (a complex of several limestone
formations-the Ocala, the Avon park, the Lake City, the Suwannee,
and the Tampa-that make up the Florida aquifer) that is separated from
the surface waters in many parts of the state by an aquiclude, the Haw-
thorne Formation (Stewart, 1980). However, other areas of recharge and
some locations, particularly in the northern and western parts of the
peninsula, have numerous springs that flow out of the limerock. The
Floridan aquifer also extends into a large area of South Georgia.
Fourth, in the populous southeastern part of the state, the shallow
Biscayne aquifer serves as the only source of storage for fresh water. This
aquifer is a local, rainfed aquifer, and water conservation areas and the
Everglades are supplied both by rainfall and by water from the drainage
basin of Lake Okeechobee (primarily the Kissimmee River, Fisheating
Creek, Harney Pond Canal, Indian Prairie Canal, and the Taylor Creek/
Nubbin Slough watershed).
Year-to-year variations in rainfall not only have serious effects on
agricultural production, but they cause wide fluctuations in the shallow
groundwater resources. These water supply and availability problems
present a challenge in meeting the expanding, competitive urban, indus-
trial, and agricultural needs while maintaining some of the wetlands
Water resource management within the state of Florida is becoming an
increasingly important function. In the early 1970s, water shortages
became acute in some areas. The State Legislature established a coordi-
nated system of water management districts covering the entire state
through the Water Resources Act of 1972. This act created five Water
Management Districts. The districts were required to develop a Water
Use and Supply Development Plan and to immediately begin accom-
plishing the ongoing functions of planning, operations, and regulation of
the water resources in their districts. Thus, districts faced the responsibil-
ity of allocating a declining water supply over an ever growing and
conflicting water demand within the urban, industrial and agricultural
Withdrawal of fresh water for irrigation represents the largest of the
state's water pumping demands (Leach, 1978). However, agricultural
areas contribute to the water harvest of surface and groundwater. Most
other withdrawal users do not generate contributions to water resources,
although they may return water to the system. In the past, the agricultural
water pumping demand has not created any serious problem; however,
with the rapid population increases and the new industrial expansions,
the potential user conflict has increased, especially during periods of low
rainfall. Thus, detailed information on measurement or estimation of
evapotranspiration (ET) is required for water resource managers and
growers to make realistic decisions concerning their water management
The main objectives of this bulletin are: (1) to present principles of the
climate dependence of potential evapotranspiration, (2) to discuss the
similarities of potential ET for various climatic regions in Florida, (3) to
evaluate several methods of estimating actual ET, (4) to discuss the
modifying influences of plant canopy and physiological stage of growth on
ET of agricultural crops, (5) to illustrate the value of irrigation to prevent
crop yield reduction, and (6) to discuss the interrelationships of ET, soil
water storage, effective rainfall, and irrigation requirements within the
framework of irrigation management systems.
2.0 EVAPOTRANSPIRATION CONCEPTS
2.1 CLIMATE DEPENDENCE
2.1.1 Evapotranspiration and the Energy Budget
Evapotranspiration is the combination of two processes: evaporation
and transpiration. Evaporation is the direct vaporization of water from a
free water surface, such as a lake or any wet or moist surface. Transpira-
tion is the flow of water vapor from the interior of the plant to the
Transpiration involves transport of water in the liquid state from the
soil, through the plant, to the leaves. It is a specialized case of evapora-
tion, because liquid water is evaporated inside the leaves, and the result-
ing water vapor diffuses mostly through small pores, called stomata, into
the atmosphere. When fully open, these stomata result in a leaf surface
that is about 1% pore space. Stomata of most higher green terrestrial
plants open during the day and close at night. If the soil is too dry, plants
will be stressed and stomata will partially close during the day to keep the
plant from losing water faster than it can be taken up by the root system.
Evaporation includes the change in state of water from liquid to vapor
and a flow of water vapor from the evaporating surface to the bulk
atmosphere. Early attempts to calculate evaporation rates were based on
vapor pressure deficits of air (the difference between saturation vapor
pressure and actual vapor pressure, or dryness of the air) and windspeed.
However, it is difficult to calculate evaporation or ET using only these
factors because of the importance of the energy source (solar radiation),
temperature, and surface conditions. One way to eliminate the complica-
tions has been to treat evaporation as one component of the energy
balance at the earth's surface.
Evapotranspiration requires energy-it is an energy flow process as
well as a water flow process. It takes about 2430 joules per gram (580
calories per gram) to convert liquid water to water vapor at a temperature
of 300C. For comparison, it takes 4.19 joules (1 calorie) to raise the
temperature of 1 gram of water by 1C. The energy balance is simply the
partitioning into energy inflows and energy outflows. These inflow or
outflow components at the surface include radiant energy flux, soil heat
flux, sensible heat flux convectivee heating or cooling of the air in contact
with the surface), latent heat flux (evaporation), and energy flux stored as
photochemical energy (photosynthesis).
The net radiant energy exchange at the earth's surface, called net
radiation, R,, can be expressed as
Rn = (1 a) R, + Rd R (1)
where R, = incoming solar radiation (direct and diffusive)
Rd = thermal radiation from the atmosphere (incoming)
R, = thermal radiation from the earth's surface (outgoing)
a = albedo (reflectively of solar radiation)
These values are usually expressed in energy flux density units such as
watts/m2 or cal/cm2 min.
At night the radiation budget is made up of only thermal radiation.
Normally, the earth's surface is warmer than the adjacent air, and it rad-
iates energy upward at a greater rate than the atmosphere radiates energy
downward. As a result, radiant energy is lost to space, and the earth's
surface cools as it loses heat which was stored in the soil during the day.
The net radiation that is absorbed by the vegetation and soil at the
earth's surface is the net energy inflow. For a vegetated surface which is
well-watered, the largest part of the net radiation goes into evapotran-
spiration. This component is called "latent heat flux" because energy is
used to evaporate water without raising the temperature. Usually a
smaller part of the net radiation goes into heating the air through convec-
tion from vegetation and soil. This movement is called "sensible heat
flux." During the daytime, a small fraction goes into heating the soil and
is called the "soil heat flux." Finally, a small fraction, usually less than
5%, goes into photosynthesis when the ground has an active vegetative
The energy outflows that balance the net radiation inflow can be
R, = ET+H+G+P (2)
where ET = evapotranspiration expressed as latent heat flux density
H = sensible heat flux density
G = soil heat flux density
P = flux density of solar radiation stored as photochemical en-
ergy in the process of photosynthesis.
Since P is usually small, it is ignored in many energy balance applications.
Figure 3 shows a typical energy budget at 2 p.m. EST on April 29, 1978,
for a short-grass pasture near Okeechobee, Florida. The energy inflow is
from net radiation (R,), which is the net amount of radiant energy
absorbed by the grassland. In this example, 62% of the net radiation (R,)
went into evapotranspiration or latent heat flux (ET), 22% into sensible
heat flux (H), and 16% into soil heat flux (G).
During the daytime, the earth's surface usually heats the air by convec-
tion, as shown by H in Figure 3. However, if a wet, freely evaporating
surface is surrounded by a large hot, dry area, the wet surface could
absorb sensible heat from the air as the wind blows across. This transfer-
ence of heat is called "sensible heat advection," and the additional
energy absorbed goes largely into evaporation of water. This effect is
called the "oasis effect." An example would be the case of evapotran-
spiration from an irrigated field in an arid climate. During the dry season
or periods of drought in Florida, an irrigated field could experience
sensible advection with higher evapotranspiration rates and negative
sensible heat flux, or at least the upward-directed sensible heat flux would
become much smaller.
2.1.2 Potential Evapotranspiration
Potential evapotranspiration is a concept popularized by the climatol-
ogist C. W. Thornthwaite (1944) as "the water loss from a moist soil tract
R, = 0.670 cal/cm2' min
G = 0.109 cal/cm2 *min
H = 0.149 cal/cm min
ET = 0.412 cal/cm2 *min
Fig. 3. Energy balance at 2:00 p.m. EST, April 29, 1978, for a short grass pasture
near Okeechobee, Florida (from Allen et al., 1978).
completely covered with vegetation (without specifying the type), and
large enough for oasis effects to be negligible. . since moisture is not
restricted, potential evapotranspiration is limited solely by available
energy." Penman (1956) defined potential transpiration as "the amount
of water transpired in unit time by a short green crop, completely shading
the ground, of uniform height and never short of water." These two
definitions imply that evapotranspiration from a well-watered, active
crop with full ground cover is determined primarily by meteorological
processes. (Stomatal closure and reduced transpiration usually are im-
portant only under limited water or other plant stress conditions). In
quantifying these processes, Penman (1956) derived an equation which
showed that potential ET could be expressed as a function of net radia-
tion, atmospheric vapor pressure deficit (dryness of air), temperature,
and windspeed. This equation expresses fully the climate dependence of
potential ET and can be adapted for estimating actual ET. The Penman
equation will be described in Section 2.3.
2.2 SURFACE DEPENDENCE OF EVAPOTRANSPIRATION
There are several surface factors which modify the energy balances
within Equation 2 and hence influence evapotranspiration. Surface fac-
tors include: (1) ground cover (vegetation); (2) soil water availability;
and (3) stomatal behavior.
Ground cover affects ET in several ways. It affects the albedo, or
reflectivity, of the surface. Vegetative surface albedos range from about
0.2 to 0.25. Water surfaces, on the other hand, would have a typical
albedo of 0.05. Soil albedo is highly variable; it is usually low (less than
0.1) when organic matter content is high, and may be as large as 0.5 for
beach sand. Albedo also varies widely with soil color, which changes with
soil water content.
Ground cover changes the amount of radiant energy impinging directly
on the soil and thus the amount of energy absorbed by the soil (soil heat
flux density term, G). Soil properties, including water content, will also
affect the amount of energy flowing into the soil compared with the
amount of direct evaporation from the soil. The height and density of
ground cover influence the efficiency of turbulent exchange of both heat
and water vapor from canopies.
Changes in soil water will cause differences in direct evaporation from
the soil, and differences in water availability to plants. As plants become
water stressed, their stomata close, which results in a reduction of plant
water loss as well as plant CO2 uptake. This is one factor that the potential
ET equation does not take into account. Stomatal resistance under
non-stressed conditions varies among plant species. However, ETdiffer-
ences are usually small, and the potential ET concept works well across
most types of vegetation with full canopies.
2.2.1 Influences of Plant Canopy and Physiological Stage of
Growth on Actual Evapotranspiration
Actual ET may be less than potential ET much of the time during the
production of an agricultural crop. Since ET, or evapotranspiration, is the
combination of two processes, soil water evaporation (E) and plant
transpiration (PT), or ET = E + PT, these two processes respond to
variations in the physical environment in different ways. Soil water
evaporation occurs on or near the soil surface, and the water vapor is
transported to the bulk atmosphere. When a bare soil surface is satu-
rated, soil water evaporation occurs at the potential evaporation rate
(Ep), which is similar to open water surface evaporation and is deter-
mined by the environment only. As the soil surface begins to dry, water is
conducted from subsurface soil regions toward the surface to replace that
water lost by evaporation. As water continues to evaporate from the soil
surface, soil below the surface begins to dry. As the soil dries, its ability to
conduct water to the surface decreases to the point that evaporation of
soil water becomes limited. Thus, soil water evaporation can be described
in two stages: (1) the time during which soil E is at the potential rate, and
(2) the time during which soil E is limited by soil water transport charac-
teristics and E is below the potential E rate. Eventually, a state of
essentially no soil E would be reached from the soil surface.
For soils exposed to wetting and drying cycles, soil water evaporation
goes through cycles of being in Stages 1 and 2, depending on the fre-
quency and amount of rainfall. Frequent rainfall tends to maintain soil
evaporation at the potential rate (Stage 1), whereas low frequency rain-
fall patterns result in evaporation being limited by soil properties much of
the time. Over a period of a week or a month, the actual E represents the
integrated effects of wetting and drying cycles on soil water evaporation.
Under normal crop growing conditions, usually the actual E over a period
of a week or a month will be less than potential E for the same period.
Management practices or rainfall distributions that tend to maintain a wet
soil surface may considerably modify the actual E to the point that it
approaches Ep during a week or month time period. Stewart and Mills
(1967) and Stewart et al. (1969) present monthly and annual E and ET
data for turfgrass cover, ranging from full cover to bare soil under
different rainfall and water table depth conditions.
Plant transpiration represents water loss from plant leaves to the
atmosphere. As water transpires from the leaves, the plant absorbs water
from the bulk soil through its root system and transports it to the leaves to
replace water transpired. Under well-watered conditions, the plants
usually absorb enough water through their root systems to maintain
transpiration rates at the potential rate, PTp, determined by the environ-
ment. However, as the soil around the root system dries, the ability of the
soil to conduct water to the roots decreases and plants can no longer
supply water fast enough to maintain the PTp rate. In order to prevent
leaf dessication, the plant has a feedback control system that causes
stomatal closure, thereby decreasing actual transpiration below the
potential rate. Thus, the transpiration process can also be discussed in
two stages: (1) potential transpiration stage, and (2) soil water limiting
The reduction in transpiration during Stage 2 depends on various soil
characteristics, the extent and density of plant roots, and on potential
transpiration. For soil-plant systems exposed to natural rainfall-drought
cycles, plant transpiration goes through cycles of being in Stage 1 or 2,
with the overall result that PT is less than PTp. However, the plant
potential T (Stage 1) is generally much longer than the soil potential E
(Stage 1), because of the bulk soil water holding capacity and the availa-
bility of that water for transpiration through the root system. Again,
these factors depend on the crop and its root system, the soil, manage-
ment practices, and the potential rate of evaporation or transpiration.
2.2.2 Seasonal Evapotranspiration of Crops
During the time course of a seasonal crop, the crop system changes
from one in which ET is entirely soil evaporation to one in which ET is
mostly plant transpiration, and finally to one in which both plant trans-
piration and soil evaporation are affected by crop senescence. Just after
planting a crop, the soil surface is bare and soil water evaporation may go
through cycles of Stages 1 and 2. The net result over time is that actual ET
would be less than potential ET when soil cover is minimal.
As plants emerge, their leaves begin to shade the soil surface. During
the time when both plants and soil are exposed to direct radiation, both
soil evaporation and plant transpiration contribute to ET. As plants grow
and shade a larger fraction of the soil surface, the contribution of PT to
ET increases. Since actual PT is affected less than actual E by rainfall
frequencies, actual ET approaches ETp as the season progresses and as
the soil becomes completely covered by the crop.
When ground cover is complete, ETis mostly PT. During this period, if
ET drops below potential ET, it would likely be due to stomatal closure
and decreased PTin reaction to reduced soil water availability in the root
zone. This drop would represent a stress on the crop and cause growth to
be reduced because of less carbon dioxide uptake and reduced leaf and
As the crop matures, leaves (especially lower leaves) begin to die and
become nonfunctional. In some crops such as corn, the dead leaves might
remain on the crop, but contribute little to transpiration. However, the
soil is still shaded and the soil water evaporation rate would be lower than
that of a bare soil. The net result is that actual ET decreases during this
maturing stage of crop growth and becomes lower than ETp.
2.3 TECHNIQUES USED TO CALCULATE
There are numerous approaches used to estimate ET and potential
evapotranspiration (ETp). The following methods are frequently used:
mass (water vapor) transfer, energy budget, watershed water budget, soil
water budget, groundwater fluctuations, and empirical formulae. The
different techniques have been developed partly in response to the availa-
bility of data for estimating ET. Each method has certain advantages and
limitations. The availability of data is often the limiting factor in the
choice of calculation technique for practical applications.
The choice of calculation technique also depends on the intended use
(Burman et al., 1981; Doorenbos and Pruitt, 1977; Jensen, 1974; Linsley
et al., 1975; Saxton, 1982) and on the time scale required by the problem.
For example, irrigation management requires daily estimates of ET to
allow producers to make rational decisions concerning the timing and
amount of irrigation. In contrast, basin level planning may require
monthly estimates of ETto project changes in water supplies and require-
ments during the year.
2.3.1 Penman Method
The Penman (1948) equation was derived by rearranging the energy
balance equation (Equation 2) without the small photosynthetic compo-
nent. When applying the Penman formula over a 24-hour period, the net
energy component going into heating the soil is small, because a large
part of the energy absorbed by the soil during the day is lost at night.
Therefore, the soil heat flux density term, G, can be dropped for 24-hour
calculations. The sensible heat flux density term, H, is replaced in the
Penman formulation by mathematical substitutions, using the saturation
vapor pressure vs. temperature relationships. Finally, these procedures
yield the Penman formula for potential evapotranspiration based on four
major climatic factors: net radiation, air temperature, wind speed, and
vapor pressure deficit. The reader is referred to Penman (1948, 1956,
1963), Tanner and Pelton (1960), Monteith (1964), Tanner and Fuchs
(1968), Jensen (1974), Doorenbos and Pruitt (1977) and Burman et al.
(1981) for more thorough discussions of the derivation and its applica-
tion. The potential ET for each day can be expressed as
ET, AR,,/ + E,(3)
EAT = -Y(3)
where ETp = daily potential evapotranspiration, mm/day
A = slope of saturated vapor pressure curve of air, mb/C
R, = net radiation, cal/cm2 day
X = latent heat of vaporization of water, 59.59-0.055 Tavg
cal/cm2 mm or about 58 cal/cm2 mm at 29C
Ea = 0.263 (ea ed)(0.5 + 0.0062u2)
ea = vapor pressure of air = (ex + emin)/2, mb
ed = vapor pressure at dewpoint temperature
Td (for practical purposes Td = Tmin), mb
u2 = wind speed at a height of 2 meters, km/day
y = psychrometric constant = 0.66 mb/C
Tavg = (T,max + Tmi,)/2, C
em, = maximum vapor pressure of air during a day, mb
emin = minimum vapor pressure of air during a day, mb
Tma = maximum daily temperature, C
T,in = minimum daily temperature, C.
From Bosen (1960), saturated air vapor pressure as a function of air
temperature, e*(T), and the slope of the saturated vapor pressure-
temperature function, A, can be computed as follows:
e*(T) = 33.8639 [(0.00738T+ 0.8072)8 0.000019
(1.8T+ 48) + 0.001316] (4)
A = 33.8639 [0.05904 (0.00738T + 0.8072)7
Values of saturated e vs. T and A/(A + y) vs. T are presented by Jensen
(1974). This equation requires daily values of maximum and minimum
temperatures, net radiation, and wind speed. Daily temperature data are
available for numerous locations in Florida and are relatively easy to
obtain. Net radiation values are not available and must be estimated from
total incoming solar radiation, R,, and the outgoing thermal or long wave
Penman (1948) proposed a relationship of the form
R, = (1 a) R, Rb (6)
where R, = net radiation in cal/cm2 day
R, = total incoming solar radiation, cal/cm2 day
Rb = net outgoing thermal or long wave radiation
ax = albedo or reflectivity of surface for R,.
Generally accepted values for a are:
a = 0.05 for water surfaces
a = 0.15 for bare soil surfaces
a = 0.23 for green vegetated surfaces.
Penman (1948) proposed a relationship for Rb of the form
Rb = T4 (0.56 0.08Ved)(1.42 R,/R,, 0.42) (7)
where a = Stefan-Boltzmann constant (11.71 x 10-8 cal/cm2 day/K)
T = average air temperature in "K (C + 273)
Ro = total daily cloudless sky radiation (values of Ro for Florida
latitudes are given in Table 1).
Table 1. Cloudless sky daily solar radiation received at the earth's surface, (Rso, cal/cm2 day) averaged for each month for three latitudes
(from Jensen, 1974, and Doorenbos and Pruitt, 1977).
Latitude' Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
25N 455 595 629 720 742 780 745 703 660 561 486 423
30*N 403 549 600 713 742 793 755 703 637 519 437 371
35N 345 496 568 700 742 800 761 697 603 474 380 313
'Florida latitudes range from 24* 33' (Key West) to 310 00' (Alabama state line).
I V I*
At some locations, values of percent sunshine hours (S) are available that
can be used to estimate R,. Fritz and MacDonald (1949) proposed a
relationship of the form
R, = (0.35 + 0.61S) Ro, (8)
Values of R, calculated from the above equation are subject to more error
than measured values of Rs, but when averaged over several days to a
month, should not be in error by more than 5% to 10%.
Where local orographic features do not strongly influence cloud cover,
daily measurements of R, at a single station can be used over large areas
for ET estimates over 5 and 10 day periods. In Florida, areas within 5 to 9
km of the coast may have significantly different cloud cover than the
central portion of the state. In these coastal areas, Rs values for stations in
the center of the state should be used with caution.
Total incoming radiation, R,, can also be estimated from cloud cover
data presented by NOAA for several locations in Florida (or obtained
from other methods). Readers are referred to a publication by Dooren-
bos and Pruitt (1977) for details on the procedure for estimating R, from
Since wind speed is measured at many different heights above the
ground surface, and since the Penman equation requires wind speed at a
height of 2 m, wind speed will normally need to be adjusted to a height of
u2 = uz (9)
where u2 = wind speed at height of 2 meters in km/day
uz = wind speed at height z in km/day
z = height of wind measurement in m.
The working Penman equation for potential ET in mm/day becomes
ET= A [(1 a) R, t T4(0.56 0.08 Va)
+ ^ [0.263 (ea ed)(0.5 + 0.0062u2)] (10)
The above equation, along with the discussed procedures for estimat-
ing R, and adjusting wind speed, is considered the most accurate method
available for estimating potential ETfrom a vegetated surface. The other
methods which are discussed briefly in the following paragraphs are less
accurate and should be used only when data are not available for the
2.3.2 Pan Evaporation Method
The open pan is the most widely used evaporation instrument today,
and its application in hydrologic design and operation has a long history.
The relationship between potential evapotranspiration (ETa) and pan
evaporation (PE) can be expressed as
ETp = k2 PE (11)
where k2 = pan coefficient (usually taken as 0.7 for Florida conditions,
but variable throughout the year) and
PE = evaporation from a National Weather Service standard Class
The pan integrates the climatic factors and provides a good estimate of
ET, if the pan is serviced as required and the area around it is properly
maintained. This situation is seldom the case; therefore, pan data should
be used with caution. Doorenbos and Pruitt (1977) point out that pan
coefficients are functions of relative humidity (vapor pressure deficit),
wind speed, and upwind fetch of the surface surrounding the pan. Class A
pan coefficients reported by Doorenbos and Pruitt (1977) for different
surrounding ground cover conditions, relative humidities, and daily wind
movements are listed in Table 2. They pointed out that for a well-watered
grass turf, 8 to 51 cm tall, the pan coefficient could vary from 0.85 for light
wind and high relative humidities to 0.40 for very strong winds and low
relative humidities. The pan coefficient was also highly dependent on
length of upwind fetch and type of ground cover (green crop or fallow).
The diurnal distribution of pan evaporation is also very different from
crop ET. Pan evaporation rates usually lag the daily cycle of solar radia-
tion by about 6 hours, whereas crop ETis very nearly in phase with solar
radiation. Leaves have very little heat capacity, and their stomata open in
response to light and close in the dark. Therefore, crop ET is very
responsive to the energy and climate factors during the day, and is
essentially nil at night. Evaporation pans continue to lose large amounts
of water well into the night (Campbell and Phene, 1976). These factors
lead to variability in the exact day-to-day correlations of pan evaporation
with ET or ETp.
2.2.3 Thornthwaite Method
Thornthwaite (1944) presented a method of estimating monthly poten-
tial ET from mean monthly temperature and day length.
PET= 16 Ld [10 T (12)
where PET= 30-day estimate of ET in mm
Ld = daytime hours divided by 12
I= (i) where i=
T, = mean monthly temperature, C, and
a = (6.75 x 10-7 1) (7.71 x 105 2)
+ 0.01792 I+ 0.49239. (13)
The Thornthwaite method greatly overpredicts ETp during the summer
months in Florida because it does not consider the increased cloud cover
(compared to winter months) in summer. In order to compare this
method with other methods, an empirical coefficient, k3, is used to
estimate ETp from PET, i.e.,
ETp = k3 PET. (14)
2.3.4 Blaney-Criddle Method
Blaney and Criddle (1950) developed a formula for arid climates for
predicting monthly ETp from percent of daylight hours in the month and
the average monthly temperature as
ETp = k4f
where ETp = monthly potential ET in mm (15)
k4 = coefficient for the Blaney-Criddle method
f= monthly ET factor = 25.4 PD (1.8 T,,, + 32)/100
T, = mean monthly temperature in C, and
PD = percent of annual daylight hours in the month.
The Blaney-Criddle equation has been widely used and modified, but
when used in Florida, it greatly overpredicts the ETfor summer months.
The equation does not account for the high percent of cloud cover in
Florida during the summer.
2.3.5 Modified Blaney-Criddle Method using Solar Radiation
The Blaney-Criddle (1950) method was modified by Shih et al. (1977).
They substituted the measured solar radiation instead of the percent of
daylight hours used in Equation 15. The equation used to estimate the
potential ET was expressed as:
ETp = k5 25.4 MR, (1.8 T, + 32)/TMR, (16)
where ETp = monthly potential ET in mm
k5 = coefficient for the Radiation-Modified Blaney-Criddle
MR, = monthly incoming solar radiation in cal/cm2
T,, = mean monthly temperature in *C, and
TMR, = annual sum of mean monthly solar radiation in cal/cm2.
Table2. Pan coefficients for Class A pan for different groundcover, levels of mean relative humidity, and 24-hour wind (from Doorenbos
and Pruitt, 1977, and Jensen, 1974).
Class A Pan, Case A: Class A Pan, Case B':
Pan surrounded by short green crop Pan surrounded by dry-fallow land
Upwind RH (mean %) Upwind RH (mean %)
distance of distance of
green crop low medium high dry fallow low medium high.
Wind (km/day) (m) (<40) (40-70) (>70) (m) (<40) (40-70) (>70)
Light 0 .55 .65 .75 0 .7 .8 .85
(<175) 10 .65 .75 .85 10 .6 .7 .8
100 .7 .8 .85 100 .55 .65 .75
1,000 .75 .85 .85 1,000 .5 .6 .7
S Moderate 0 .5 .6 .65 0 .65 .75 .8
(175-425) 10 .6 .7 .75 10 .55 .65 .7
100 .65 .75 .8 100 .5 .6 .65
1,000 .7 .8 .8 1,000 .45 .55 .6
Strong 0 .45 .5 .6 0 .6 .65 .7
(425-700) 10 .55 .6 .65 10 .5 .55 .65
100 .6 .65 .7 100 .45 .5 .6
1,000 .65 .7 .75 1,000 .4 .45 .55
Very strong 0 .4 .45 .5 0 .5 .6 .65
(>700) 10 .45 .55 .6 10 .45 .5 .55
100 .5 .6 .65 100 .4 .45 .5
1,000 .55 .6 .65 1,000 .35 .4 .45
'For extensive areas of bare-fallow soils and no agricultural development, reduce k2 values by 20% under hot windy conditions, and by 5-10% for moderate
wind, temperature, and humidity conditions.
2.3.6 Stephens-Stewart Method
Stephens and Stewart (1963) introduced a radiation method adjusted
for mean monthly temperature and calibrated it against 30 months of ET
data from St. Augustinegrass growing in non-weighing lysimeters at Fort
ETp = 0.01476 (Tm + 4.905) MR,IX (17)
where ETp = monthly potential ET in mm
T, = monthly mean temperature in C
MR, = monthly solar radiation in cal/cm2, and
X = latent heat of vaporization of water, 59.59 0.055 T,,
This method was called the "Fractional Evaporation-Equivalent of Solar
Energy" method by Stephens and Stewart, but it is essentially the same
form as the original Jensen and Haise (1963) method that has been used
frequently under western conditions. Stephens (1965) determined coef-
ficients for data from Waynesboro, North Carolina, and Davis, Califor-
nia, and for. western and midwestern crops reported by Jensen and Haise
(1963). The slope of the line increased with increasing aridity of the
location of the ET data sets, which implies greater ET per unit of solar
radiation energy as aridity (or saturation vapor pressure deficit) of the
2.4 POTENTIAL EVAPOTRANSPIRATION FOR
VARIOUS REGIONS IN FLORIDA
Seven climatological districts are defined for Florida by the National
Climatic Center for reporting climatological data: northwest, north,
north central, south central, Everglades and southwest coast, lower east
coast, and Keys. These seven districts are located within the range from
24030' to 310 north latitude. Three weather stations located within this
range of latitudes were selected as examples to demonstrate the tech-
niques of the Penman method used to compute the potential ET for
various regions of Florida. These three stations selected were Hialeah
(25050' N. lat., 8017' W. long.), Lakeland (28001' N. lat., 81055' W.
long.), and Milton (30047' N. lat., 87008' W. long.). (See Figure 2.)
Estimates of surface albedo are required for applying the Penman
equation (Equation 10) to calculate ETp. As discussed earlier, surface
albedo is not constant but varies with ground cover, soil type, and crop
stage of growth. Because of the difficulty in estimating seasonal changes
in surface albedo, one approach for applying the Penman method is to
use an albedo for a free water surface (a = 0.05) to estimate potential
evaporation rate for a free water surface (Eo), and then to multiply that
Table 3. Application of Penman method to calculate free water evaporation (Eo) and monthly potential evapotranspiration (ET,) for
three locations in Florida. ETp was calculated using Equation 10 and a fixed surface albedo of a = 0.23 and by multiplying Eo by
kl = 0.7 for comparison.
Hialeah, 25050' N. Lat. Lakeland, 28001' N. Lat. Milton, 30047' N. Lat.
a=0.05 a= 0.05 a =0.05
a=0.23 a=0.05 kl =0.7 a= 0.23 a=0.05 kt=0.7 a=0.23 a=0.05 k1=0.7
Month (ET,) (E.) (ETp) (ET,) (Eo) (ETp) (ETp) (Eo) (ET,)
Jan 68 90 63 56 75 52 46 60 42
- Feb 89 116 81 80 104 73 66 85 60
Mar 113 146 102 105 135 95 88 114 80
Apr 142 180 126 137 175 123 127 162 113
May 151 191 134 156 198 138 150 191 134
Jun 145 183 128 161 202 141 166 209 147
Jul 156 197 138 157 197 138 159 200 140
Aug 149 189 131 146 184 129 149 189 132
Sep 126 158 111 125 158 110 119 152 106
Oct 102 130 91 96 123 86 85 113 79
Nov 74 98 68 67 89 62 51 70 49
Dec 59 78 54 48 66 46 39 53 37
Annual 1374 1756 1227 1334 1706 1193 1245 1598 1119
L r -
estimate of Eo by an empirical constant to estimate ETp for a crop, i.e.,
ETp = k, Eo. (18)
Van Bavel and Verlinden (1956) used a = 0.05 and found that k1 = 0.7 in
Equation 10 provided a good estimate of ET, for a well-watered vege-
tated surface. In applying the Penman method to estimate ETp for gen-
eral applications, the use of a = 0.05 and Equation 18 is recommended
for reference ETp values. Actual ETcan then be estimated by multiplying
the reference ETp estimate by a crop coefficient which will be discussed in
later sections. For comparison purposes, we used a = 0.05 in Equation
10 to estimate Eo and k, = 0.7 in Equation 18 to estimate ETp. The results
of potential ET estimated by the Penman method for the stations of
Hialeah, Lakeland, and Milton are listed in Table 3. In Table 3, ETp was
also calculated by using a = 0.23 in Equation 10.
Assuming the basin-wide vegetated crop reached full canopy, the
annual total ETp varied from 1119 mm in Milton to 1227 mm in Hialeah.
The difference is only 108 mm or 9%. This means that the annual total
ETp in Florida increases slightly from north to south.
However, if ETp was computed on a November-to-April and on a
May-to-October basis as shown in Table 4, the difference between the
southern and northern parts of the state was shown to occur during the
November-to-April period. This difference amounts to a 30% deviation.
The May-to-October period did not show a significant difference in ETp
over the entire state. The climate is usually similar throughout the state
during the May-to-October period, i.e., hot and humid. However, during
the November-to-April period, the climate is cooler and usually wetter at
Milton than at Hialeah.
Table 4. Calculated seasonal potential evapotranspiration by Penman method
for three locations in Florida.
Hialeah Lakeland Milton
Time of (25050' (28001' (3047'
Parameter' Year N. Lat.) N. Lat.) N. Lat.)
Nov.-April 545 494 416
a=0.23 May-Oct. 829 841 828
Nov.-April 709 634 544
a=0.05 May-Oct. 1046 1061 1055
a = 0.05 Nov.-April 496 451 381
k= 0.7 May-Oct. 732 743 738
'The Penman Method (Equation 10) was applied in three ways: (1) with a = 0.23 to
represent a full crop canopy, (2) with a = 0.05 to represent a freely evaporating water
surface, and (3) with a = 0.05 and k, = 0.7 to estimate a reference ETp value using
Equations 10 and 18.
In conclusion, Tables 3 and 4 show that potential ET over the state is
essentially the same for the May-to-November period because of small
variations in climate. The differences in potential ETfrom south to north
occur mostly in the winter months but amount to only about 9% on an
annual basis. Net radiation is very similar over the 250 to 310 north
latitude band throughout Florida, especially for the six or eight warmer
months when ET rates are highest. Vapor pressures, temperatures, and
vapor pressure deficits are also very similar during these periods. Wind
movement is not extremely variable either. Therefore, potential ET, as a
climate-driven process, is very similar throughout the state, with most of
the differences occurring during the winter portion of the year.
3.0 EXPERIMENTAL VERIFICATION
The modified Penman combination method for computing ETp is based
on sound analysis of physical processes (rather than totally on correlation
techniques). We have further shown that ETp is similar throughout
Florida, except during the winter season. There are several cases where
climatic data and experimental data are available simultaneously, so that
calculations of potential ET can be verified using the water budget
method. The water budget method equates water input to an area during
a specified time period: rainfall (RF) plus irrigation (IR) is equated to the
change in soil water storage (AS) plus water output from the area during
the same time period in the form of surface runoff (RO), ET, and
percolation from the root zone (PN). A measure of ET is obtained by
measuring all other variables in the following equation:
ET = RF + IR RO PN + AS, (19)
where all variables are expressed in depth per unit time.
Experimentally, the measurements of RO, PN, and AS are difficult,
and data that can be used for water budget measurements are scarce.
Here, we present data from several watersheds in Florida with varying
types of vegetation and surface types to demonstrate their effects on ET.
Natural surfaces in Florida watersheds (based on wetness) could be
classified as: (1) open bodies of water (lakes and streams), (2) wetlands
(swamps, marshes), (3) upland soils (pine/palmetto, hardwood ham-
mocks, range, pastures, croplands), and (4) muck soil croplands. The
wetter surfaces would be expected to have higher ET than the drier
surfaces. Often these surfaces grade from one to another.
Mineral soils of Florida vary widely in the amount of water they hold
(30-fold, derived from Stewart et al., 1963, and Carlisle et al., 1978).
During drought conditions, soils with low water holding capacities will
limit ET sooner than soils with high water holding capacities.
The water budget method is also applied to estimating ET for specific
agricultural crops grown in experimental fields and in lysimeters with
adequate fetch to prevent an "oasis" effect. The effects of physiological
stage of growth on seasonal ET for the crops will be demonstrated.
3.1 BASIN-WIDE WATER BUDGET
Two types of water budget data were used to measure actual ET. The
first was an annual basis which included upper Kissimmee Basin, Jane
Green Creek, Wolf Creek, Taylor Creek (Okeechobee County), Mon-
reve Ranch (Martin County), the Green Swamp Area, and the Ever-
glades Agricultural Area. The second was a monthly basis for the Ever-
glades Agricultural Area. All of these watersheds except the Green
Swamp Area are considered to have negligible deep aquifer recharge or
leakage (Stewart, 1980).
3.1.1 Annual Water Budget
The long-term average annual water balances for several basins in
southern Florida, ranging from the Green Swamp Area to the Everglades
Agricultural Area, are presented in Table 5.
The differences in ET among these basins are attributable mainly to
surface cover or land use differences. The ET of the Upper Kissimmee
Basin (S-65) is high because of the relatively large cover of lakes and
wetlands. (The USGS estimated a lake evaporation of 1300 mm per
year.) Topographical maps of the Jane Green Creek and Wolf Creek
areas show more gradient in the landscape. LANDSAT satellite images
show a mixture of wetlands and well-drained lands, with no lakes, in the
Jane Green Creek area. Wolf Creek has essentially no wetlands, and
most of the watershed is cleared (probably for pasture).
Taylor Creek watershed land use is predominantly pasture, with some
rangeland, woodland, and wetlands. Monreve Ranch includes a greater
proportion of ponded wetlands than Taylor Creek watershed, and has
more annual rainfall because it is closer to the southeast coast of Florida.
The Everglades Agricultural Area (EAA) showed a higher annual ET
than the pasture areas of the Taylor Creek Basin. This higher ET prob-
ably is caused by more winter crop production in the EAA, by the earlier
and longer vigorous growing season for sugarcane (the most prevalent
crop), and by high water tables (irrigation) within the EAA. During the
spring in the Taylor Creek area, most pastures of subtropical grasses do
not begin active growth until the rainy season starts. Water tables are
typically low from November to April. Vegetative cover density is low
because of continuous grazing throughout the winter and spring.
In general, the higher annual ET values are associated with more
available water (lakes and wetlands) or with longer periods of full canopy
active vegetation, and the lower ET values with drier surface conditions
Table 5. Comparison of several long-term basin-wide water budgets on an annual basis. Pan evaporation is included for comparison.
Osceola County Taylor Creek Monreve Green
Area1 Watershed2 Ranch3 EAA4 Swamp5
Item S656 Creek Creek W-29 W-310 W-511 W-412 W-413
Precipitation 1320 1320 1320 1240 1210 1320 1420 1550 1340 1320
Discharge 240 --=380 480 350 300 390 380 480 290 250
ET14 1080 940 840 890 910 920 1040 1070 1050 1020
Pan evaporation 1520 1520 1500 1490 1480 1550 -
'Source: USGS Open File Report 79-1289. Lake evaporation estimated to be 1300 mm.
NJ 2Source: Hydrology and Hydrogeology of the Upper Taylor Creek Watershed, Okeechobee County, Florida. USDA Technical Bulletin (In Press).
Knisel, Yates, Sheridan, Woody, Allen, and Asmussen. Monthly ET was also reported.
"Source: Allen, et al. (1982). Monthly ET was also reported.
'Source: South Florida Water Management District, R. Mierau, (personal communication). Area = 242,000 ha in this part of the Everglades Agricultural
Area 11-year record (1962-1972).
"Source: USGS Water-Resources Investigations 78-99 (Grubb and Rutledge, 1979). The water balance also includes about 50 mm of flow to the Floridan
Aquifer (Pride et al., 1966). Lake Helene evaporation in 1962 was 1350 mm (Pride, et al. 1966).
6Lake Kissimmee discharge structure. Area includes lakes and wetlands, as well as uplands.
7Eastern Osceola County flow to St. Johns River. Area includes wetlands but not lakes.
"Eastern Osceola County flow to St. Johns River. Area includes no wetlands or lakes.
'15-year record (1959-1973). Area of 27,066 ha.
"15-year record (1959-1973). Area of 4,947 ha.
"8-year record (1965-1972). Area of 9,169 ha.
"51/2-year record (Dec. 1967-May 1973). Area of 1,606 ha. Area includes ponds and wetlands.
"13-year record (1960-1972). Area of 1,606 ha. Precipitation includes an average of 76 mm per year of applied irrigation.
"Precipitation minus discharge. This assumes no net change in soil storage and no leakage to the aquifer.
(which may also tend to decrease vegetative canopy cover during the
typically dry spring months).
The annual basinwide ET in these watersheds ranged from 840 mm to
1080 mm. They are all less than the ETp value of 1200 mm given in Table
3. The most likely reason is that the watersheds do not meet the Penman
criterion of "full canopy cover, never short of water" over all areas of the
watershed throughout the entire year.
3.1.2 Monthly Water Budget
The EAA consists of about 190,000 hectares, for which the water
budget method was applied (Mierau, 1974). In 1980 about 75% of the
area was used for sugarcane production and the remaining 25% of the
area was under truck crop and pasture production. Several unique fea-
tures exist in the EAA. First, only organic soil is found in this area.
Second, it is entirely devoted to intensive agricultural production. Third,
the crops are produced year around. Fourth, it is surrounded by levees
and the inflow and outflow are well regulated by gates and pump systems.
Fifth, the canal system within the basin is well maintained for seepage
irrigation and drainage. Sixth, historical water budget components have
been monitored extensively. Seventh, it is an area of negligible surface
outflow and Floridan aquifer leakage (Stewart, 1980). Because of these
unique features, the EAA is considered to be ideal for demonstration of
the application of the techniques introduced in the previous section
(Shih, et al., 1983).
Climatological data were obtained from the weather station located at
the University of Florida's Agricultural Research and Education Center
(AREC) in Belle Glade. Averages based on 52 years' (1924-75) data of
pan evaporation, rainfall, and mean are temperature; 19 years' (1958-76)
relative humidity data, and 7 years' (1934-39, 1978-79) wind velocity
data at 2 m height are given in Table 6. Twelve years' (1967-78) data of
the percentage of possible sunshine obtained from the Tampa Interna-
tional Airport Weather Station are also included in Table 6.
The potential ET was computed on a monthly basis using the Penman
method as defined in Equation 10 (a = 0.05) and 18 (k, = 0.7) and using
weather data from Table 6. The resulting ETp values are listed in Table 7.
The annual Eo of 1479 mm compares favorably with the annual pan
evaporation of 1569 mm (a 6% difference). Applying the 0.7 kl value in
equation (18), ETp was estimated to be 1035 mm. The 10 years' (1962-71)
water budget ETdata for the EAA as reported by Mierau (1974) are also
shown in Table 7. The average annual ET as measured by water budget
was only 17 mm more than ETp. A rounded value of 1020 mm was then
used to establish the coefficients k2, k3, k4, and ks used in Equations 11,
14, 15, and 16, respectively. All coefficients except for the Penman
method (kl) are "calibrated" against this watershed. The results are
Table 6. Average monthly climatological data for the Belle Glade, Florida, weather station.
Solar Possible air Relative at 2 m Pan
radiation sunshine1 temp. humidity 1934-39 evap. Rainfall
Month 1971-79 1967-78 1924-75 1958-76 1978-79 1924-75 1924-75
cal/cm2 day % 0C % km/day ---------------mm-----------
Jan 282 62 17.8 76 121 84 51
Feb 340 64 17.8 74 131 99 50
Mar 436 73 19.4 73 125 148 82
S Apr 510 79 21.1 72 119 166 75
May 507 71 23.9 75 100 180 120
June 472 63 25.6 78 87 158 231
Jul 452 62 26.7 78 83 159 218
Aug 447 59 27.2 78 75 156 209
Sep 405 60 25.6 79 85 130 224
Oct 370 61 22.8 77 100 118 144
Nov 302 62 19.4 76 106 93 44
Dec 281 60 17.8 75 106 78 46
Totals 1569 1494
'Data from Tampa International Airport Weather Station.
Table 7. Monthly evapotranspiration estimated by different methods for the Everglades Agricultural Area for 10 years (1962-1971).
Penman Pan evap. Thornthwaite Stephens- Water2
method method method Original Modified1 Stewart budget
Month kl =0.70 k2=0.65 k3=0.85 k4=0.55 k5=0.55 method method
---------------------------------------- n------------mm--------------------------------------- ----
Jan 48 55 37 65 52 50 48
Feb 62 65 37 67 64 54 36
Mar 86 96 57 78 86 83 71
Apr 107 108 77 88 106 101 109
May 117 117 110 99 112 115 152
Jun 113 103 131 108 108 110 112
Jul 113 103 146 106 106 112 107
Aug 110 101 143 102 105 113 119
Sep 96 85 119 94 94 94 114
Oct 81 77 87 82 80 80 46
Nov 56 60 51 71 60 56 46
Dec 46 51 38 64 53 50 58
Total 1035 1021 1033 1024 1026 1018 1018
'This modification of the Blaney-Criddle method was developed by Shih et al. (1977) and uses solar radiation instead of percent daylight hours.
2Data reported by Mierau (1974) and Mierau (personal communication).
Table 8. Evapotranspiration deviation between water budget data and model predictions for the Everglades Agricultural Area.
Blaney-Criddle Method Stephens-
Penman Pan evap. Thornthwaite Stewart
Month method method method Original Modified method
Jan 0 + 7 -11 +17 + 4 + 2
Feb +26 +29 + 1 +31 +28 +18
Mar +15 +25 -14 + 7 +15 +12
Apr 2 1 -32 -21 3 8
May -35 -35 -42 -53 -40 -37
Jun +1 -9 +19 -4 -4 -2
Jul + 6 4 +39 1 1 + 5
Aug 9 -18 +24 -17 -14 6
Sep -18 -29 + 5 -20 -20 -20
Oct +35 +31 +41 +36 +34 +34
Nov +10 +14 + 5 +25 +14 +10
Dec -12 7 -20 + 6 5 8
deviations 169 209 253 238 182 162
deviation 14 17 21 20 15 14
listed in Table 7. The potential ET computed by those four methods
varied slightly from the 1020 mm value because of rounding the coef-
ficients to the nearest 0.05 value. The results of the Stephens-Stewart
method are also listed.
The deviations between the water budget and model predictions for
each month were computed. Also the totals of the absolute values of
these deviations were computed (Table 8). The total absolute deviation is
used as a criterion for judging the applicability of each method for
estimating ETfor the basin. The smaller absolute deviation means better
prediction. Several observations can be made from Table 8.
A positive value in Table 8 indicates that the method overpredicted
monthly ET. The pan evaporation, Blaney-Criddle, and Radiation-
Modified Blaney-Criddle methods over-predicted ET during winter
months October through March and under-predicted ET during summer
months. In contrast, the Thornthwaite method over-predicted ETin the
summer and fall (June through November) and predictions were gen-
erally low December through May.
The soil water storage in the EAA changes significantly at both the
beginning (May) and the ending (October) of the wet season. The water
table in the area is maintained about 200 mm higher in the wet season
than in the dry season. In muck soil, a change in the water table depth of
200 mm results in a change in soil water storage of about 33 mm. The
water budget method used to estimate ET for the EAA did not take this
change in soil water storage into account. The large deviations between
the water budget and the Penman and other methods in May and October
(Table 8) can be partially explained by increased soil water storage in
May and decreased soil water storage in October.
The total of the absolute values of the deviations varied from 162 for
the Stephens-Stewart method to 253 mm for the Thornthwaite method,
an average deviation of 14 to 21 mm per month. On the basis of seasonal
trends in deviations and on average deviations, the Penman and
Stephens-Stewart methods best predicted monthly ET and the Thorn-
thwaite method predicted poorest. In third place was the Shih et al.
(1977) Radiation-Modified Blaney-Criddle method. In fourth and fifth
places were 209 mm for the pan evaporation method, and 238 mm for the
original Blaney-Criddle method, respectively. Stephens and Stewart
(1963) also found that the ranking of accuracy of methods was Stephens-
Stewart = Penman > Blaney-Criddle > Thornthwaite in their compari-
son of nine methods of computing monthly ETp at Ft. Lauderdale.
3.2 EXPERIMENTAL PLOT WATER BUDGET VERIFICATION
3.2.1 Monthly Water Budget
Figure 4 shows 5-day average ET rates for turfgrass grown in lysimeters
at Fort Lauderdale in 1965 with water tables maintained 30 cm below the
1- I ,I
5-day average ET
0 1 1 1 1I
0 100 200 300
Fig. 4. Five-day averages of potential evapotranspiration for turfgrass and mea-
sured evapotranspiration from Tifway bermudagrass in 30-cm water table
lysimeters (data from Plantation Field Laboratory, Fort Lauderdale,
Florida, 1962, reported by Allen et al., 1978).
soil surface (Allen et al. 1978). Although water table depths of 60 and 90
cm were also maintained in the lysimeters, the 30 cm water table treat-
ment more closely fits the definition of a well watered vegetative surface.
Daily potential ET for 1962 was computed using the Penman method
(Equations 10 and 18, ua = 0.05, k, = 0.7) and using climatological data
at the turfgrass site. Daily ETp data were averaged over 5-day periods and
are plotted in Figure 4 for comparison with experimental ET based on
water budgets for the lysimeters. There is good agreement between ETp
and experimental ET; both data sets demonstrate periods of reduced ET
during the summer. Potential ET exceeded experimental ET during the
Table 9 shows monthly ET for various crops in Florida measured by
water budget method. Data for citrus were obtained from the SWAP
project in Ft. Pierce (Rogers et al., 1983) and from Lake Alfred (Koo and
Sites, 1955). The SWAP orchard had Bahiagrass cover under the trees,
which resulted in essentially complete vegetative cover over the soil.
Annual ET (1089 mm) was slightly lower than the ETp given by the
Penman Method (oa = 0.05, kl = 0.7) of about 1193 mm per year (Table
3, data from Lakeland). Monthly ET values were also slightly lower than
ETp values for January through September and slightly higher the rest of
the year. Annual ETin 1955 for Lake Alfred citrus (1001 mm) was lower
than citrus ET at Ft. Pierce and lower than the ETp for Lakeland. This
difference was due mostly to lower ET values at Lake Alfred during the
dry months of February through May in 1955. Lake Alfred citrus ETfor
one year was 16% lower than the Penman based ETp estimate of
In 1973, annual ET for peaches (Gainesville) was about 18% lower
than annual ETp based on Penman method (Table 3, 1119 mm for
Milton). Monthly ET data were also consistently lower than monthly
ETp. Management practices that reduce grass cover between trees in the
orchard would reduce ET, as explained earlier. The low ET of this peach
orchard was probably the result of incomplete vegetative cover.
Table 9. Water budget estimates of monthly evapotranspiration for citrus, pas-
ture, sugarcane, rice, and peaches for various locations in Florida.
Month Citrus1 Citrus2 Pasture3 Sugarcane4 Rice5 Peaches6
Jan 56 50 51 36 39
Feb 56 47 64 28 37
Mar 81 68 85 64 82 71
Apr 90 84 107 86 97 94
May 115 78 132 122 107 154
Jun 129 124 108 152 129 114
Jul 127 121 122 165 170 118
Aug 118 113 122 170 178 120
Sep 105 104 98 130 114 51
Oct 93 103 87 132 121 47
Nov 63 59 63 81 95 15
Dec 56 50 49 66 54
Annual 1089 1001 1088 1232 912
'Based on 8 years of water balance data from the SWAP project at Ft. Pierce ARC (Rogers,
et al., 1983).
2Koo and Sites (1955), Lake Alfred, Florida.
3Stewart and Mills (1967). Mean monthly values averaged over 5 years (3 years Tifway
Bermuda grass and 2 years St. Augustinegrass) and over water table depths of 30, 60, and 90
cm maintained in lysimeters at Fort Lauderdale, Florida. These turfgrass ET values are
assumed to be valid for pastures.
4From Shih and Gascho (1980), Belle Glade, Florida.
'Shih, et al. (1982), Belle Glade, Florida. Averages from 5 planting dates (March, April,
May, June, and July) and adjusted to a rough grain yield of 5500 kg/ha.
61973 water balance data from Gainesville, Florida. From T. Phung and J. F. Bartholic.
Water balance in a peach orchard. In: Disposition of Water from Fruit Crops and
Approaches to Increase Water Use Efficiency. Water Resources Research Center, Publica-
tion No. 33, OWRT Project Number B-014-FLA. Matching Grant Agreement Number
DI-14-31-0001-3868. February 1976.
The ET data presented for pasture turfgrass ET data were reported by
Stewart and Mills (1967) and Stewart et al. (1969). Annual ETwas 11%
lower than annual ETp for Hialeah (Table 3). Monthly ET values were
consistently lower than monthly ETp values.
Sugarcane ET data reported by Shih and Gascho (1980) were taken in
lysimeters in Belle Glade. On an annual basis, sugarcane ET was about
equal to ETp for Hialeah given by the Penman Method (Table 3). On a
monthly basis, these sugarcane ET data were higher than ETp values
during June through December and lower during the other months.
Monthly ET values for rice were similar to sugarcane ET. The data
shown in Table 9 were averaged from five different planting dates
(March, April, May, June, and July) and were also adjusted (normalized)
to a rough grain yield of 5500 kg/ha.
In summary, data presented in Figure 4 and in Table 9 verify the utility
of the Penman method for estimating monthly ETfor situations in which
conditions exist for potential ET to occur. The peach orchard also dem-
onstrated a case in which incomplete vegetative cover and management
practices could significantly influence ET.
3.2.2 Seasonal Crop Evapotranspiration
Many crops are produced from seed to harvest over a 3- to 5-month
time period. The ET for these seasonal crops changes with physiological
development and ground cover as plants grow. Examples of water budget
ET data are presented for several seasonal crops in Figures 5-9.
Doss et al. (1962) presented ET by irrigated corn from four years of
experiments at Thorsby, Alabama. Figure 5 shows the four-year average
ET by corn from those experiments. Since the summer climate in the
Thorsby area is similar to that in north Florida, the potential ET and free
water surface evaporation (Eo) calculated by the Penman method for
Milton are plotted in Figure 5 for comparison. The ETp calculations are
based on a = 0.05 and k, = 0.7, using Equations 10 and 18. These data
demonstrate the canopy development, full canopy, and senescing canopy
stages of seasonal crop ET. During the corn canopy development from
emergence to early June, ETwas below ETp. Between July 1 and August
1, ET was higher than the reference ETp, possibly because of advection
from surrounding areas that were not irrigated. The ET dropped below
ETp as senescence progressed between August 1 and August 15 (matur-
ity). Actual ET was lower than Eo throughout the season.
Figure 6 shows water budget ET data from Doss et al. (1965) for
bahiagrass experimental plots averaged over the 1957 and 1958 seasons.
These plots were irrigated when available soil water depletion reached 65
percent. This level of irrigation could have allowed some stress, since ET
was consistently lower than ETp. Also, these grass plots were periodically
clipped, causing reductions in ET immediately after clipping. Figure 7
0o I I_ LI I
15 1 15 1 15 1 15 1 15 1
May June July August
Fig. 5. Evapotranspiration (ET) for corn (reported by Doss et al. (1962) for
NOTE: Crop ET data for Figs. 5-9 are from water budgets of field or lysimeter plots.
Potential evapotranspiration (ET,) and free water surface evaporation (E,) cal-
culations are based on Milton, Florida, weather data.
shows ET data averaged over four years (1957, 1958, 1960, 1961) for
'Sart' sorghum from the same study of Doss et al. (1965) at Thorsby,
Alabama. They concluded that there was little difference in full canopy
ET between well-irrigated corn, sorghum, and cotton in their study. The
ET for a second sorghum crop reached potential rates, whereas the first
crop ET was slightly higher than ETp for 30 days (June 15 to July 15).
Water budget ET data were presented by Saxena et al. (1971) for
tomatoes grown in experimental field plots near Live Oak in 1969. We
averaged their data over irrigation and non-irrigation treatments and
smoothed them using 1:3:1 weighting values for previous, current, and
subsequent values, respectively. Data from May 15 to 23 and July 2 to 17
were omitted because heavy rainfall produced percolation that was not
measured. The ETof tomatoes (Figure 8) increased rapidly from about 2
mm/day on April 27 to peak values of near 5 mm/day on June 1. The
tomatoes were transplanted as seedlings, which allowed more rapid
7 E---000---00-OWN-------------- aft*-------
15 1 15 1 15 1 15 1 15 1
May June July August
Fig. 6. Evapotranspiration (ET) for bahiagrass (reported by Doss et al. (1965)
for Thorsby, Alabama).
NOTE: See Note, Fig. 5.
development of the canopy and vegetative ground cover than that of
seed-planted crops. Note that ET was nearly equal to ETp for 20 to 30
days. The decline after June 15 was apparently due to maturity and
Stansell et al. (1976) presented seasonal ETdata from three varieties of
peanuts grown under well-irrigated conditions in lysimeters in Tifton,
Georgia. Data from the three varieties were averaged and are presented
in Figure 9. The peak ET rate was about 5.5 mm/day, similar to the peak
rates shown previously for corn, sorghum, and tomatoes. This peak ET
rate is higher than the peak ETp rate predicted by the Penman method (a
= 0.05, ki = 0.7) for Milton, Florida (4.9 mm/day) and is lower than the
peak Eo rate (7.0 mm/day). Peak corn and peanut ET rates were more
closely described by the Penman ETp for Milton, based on a = 0.23 in
Table 3, 5.5 and 5.1 mm/day for June and July, respectively.
On the basis of these comparisons, we conclude that the Penman
method provides a good approximation of potential ET for crops during
full canopy. Measured ET values for tomato, bahiagrass, and sorghum
during full canopy and under well-watered conditions were in reasonable
agreement with Penman ET values. Peak ET rates of corn and peanut
crops were higher than ETp based on a = 0.05, ki = 0.7 but were better
described by ETp based on a = 0.23 in Equation 10. The ETrates of crops
prior to full ground cover are lower than ETp values for the same time
periods. The length of canopy development, full canopy, and senescing
canopy stages depends on crop, soil, climate, and management factors.
There are possibilities of using more sophisticated models for predicting
actual ET as climatic conditions fluctuate and the crop goes through its
developmental stages. The more promising of these approaches includes
modified Penman methods for predicting ETp and partitioning it into soil
evaporation and crop transpiration (Ritchie 1972; Kanemasu et al. 1976).
This area is beyond the scope of this publication. Rather, we will consider
the use of crop coefficients, which vary with crop development stages, to
predict crop ET from a reference ETp value calculated by Penman's
method with a = 0.05 and k1 = 0.7.
3.3 CROP COEFFICIENTS
Data presented in Figures 5-9 demonstrate the deviation of actual ET
from calculated ET, as crops develop. Therefore, ETp calculations should
be considered to be reference crop ETvalues corresponding to the rate of
5 First crop Second crop
5 1 15 1 15 1 15 1 15 1 15 15
May June July August Sept. Oct.
Fig. 7. Evapotranspiration (ET) for sorghum (reported by Doss et al. (1965) for
NOTE: See Note, Fig. 5.
2 First crop Second crop
15 1 15 1 15 1 15 1 15 1 15 1 15
May June July August Sept. Oct.
Fig. 7. Evapotranspiration (ET) for sorghum (reported by Doss et al. (1965) for
NOTE: See Note, Fig. 5.
o 1- -1
15 1 15 1 15 1 15 1 15
May June July August
Fig. 8. Evapotranspiration (ET) for tomatoes (reported by Saxena et al. (1971)
for Live Oak, Florida).
NOTE: See Note, Fig. 5.
ET "from an extended surface of 8 to 15 cm tall green grass cover of
uniform height, actively growing, completely shading the ground and not
short of water" (Doorenbos and Pruitt, 1977). Actual crop ET is esti-
mated by multiplying ETp estimates by a crop coefficient (kc), or:
ET = k, ET, (20)
Crop coefficients are empirical factors that describe the net effects of soil
and crop conditions on actual ET. They are measured under well-watered
conditions and depend on the rainfall and irrigation patterns and on the
method for estimating ET,. For example, the Blaney-Criddle and Thorn-
thwaite methods do not consider vapor pressure deficit (or aridity). Thus,
crop coefficients based on those methods are influenced by local climatic
conditions. Because local environmental conditions are taken into
account by the Penman method, the associated crop coefficients are
essentially independent of climate. The seasonal variations in crop coef-
ficients correspond roughly to incomplete soil cover, full canopy, and
senescing crop growth stages outlined previously (Jensen, 1968; Dooren-
bos and Pruitt, 1977). These crop coefficients thus depend on physiologi-
cal stage of crop growth and degree of canopy coverage.
In applying Equation 20 to predict ETfor a crop, estimates of ETp are
required. Several methods have been presented for calculating ETp.
Since values of ETp vary with method (Table 7), crop coefficients will also
vary and must be developed for use with a specific method for calculating
Examples of crop coefficients for citrus, pasture and turfgrass, and corn
are presented in Figures 10a, 10b, and 11, respectively. The kc values
shown by solid lines were taken from Soil Conservation Service Technical
Release No. 21 (1967). These kc values were developed from a modified
Blaney-Criddle formula. Also on each graph are crop coefficients (k')
calculated from data presented in this bulletin and based on the Penman
method for calculating ETp (a = 0.05, k, = 0.7). The curves may not be
directly comparable, since the curves from the Soil Conservation Service
7 E-T- ft.. -
Fig. 9. Evapotranspiration (ET) for peanuts (reported by Stansell et al. (1976)
for Tifton, Georgia).
NOTE: See Note, Fig. 5.
"-" rk, citrus (Soil Conservation Service, 1967)
---- k citrus, Fort Pierce
-- k, citrus, Lake Alfred
0.00 I I I II I I I
15 15 15 15 15 15 15 15 15 15 15 15
Jan. March May July Sept. Nov.
k, pasture grass (Soil Conservation Service, 1967)
.- k turfgrass, Fort Lauderdale
0.25 k. bahiagrass, Thorsby, Alabama
0.00 I I I I I
15 15 15 15 15 15 15 15 15 15 15 15
Jan. March May July Sept. Nov.
Fig. 10. Crop coefficients for (a) citrus and (b) pasture and turfgrass reported by
USDA-SCS (1967) for a modified Blaney-Criddle approach applied to
estimate ETp (kc) and those calculated from data presented in this report,
using Penman estimates of ETp (k').
2 0.6- -
0.4 k corn (Soil Conservation Service, 1967)
k -- corn (Thorsby, Alabama)
0.0 I I I I I I I
0 20 40 60 80 100
Percent of Growing Season
Fig. 11. Crop coefficients for corn reported by USDA-SCS (1967) for a modified
Blaney-Criddle approach to estimate ET, (kc) and those calculated from
data presented in this report, using Penman estimates of ET, (k').
publication are based on ET values for subhumid and arid conditions of
the Western States. There are, however, some important observations
that can be made.
The k' curve for citrus at Ft. Pierce does not follow the pattern for the
Blaney-Criddle method kc (Figure 10a). The citrus k' at Ft. Pierce varies
from 0.9 to 1.1 during the months June through January. From February
through May, k' varied between 0.7 and 0.9. The k' for citrus at Lake
Alfred followed a similar trend. The low k' values calculated for Febru-
ary through May were probably due to limited rainfall and low available
soil water during that time period.
In contrast, kc peaked at about 0.73 in the summer months and
dropped to about 0.65 in the winter. The Blaney-Criddle method nor-
mally overestimates ETp in the winter months, which would tend to
produce lower kc values. The citrus k' for both Florida locations was
higher than the citrus kc presented by Soil Conservation Service (1967).
The seasonal pattern of monthly turfgrass k' values was nearly con-
stant, varying between 0.8 and 0.9 most of the year (Figure 10b). The k'
values for Fort Lauderdale turfgrass were based on Penman ETp for
Hialeah. The pasture k' values for the bahiagrass data from Thorsby,
Alabama, were based on Penman ETp data for Milton, Florida. The k'
values were low in May (0.7), increased to a peak of 0.97 in July, and
dropped to 0.85 in September. The shape of this k' curve is similar to that
of an annual crop as it goes through its growth and maturity stages.
Turfgrass k' was more constant because of its annual growth pattern. In
the summer months, kc varied from 0.85 in April and September to a
peak of 0.92 in June and July. Winter kc for pasture dropped to 0.45 in
The corn ET data in Figure 5 were divided by Penman ETp data from
Milton, Florida, to obtain corn k'. The corn growing season for those
data extended from April 21 through August 15 (Doss et al. 1962). The
resulting data were plotted in Figure 11. The k' and kc for corn started the
season at about the same values of 0.42. The peak k' of 1.23 was larger
than the peak kc value of 1.08. However, the kc curve lagged behind the
kc curve for about 10% of the growing season, or about 12 days.
4.0 APPLICATION OF POTENTIAL
EVAPOTRANSPIRATION CONCEPTS TO
4.1 IRRIGATION TO PREVENT CROP YIELD REDUCTIONS
Assuming that other factors, such as fertility levels, pest or disease
activity, and climatic parameters, are not limiting, crop growth and yield
are maximized by maintaining optimum soil water content throughout
the growing season. Optimum soil water content for plant growth is
normally that near field capacity, such that capillary water potentials are
high and water can readily be removed from the soil by plants, yet not so
high that gaseous diffusion in the soil is restricted.
Many researchers have shown that crop dry matter production is
directly related to water use by the crop throughout its growth cycle
(Briggs and Shantz, 1914; Staple and Lehane, 1955; Allison et al., 1958;
Chang, 1968; Tanner, 1981; Tanner and Sinclair, 1983; deWit, 1958;
Stanhill, 1960; Viets, 1962; Arkley, 1963; Chang et al., 1963; Hanks et
al., 1969). On the other hand, it has been found that for grain and fruit
yields, sensitivity to drought varies with physiological growth stage (Hiler
and Clark, 1971; Sudar et al., 1981). These yield relationships are dis-
cussed under separate subheadings in the following sections.
4.1.1 Dry Matter Yield
Dry matter yield was related to cumulative transpiration by deWit
Y_ KCT (21)
where Y = yield (kg/ha)
K, = crop factor (kg/ha),
CT= cumulative transpiration from beginning of growth season
Eo = average free water surface evaporation rate (mm/day).
This equation allowed yield to be expressed as a function of water use
by the crop. It also implied that there were no critical periods of crop
growth during which water use was more important than at any other
For a given crop and yield, Hanks (1974) calculated relative yield as a
function of relative transpiration:
Y = CT (22)
YP C TP
where Yp = potential crop yield when water is not limiting (kg/ha),
and CTp = cumulative transpiration that occurs when soil water does
not limit transpiration (mm).
This equation was found to produce good results when the transpiration
component was separated from evapotranspiration. It also allowed the
prediction of relative yield at any time during the growing season as a
function of transpiration to that time.
One of the best examples of the linear relationship between dry matter
yield and evapotranspiration in the humid southeast can be found from
South Carolina lysimeter studies reported by Allison et al. (1958). These
data were obtained for several different crops (corn, millet, cowpeas,
crotalaria, cotton, and soybeans) and crop rotations over a 12-year period
from 1933 to 1945. These data are plotted in Figure 12, and give a linear
relationship with a correlation coefficient of 0.959. Under these ex-
perimental conditions, they estimated bare soil evaporation rate to be
about 406 mm per year.
Several other researchers including Arkley (1963), Hanks et al. (1969),
Leggett (1959), Powers et al. (1961) and Whittlesly and Colzar (1968)
verified the results given in Equations 21 and 22. The experimental
results summarized by these equations demonstrate the important fact
that a reduction in transpirational water use below potential rate results
in a concomitant decrease in crop biomass yield. Bierhuizen and Slayter
(1965), Tanner (1981), and Tanner and Sinclair (1983) showed that
diffusion of CO2 into the stomata and loss of water vapor from the
stomata was the coupling mechanism between biomass yield and ET.
4.1.2 Grain Yield
Grain yields are more sensitive to water shortages during certain stages
of growth than in others. Several approaches predict grain yields from
physically based models which relate water stresses during various stages
of crop growth to final yield. An approach frequently used is to interpret
transpiration reduction below potential levels, CT/CTp, as an integrator
of the effects of climatic conditions and soil water status on grain yield.
Y = 0.2634X 12.049
40 50 60 70 80
Fig. 12. The relationship between crop dry matter yields and water use (from
Allison et al., 1958).
Because it was observed that interactive effects between crop growth
stages existed (i.e. reduced vegetative growth during early stages caused
a reduction in photosynthetic material for fruit production at later
stages), multiplicative models were formulated.
One such model was that developed by Jensen (1968):
Y = ET ET 2 ... x ET Sn
Y, ETPT ETp2 ET,p
P = relative yield
relative evapotranspiration during the ith stage of physio-
ETP logical development
8i = factor expressing crop sensitivity to water stress during the
ith growth stage, i = 1,... n,
n number of plant growth stages.
The crop sensitivity factors for four growth stages of Bragg soybeans
calculated from yield data of Hiler and Clark (1971) were 0.24,0.48,0.84,
and 0.26 for vegetative, flowering, early pod fill, and late pod fill stages,
respectively. Those data indicated that water stress effects during the
early pod fill stage reduced yield more than stress at other growth stages.
Other researchers have found similar results with other field crops,
including corn, grain sorghum, and soybeans (Sudar et al. 1981). Effects
of water stress and reduced ET at various growth stages in yield reduc-
tions for many crops are summarized by Doorenbos (1979). With the
exception of ongoing research at the IFAS Irrigation Research and
Education Park, no known research of this type has been conducted
under Florida conditions. There is, however, evidence to indicate that the
general patterns of stress effects on yield are similar in this climatic
regime to those in other areas. The general relationships between corn,
peanut, and soybean yields and estimated ET (based on Penman and
daily water balance calculations) are shown in Figures 13, 14, and 15.
These data resulted from several years of water management studies on
well-drained sandy soils in North Central Florida (L. C. Hammond, et
al., unpublished data, Florida Agricultural Experiment Station, Gaines-
ville). Different levels of water use were obtained by imposing various
water management treatments-nonirrigated and different seasonal
amounts of irrigation. In some irrigation treatments, timing and amount
per application were managed so as to place the crop under water stress
for short periods throughout the season. The general relationships be-
tween grain yield and ET were linear, in agreement with the findings of
others (Stegman, et al. 1980). However, these linear responses may be
altered by the timing of stress. These data demonstrate the fact that
8- o 1977 o
y = 0.661x 22.90
S 6 P = 0.90
0 K4 I --- i --- ---------i
0 30 35 40 45 50 55
Simulated ET cm
Fig. 13. The relationship between grain yield and estimated ETfor corn grown in
deep sandy soil in north Florida (from L. C. Hammond, unpublished
data, Gainesville, Florida).
Y = 0.153X 328
r2 =0.85 o
36 38 40 42 44 46 48 50
Simulated ET, cm
Fig. 14. The relationship between crop yield and estimated ETfor peanuts grown
in deep sandy soil in north Florida (from L. C. Hammond, unpublished
data, Gainesville, Florida).
periodic droughts in Florida will result in yield reductions if irrigation
management practices do not supply enough water to meet the crop ET
Some research has been conducted to determine the effects of water
stress on yield and growth of citrus, but the effects have not been
quantified through a model such as Equation 23. In most field studies,
plant stresses are implied by low soil moisture availability. In general,
irrigation has been shown to increase fruit yield and rate of tree growth of
citrus in Florida (Koo, 1979) and in Arizona (Hilgeman, 1977). Also, for
fruit yield, irrigation has been shown to be more critical in January to
June than during the latter part of the year (Koo, 1963; Hilgeman, 1977),
but for tree growth, the response is related to irrigation throughout the
year (Koo, 1979; Hilgeman, 1977). Environmental conditions that pro-
duce plant water stresses are strongly implicated in excessive fruit abscis-
sion (Palmer, et al., 1977).
Y = 0.144X 4.12
/ = 0.88
0 --I I I
0 30 40 50
Simulated ET, cm
Fig. 15. The relationship between crop yield and estimated ET for soybeans
grown in deep sandy soil in north Florida (from L. C. Hammond,
unpublished data, Gainesville, Florida).
4.2 EFFECTIVE RAINFALL
Effective rainfall is defined as rainfall which is temporarily stored in the
soil and available to meet the ET requirements of crops. It does not
include runoff or percolation below the crop root zone. Effective rainfall
is used to estimate the supplemental irrigation requirements. Rainfall
amounts and temporal distributions, ETrates, and soil hydraulic charac-
teristics are the primary factors influencing rainfall effectiveness.
In humid areas such as Florida, the magnitude and intensity of storm
events are usually different in wet and dry seasons. In the wet season,
convective and tropical depression storms of large magnitude and high
intensity occur frequently. Once soil water is restored after the dry
season, these storms can produce a large amount of water in excess of that
which can be stored in the soil profile (Allen et al., 1982). This excess is
lost either to surface runoff or to percolation below the root zone. Thus,
the effectiveness of rainfall in the wet season is relatively low. Con-
versely, during the dry season frontal storms of small magnitude and low
intensity are typical. Under nonirrigated conditions, this amount of rain-
fall can usually be stored in the soil, resulting in a higher rainfall effective-
When the ET rate is high, the available moisture in the soil profile is
depleted rapidly, thus providing a relatively large storage capacity for
receiving rainfall. That is, a higher ETrate will increase the effectiveness
of rainfall, while lower ET rates will decrease rainfall effectiveness.
When the soil-water storage capacity is large, the potential to store
rainfall is high. Thus, the effectiveness of rainfall would be relatively
high. Conversely, if the soil-water storage capacity is low, only a small
amount of rainfall can be stored in the soil and the resulting rainfall
effectiveness will be low. Infiltration rates are usually lower and the
potential for runoff is greater for soils with greater water-holding capaci-
Antecedent soil water content influences the amount of rainfall which
can be stored. For this reason, an untimely or excessive irrigation would
reduce rainfall effectiveness. If soil water levels are maintained high by
irrigation, rainfall effectiveness will be lower than for nonirrigated areas
which are otherwise identical.
Soil Conservation Service scientists analyzed 50 years of rainfall rec-
ords at each of 22 locations to develop a predictive technique for effective
rainfall (1967). They predicted effective rainfall (ER) from:
ER = (0.83548R.82416 0.29352)(100.09553(E7))(F) (24)
where ER = average effective monthly rainfall, cm;
R, = total monthly rainfall, cm;
ET= monthly evapotranspiration, cm; and
F= soil water storage factor.
The soil water storage factor was defined by:
F= 0.531747 + 0.116206D 0.008943D2 + 0.000232D3 (25)
where D is the usable soil water storage in cm. D is usually calculated as
50% to 70% of the available water capacity of the soil, depending upon
the irrigation management practices used.
A solution to Equation 24 for D = 7.6 cm is given in Table 10. For
other values of D, the ER values must be multiplied by the corresponding
soil water storage factor given in the lower portion of Table 10.
Table 10. Average monthly effective rainfall' as related to mean monthly rainfall and average monthly consumptive use (adapted from
Soil Conservation Service, 1967).
Monthly Average Monthly Consumptive Use, ET, in cm
Mean 0.0 2.5 5.1 7.6 10.2 12.7 15.2 17.8 20.3 22.9 25.4
R, cm Average Monthly Effective Rainfall, ET, in cm
0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
1.3 0.71 0.76 0.81 0.86 0.91 0.97 1.02 1.07 1.14 1.19 1.27
2.5 1.50 1.60 1.68 1.78 1.88 1.98 2.11 2.24 2.36 2.49 2.54
3.8 2.21 2.36 2.49 2.62 2.77 2.95 3.10 3.28 2.48 3.68 3.80
5.1 2.90 3.07 3.23 3.43 3.63 3.84 4.04 4.29 4.52 4.78 5.05
6.4 3.53 3.73 3.96 4.19 4.42 4.67 4.95 5.23 5.54 5.84 6.20
7.6 4.39 4.65 4.93 5.21 5.51 5.42 6.15 6.50 5.88 7.26
8.9 5.03 5.33 5.64 5.97 6.30 6.65 7.04 7.44 7.87 8.33
10.2 5.66 5.99 6.32 6.69 7.09 7.49 7.92 8.36 8.84 9.35
11.4 6.63 7.01 7.42 7.85 8.28 8.76 9.27 9.80 10.36
12.7 7.26 7.67 8.13 8.59 9.07 9.60 10.16 10.74 11.35
14.0 NOTE: Values 7.87 8.33 8.81 9.32 9.86 10.41 11.02 11.66 12.32
15.2 below line exceed 8.97 9.50 10.03 10.62 11.23 11.86 12.55 13.28
16.5 monthly consump- 9.63 10.16 10.74 11.38 12.01 12.70 13.44 14.22
17.8 tive use and are to 10.28 10.82 11.46 12.12 12.80 13.54 14.33 15.14
19.0 be used for inter- 11.48 12.14 12.85 13.59 14.35 15.19 16.05
20.3 polation only 12.14 12.83 13.56 14.33 15.16 16.05 16.97
'Calculations assume 7.6 cm net depth of application. For other net depths of application, multiply by the factors shown below.
Net Depth of
Application (D) 1.9 2.5 3.8 5.1 6.4 7.6 10.2 12.7 15.2 17.8
Factor (F) .72 .77 .86 .93 .97 1.00 1.02 1.04 1.06 1.07
4 SI. *
The average monthly effective rainfall calculated by Equation 24 can-
not exceed either average monthly rainfall or average monthly ET. If the
application of this equation results in an ER which exceeds either, then
ER must be reduced to the lesser of the two.
Equations 24 and 25 were developed using a daily soil moisture balance
procedure. Such a procedure necessarily fails to consider two factors
which may affect ER. These factors, soil infiltration rates and rainfall
intensities, were not considered because sufficient data were not avail-
able, and they are too complex to be readily considered. If, in a specific
apphcation, infiltration rates are low and rainfall intensities are high,
large amounts of rainfall may be lost to surface runoff. Sloping land
surfaces would further reduce infiltration amounts. In such cases, the ER
values obtained from Equation 23 would need to be modified appro-
The Everglades Agricultural Area (EAA), as described in Section 3.1
and by Mierau (1974), was used as example to demonstrate the applica-
tion of Equations 24 and 25 for determining effective rainfall. The aver-
age available water capacity in the top 60 cm layer of Pahokee muck
(Lithic Medisaprists) is about 0.3 cm of water per cm of soil (Stewart et
al., 1963). If irrigations are scheduled at 50% of the available water
capacity, the usable soil water storage, D, is 9 cm.
Effective rainfall was computed from Equation 24, using rainfall data
(Table 5), potential ET from the Penman method (Table 6), and D = 9
cm. The results are shown in Table 11. Effective rainfall was computed to
be 825 mm compared with 1494 mm of rainfall (55% of rainfall). How-
ever, if only the irrigation period (Nov.-May) was considered, the effec-
tive rainfall was computed to be 312 mm compared with 468 mm of
rainfall (67% of rainfall).
Field data, based on a water budget for the EAA, showed that the
actual effective rainfall values were 305 mm during the irrigation season
and 805 mm on an annual basis. The good agreement between the field
data and the model predictions for effective rainfall implied that the
effective rainfall estimated from the Soil Conservation Service (1967)
method is a satisfactory approach for the south Florida organic soil
4.3 IRRIGATION REQUIREMENT
Irrigation requirement is defined as the quantity of water, exclusive of
precipitation, that is required by a crop growing without water stress. It is
calculated as the difference between evapotranspiration and effective
IRR = ET- ER (26)
Table 11. Example predictions of effective rainfall and irrigation requirement for the Everglades Agricultural Area, using average
weather conditions from 1924 to 1975.
Effective rainfall' Irrigation
(Penman ETp Irrigation actually
Month Rainfall method) Predicted Data2 Requirement applied'
Jan 51 48 32 26 16 22
Feb 50 62 33 21 29 15
Mar 82 86 54 43 32 28
Apr 75 107 53 56 54 53
May 120 117 81 111 36 41
SJun 231 113 113 112 -
Jul 218 113 113 107 -
Aug 209 110 110 119 -
Sep 224 96 96 114 -
Oct 144 81 81 46 -
Nov 44 56 29 24 27 22
Dec 46 46 30 26 16 32
Total 1494 1035 825 805 210 213
Nov-May 468 312 305 210 213
'The Penman method was used to estimate ET in the effective rainfall equation.
'Data from Mierau (1974) for a 10-year period, 1962-1971.
where IRR = irrigation requirement in mm,
ET= evapotranspiration in mm, and
ER = effective rainfall in mm.
For the EAA, predicted monthly irrigated requirements were calcu-
lated and compared with actual irrigation schedules reported by Mierau
(1974) (Table 11). No irrigation was required during the period June
through October, the wet season. This result cannot be extrapolated to
other parts of the state having different soil characteristics and rainfall
patterns. Many of the sandy soils in Florida have low water holding
capacities compared with the 9 cm for muck soil. Thus, short periods of
water stress may develop in the wet season and result in a need for crop
Extreme caution should be exercised in the use of Equation 26 to
compute irrigation requirements for any given year, because actual
irrigation requirements depend upon the frequency of occurrence and
magnitude of each rainfall event rather than long-term averages. The
fallacy of using long-term average monthly rainfall amounts for estimat-
ing irrigation requirements is demonstrated in Table 11. In only four
months (February, March, April, and November) does long-term aver-
age ETp exceed average rainfall. During these four months, predicted
ETp exceeded average rainfall by only 60 mm compared with 210 mm of
irrigation required during the year. This fallacy is more severe when
applied on an annual basis. Average annual rainfall exceeded average
annual ETp by 459 mm. However, its seasonal distribution results in the
observed irrigation needs. These data emphasize the need for continuous
daily evaluations of water balances to adequately determine the crop
irrigation requirements for a specific application.
4.4 EXAMPLE OF IRRIGATION MANAGEMENT
CONCEPTS FOR HUMID REGIONS
Many humid regions like Florida have variable rainfall distribution and
soils with low water-holding capacity. This combination does not sustain
optimal plant growth throughout the growing seasons of all crops. There-
fore, irrigation is needed, but normally in amounts less than 50% of the
total water needed to produce a crop. By contrast, in arid regions, nearly
all of the seasonal water needs of crops may be supplied by irrigation.
Moreover, water in excess of ET must be applied to leach excess salts to
avoid salt buildup in the plant root zone. Most of the arid region soils that
are used for crop production are medium to fine-textured and hence have
a relatively high water storage capacity.
Irrigation management systems for arid region conditions have been
developed over a long period of time. Currently, progress is being made
in the development of management systems for sandy soils with low water
storage capacities and to climates characterized by alternate periods of
rainfall excesses and deficiencies. The objective is to schedule irrigation
(time and amount) to maximize evapotranspirational use of irrigation
and rainfall while minimizing leaching loss of water, fertilizers, and
pesticides during the crop-growing season. A near ideal water manage-
ment result would be: (1) a soil water profile near full storage at planting
and about 50% depleted over a deep root zone at harvest, and (2) a
crop-season rainfall and irrigation distribution pattern such that plants
were never stressed and no drainage occurred below the root zone. The
ideal is not possible, given the real-world conditions of rainfall uncertain-
ties and low water-storage capacity of sandy soils. However, economic
and resource factors will make the achievement of improved water man-
agement practices a continuing goal.
Humid region conditions we have just described suggest a water man-
agement scheme as follows: (1) irrigation before the crop shows visible
water stress (wilting), and (2) application of an amount of water which
does not completely restore the water-depleted root zone. This scheme is
designed to assure a water supply equal to the ET demand of the atmo-
sphere and, at the same time, to allow the capture of more rainfall than
would be the case if the entire root zone was restored to capacity. It is not
easy to accomplish this management objective, and much current re-
search deals with this problem. However, the following example calcula-
tions demonstrate the principles involved and, in particular, the use of
estimated ET and soil water characteristic data.
The example is for central Florida during an intermittent drought in
May and June. The needed daily ET estimates are calculated from the
monthly values of ETp (a = 0.05, kl = 0.7) for Lakeland (Table 3)..The
following soil water profile data were used: maximum water storage in
the root zone, 9 cm; stored water not readily available to plants, 2 cm;
water content available to plant, 9 2 = 7 cm. It is assumed that a
depletion of up to 55% of the available water supply (4 cm in this case)
would not cause plant water stress. Also, assume that a 2-cm irrigation
can be made on the day when a water need is calculated.
The water management plan contains two scheduling scenarios: (1) the
initial irrigation after depletion from a full soil water profile, and (2)
subsequent irrigations in a continuing drought.
May 1 Assume a full soil profile (7 cm of available water); ET =
0.445 cm/day during May; negligible drainage.
10 Initial irrigation, 2 cm; 4 cm of water have been depleted
by 9 days of ET, and the irrigation leaves 2 cm of storage
14 Subsequent irrigation, 2 cm; estimated ET (4 days) is 1.8
cm or nearly equal the 2-cm May 10 irrigation.
19 Subsequent irrigation, 2 cm; estimated ET of 2.2 cm (5
days); 2 cm storage unfilled; no drainage.
20 Rainfall of 3 cm; estimated ET of 0.4 cm (1 day); profile
temporarily overfilled; drainage loss of 0.6 cm; root zone
restored to 7 cm of available water as on May 1.
29 Initial irrigation, 2 cm; estimated ET of 4 cm (9 days);
rainfall of 1 cm after irrigation; 1 cm of storage unfilled;
June 5 Subsequent irrigation, 2 cm; estimated ET of 3.2 cm (7
days, 3 at 0.445 cm/day and 4 at 0.47 cm/day); 2.2 cm
storage unfilled; no drainage.
June 9 Subsequent irrigation, 2 cm; estimated ET of 1.9 cm (4
days); 2.1 cm storage unfilled; no drainage.
June 11 Rainfall of 4.3 cm; estimated ET of 0.9 cm (2 days);
storage temporarily overfilled; drainage of 1.3 cm; root
zone restored to capacity as on May 1.
The consequences of rainfall on May 20 and 29 and on June 11 demon-
strate the value of a management practice which provides some soil
storage capacity for trapping rainfall. To use the above or any other
management strategy successfully, the irrigator will need to continually
test the consequences of his irrigation decisions. Do the plants wilt a day
or so before irrigation is scheduled? Is the soil wetted to a depth of 20 to
30 cm about a day after irrigation? On the other hand, is the soil wetted to
greater depths-an indication that unfilled storage is less than calculated?
Are the measured or estimated soil and plant characteristics (available
water, root depth, plant canopy), appropriate for the major portion of
the specific field being irrigated?
Additional questions should be considered. What is the forecast for
rainfall? Are the weather conditions (clouds, relative humidity and
temperature) such that estimates of ET need to be adjusted up or down
during the rain-free periods? What is the crop condition-stage of growth
and degree of ground cover-and appropriate k' value? How much
leaching loss of water and nutrients has occurred? Is the loss of mobile
nutrients serious enough to require correction by top dressing with fertil-
izer? Is the irrigation system design appropriate to the soil-crop-land area
In spite of the obvious complexity of the irrigation management prob-
lem, progress is being made in developing more efficient management
systems for crop production in Florida. The aspects of the problem,
treated briefly here, emphasize the need for more knowledge about the
crop's physical and biological environment and the grower's management
potential in it.
Evapotranspiration is the natural loss of water vapor from soil, vegeta-
tion, and open water surfaces to the atmosphere. Evapotranspiration
from a well-watered, active crop with full ground cover is determined
primarily by meteorological processes and is referred to as potential
evapotranspiration. Potential ET can be estimated using net radiation,
temperature, saturation vapor pressure deficit or dryness of the air, and
windspeed. Surface factors also influence ET. Soil water availability,
vegetative cover, and the albedo of soil and vegetation interact to compli-
cate the procedures for estimating actual ET.
Potential ET was discussed as a useful concept in understanding the
factors affecting actual ET and as a basis for estimating actual ET for
various crops. The Penman method for estimating ETp was described.
Different surface albedo values were used to demonstrate the use of the
Penman method in estimating evaporation from a free water surface, Eo
(a = 0.05), and ET, for a vegetative crop surface (a = 0.23). Because of
the difficulty in estimating average surface albedo for crops as they
emerge, expand to cover the soil, and then mature and die, the approach
developed by van Bavel and Verlinden (1956) was presented. In this
approach, Eo is estimated by using a = 0.05 in the Penman equation, and
the result is multiplied by a kl value of 0.7 to provide a reference estimate
of ETp. This reference ET, value can then be modified by an empirical
coefficient that varies with time in the season due to the factors discussed
above. The use of this approach may result in ETp estimates lower than
actual ET as demonstrated for corn in Figure 6 and corresponding k'
values that are larger than 1.0. However, as a practical tool, this proce-
dure eliminates the requirement for descriptions of how surface albedos
change with crop stage and ground cover. More sophisticated models that
estimate actual ET have been developed and show promise. However,
unless specific data are available, it is recommended that the Penman
method using ac = 0.05 and kl = 0.7 be used to estimate a reference ET,
value for Florida conditions. The k' crop coefficients presented in this
bulletin are based on ET, estimates using this approach.
A crop water budget method was used to estimate annual monthly ETp
for various watersheds in Florida and for various crops. In addition, field
plot and lysimeter data were used to estimate daily ET for various crops
in Florida and the humid Southeast region. These water balance esti-
mates of ET were then compared with other methods for estimating ETp:
the Penman, Thornthwaite, pan evaporation, Blaney-Criddle, a solar-
radiation modified Blaney-Criddle, and Stephens-Stewart methods. It
was demonstrated that the Penman and Stephens-Stewart methods were
superior to the other methods tested. However, the Penman method was
adopted as a standard because it is based on physical derivations with less
empiricism. The Stephen-Stewart method, while convenient, requires
calibration for different climates. Crop coefficients, based on the Penman
ETp estimates, were presented for citrus, turfgrass, bahiagrass, and corn
for Florida conditions.
The importance of ET as a concept in irrigation management was
discussed. Irrigation management practices that reduce ETbelow ETp for
a well-watered crop will reduce dry matter growth in an amount directly
proportional to the reduction in ET. More importantly, grain yields of
crops are reduced in proportion to the reduction in ET caused by periodic
droughts and insufficient irrigation. The dependence of grain and pod
yields of corn, soybean, and peanuts on seasonal ET were presented to
illustrate the importance of optimal irrigation practices. Research in
other states indicates that grain yield of crops is more sensitive to short-
ages of water during certain stages of growth than in others. More
research is needed to quantify such relationships for Florida climate and
A technique was presented for estimating monthly irrigation require-
ments based on monthly ETp, water-holding capacity of soil, and monthly
effective rainfall. This technique is recommended for use in estimating
monthly irrigation requirements on a large area basis for planning pur-
poses. It is not recommended for use in day-to-day crop irrigation man-
agement decisions. An example was discussed to illustrate the application
of water management principles in humid regions.
APPENDIX I. NOTATION
The following symbols are used in this bulletin:
a = superscript coefficient for Thornthwaite method;
CT = cumulative transpiration from the beginning of growth stage,
CTp = cumulative transpiration that occurs when soil water does
limit transpiration, mm;
D = 50-70% of the available water capacity of the soil, cm;
E = evaporation, mm;
EAA = Everglades Agricultural Area;
E, = 0.263 (ea ed)(0.5 + 0.0062u2);
ea = vapor pressure of air = (em,, + emi,)/2, mbar;
ed = vapor pressure at dewpoint temperature, mbar;
em, = maximum vapor pressure of air during a day, mbar;
ei, = minimum vapor pressure of air during a day, mbar;
Eo = average free water surface evaporation rate, mm/day;
Ep = potential evaporation, mm;
ER = average effective monthly rainfall, cm;
EST = eastern standard time in United States;
ET= evapotranspiration, mm;
ETp = potential evapotranspiration, mm;
e(T) = saturated air vapor pressure as a function of air temperature,
ETIETp = relative evapotranspiration during a given stage of physiolog-
F= soil water storage factor;
f= monthly evapotranspiration factor = 25.4 PD (1.8 Tm +
G = soil heat flux density, cal/cm2 day;
H = sensible heat flux density, cal/cm2 day;
I = Thornthwaite's temperature-efficiency index;
i= heat index;
IR = irrigation, mm;
IRR = irrigation requirement, mm;
j = index of growth stage;
kc = crop coefficient from Soil Conservation Service (1967), based
on modified Blaney-Criddle estimates of ET,;
k' = crop coefficient based on Penman method estimates of ETp;
k, = coefficient for Penman method;
k2 = coefficient for pan evaporation method;
k3 = coefficient for Thornthwaite method;
k4 = coefficient for Blaney-Criddle method;
k5 = coefficient for modified Blaney-Criddle method;
km = crop factor;
Ld = daytime hours divided by 12;
MR,, = monthly net solar radiation in cal/cm2;
MR, = monthly incoming solar radiation, cal/cm2;
n = number of growth stages;
P = flux density of solar radiation stored as chemical energy in
photosynthesis process, cal/cm2 day;
PD = percent of annual daylight hours in the month;
PE = evaporation from U.S. Weather Bureau standard class pan,
PET= monthly potential evapotranspiration estimated by Thorn-
thwaite's method, mm;
PN = percolation from root zone, mm;
PT = plant transpiration, mm;
PTp = potential plant transpiration, mm;
Rb = net outgoing thermal or long wave radiation, cal/cm2 day;
RO = runoff, mm;
RF = rainfall, mm;
R, = net radiation, cal/cm2 day;
R, = incoming solar radiation (direct and diffuse), cal/cm2 day;
R, = total monthly rainfall, cm;
Rso = cloudless sky radiation, cal/cm2 day;
R, = thermal radiation from the earth's surface (upward), cal/cm2
S = percent sunshine hours;
T= average air temperature in K, (273 + C);
T,,g = average air temperature = (T,, + Tmi)/2, C;
T, = dewpoint temperature, C;
T, = monthly average air temperature, C;
T,,ax = maximum daily temperature, C;
Tmin = minimum daily temperature, C;
TMR, = annual sum of mean monthly solar radiation in cal/cm2;
u2 = wind speed at height of 2 m, km/day;
uz = wind speed at height of z m, km/day;
Y= crop yield, kg/ha;
Yp = potential crop yield when water is not limiting for production,
Y/Yp = relative crop yield;
z = height of wind speed measurement, m;
a = albedo (reflectivity of solar radiation);
X = latent heat of vaporization of water, cal/cm2 mm;
y = psychrometric constant, mbar/C;
A = slope of saturated vapor pressure curve of air, mbar/C;
AS = change in soil water storage;
a = Stefan-Boltzmann constant (11.71 x 10-8 cal/cm2 day/K);
b8 = crop sensitivity to water stress during the jth growth stage.
U.S. CUSTOMARY-SI CONVERSION FACTORS
To Convert To by
inches (in.) millimeters (mm) 25.40
inches (in.) centimeters (cm) 2.540
inches (in.) meters (m) 0.0254
feet (ft) meters (m) 0.3048
yards (yd) meters (m) 0.9144
miles kilometers (km) 1.609
square inches (sq in.) square centimeters (cm2) 6.45
square feet (sq ft) square meters (m2) 0.093
square yards (sq yd) square meters (m2) 0.836
acres (acre) hectares (ha) 0.4047
square miles (sq miles) square kilometers (km2) 2.59
cubic inches (cu in.) cubic centimeters (cm3) 16.39
cubic feet (cu ft) cubic meters (m3) 0.02832
cubic yards (cu yd) cubic meters (m3) 0.765
acre-feet (acre-ft) cubic meters (m3) 1233.5
gallons (gal) liter (L) 3.79
D. WEIGHT AND YIELDS
pounds (lb) kilogram (kg) 0.4536
tons (ton) megragram (Mg) 0.9072
pounds per acre (lb/acre) kg per hectare (kg/ha) 1.121
feet per second (ft/sec) meters per second (m/sec) 0.3048
mile per day (mile/day) kilometers per day
To Convert To by
atmosphere (atm) kilopascals (kPa) 101.3
= 76 centimeters of
mercury (cm Hg)
= 1.013 bar
inch of mercury (inch Hg) kilopascals (kPa) 3.386
= 0.334 atm
= 0.0339 bar
inch of water (inch HzO) pascals (Pa) 249
= 2.49 millibar (mbar)
millibar (mbar) pascals (Pa) 100
= 0.75 millimeters of
mercury (mm Hg)
pounds per square inch kilopascals (Pa) 6.894
= 51.72 millimeters of
mercury (mm Hg)
G. RADIATION TO EQUIVALENT DEPTH
OF EVAPORATION (at 290C)
cal/cm2 mm 1/58
cal/cm2 min mm/hr 1.035
cal/cm2 day mm/day 1/58
mW/cm2 mm/hr 0.01483
W/m2 mm/hr 0.1483
1. Allen, L. H., Jr., J. S. Rogers, and E. H. Stewart. 1978. Evapotranspiration
as a benchmark for turfgrass irrigation. Proc. of the 26th Annual Florida
Turf-Grass Management Conference, Vol. XXVI:85-97.
2. Allen, L. H., Jr., W. G. Knisel, Jr., and P. Yates. 1982. Evapotranspiration,
rainfall, and water yield in South Florida research watersheds. Proc. Soil
Crop Sci. Soc. Fla. 41:127-139.
3. Allison, F. E., E. M. Roller, and W. A. Raney. 1958. Relationship between
evapotranspiration and yields of crops grown in lysimeters receiving natural
rainfall. Agron. J. 50:253-259.
4. Arkley, R. J. 1963. Relationships between plant growth and transpiration.
5. Bierhuizen, J. F., and R. O. Slatyer. 1965. Effect of atmospheric concentra-
tion of water vapor and CO2 in determining transpiration-photosynthesis
relationships of cotton leaves. Agric. Meteorol. 2:259-270.
6. Blaney, H. F., and W. D. Criddle. 1950. Determining water requirements in
irrigated areas from climatological and irrigation data. U.S. Dept. Agr., Soil
Conservation Service. SCS-TP-96, 48 pp.
7. Blaney, H. F., and W. D. Criddle. 1962. Determining consumptive use and
irrigation water requirements. U.S. Dept. Agr., Agricultural Research Ser-
vice, Tech. Bul 1275. 59 pp.
8. Bosen, J. F. 1960. A formula for approximation of saturation vapor pressure
over water. Monthly Weather Rev. 88:275-276.
9. Briggs, L. J., and H. L. Shantz. 1914. Relative water requirements of plants.
Agr. Res. 3:1-64.
10. Burman, R. O., P. R. Nixon, J. L. Wright, and W. O. Priutt. 1981. Water
requirements. Pp. 189-232 in Design and Operation of Farm Irrigation
Systems (ed., M. E. Jensen). American Society of Agricultural Engineers
Monograph #3, St. Joseph, Mich.
11. Butson, K. D., and G. M. Prine. 1968. Weekly rainfall frequencies in
Florida. Cir. S-187, Agr. Expt. Station, Inst. of Food and Agr. Sci., Univ. of
Fla., Gainesville. 41 pp.
12. Campbell, R. B., and C. J. Phene. 1976. Estimating potential evapotrans-
piration from screened pan evaporation. Agric. Meteorol. 16:343-352.
13. Carlisle, V. W., R. E. Caldwell, F. Sodek, L. C. Hammonid, F. G. Calhoun,
M. S. Granger, and H. L. Breland. 1978. Characterization data for selected
Florida soils. Soil Science Research Report No. 78-1. University of Florida,
Institute of Food and Agricultural Sciences, Soil Science Department, Soil
Characterization Laboratory, in cooperation with the U.S. Department of
Agriculture, Soil Conservation Service. 335 pp.
14. Chang, J. H., R. B. Campbell, and F. E. Robinson. 1963. On the relationship
between water and sugar cane yield in Hawaii. Agron. J. 55:450-453.
15. Chang, J. H. 1968. Climate and agriculture, an ecological survey. Aldine
Publishing Co., Chicago. 304 pp.
16. deWit, C. T. 1958. Transpiration and crop yields. Institute voor Biologish en
Scheikungdi Onderzoek van Landbouwgewassen. Wageningen, Medeling..
17. Doorenbos, J., and W. D. Pruitt. 1977. Guidelines for predicting crop water
requirements. Food and Agriculture Organization of the United Nations,
Irrigation and Drainage Paper No. 24. 144 pp.
18. Doorenbos, J. 1979. Yield response to water. Food and Agriculture Orga-
nization of the United Nations, Irrigation and Drainage Paper No. 33.
19. Doss, B. D., O. L. Bennett, and D. A. Ashley. 1962. Evapotranspiration by
irrigated corn. Agron. J. 54:497-498.
20. Doss, B. D., O. L. Bennett, and D. A. Ashley. 1965. Moisture used by
various plant species and its relation to pan evaporation and net radiation.
ARS 41-112, U.S. Department of Agriculture. 16 pp.
21. Fritz, S., and J. H. MacDonald. 1949. Average solar radiation in United
States. Heating and ventilating, Vol. 46:61-64.
22. Grubb, H. F. and A. T. Rutledge. 1979. Long-term water supply potential,
Green Swamp Area, Florida. U.S. Geological Survey. Water Resources
Investigations No. 78-99, Tallahassee, Florida. 76 pp.
23. Hanks, R. J., H. R. Gardner, and R. L. Florian. 1969. Plant growth-
evapotranspiration relations for several crops in the Central Great Plains.
Agron. J. 61:30-34.
24. Hanks, R. J. 1974. Model for predicting plant yield as influenced by water
use. Agron. J. 60:660-665.
25. Hiler, E. A., and R. N. Clark. 1971. Stress-day index to characterize effects
of water stress on crop yields. Trans. Am. Soc. Agr. Eng. 14(4):757-761.
26. Hilgeman, R. H. 1977. Response of citrus trees to water stress in Arizona.
Proc. Int. Soc. Citriculture 1:70-74.
27. Jensen, M. E. 1968. Water consumption by agricultural plants. Pp. 1-22 in T.
T. Kozlowski (ed.) Water deficits and plant growth 2, Academic Press, New
28. Jensen, M. E. (ed). 1974. Consumptive use of water and irrigation water
requirements. Technical Committee on Irrigation Water Requirements,
Irrigation and Drainage Div. ASCE, New York. 215 pp.
29. Jensen, M. E., and H. R. Haise. 1963. Estimating evapotranspiration from
solar radiation. J. Irr. and Drain. Div., Am. Soc. Civ. Engr. 89:15-41.
30. Kanemasu, E. T., L. R. Stone, and W. L. Powers. 1976. Evapotranspiration
model tested for soybean and sorghum. Agron. J. 68:569-572.
31. Knisel, W. G., Jr., P. Yates, J. M. Sheridan, T. K. Woody, III, L. H. Allen,
Jr., and L. E. Asmussen. 1984. Hydrology and Hydrogeology of Taylor
Creek Watershed, Okeechobee County, Florida. USDA Technical Bulletin
32. Koo, R. C. J. 1963. Effects of frequency of irrigation on yields of oranges and
grapefruit. Proc. Fla. State Hort. Soc. 75:1-5.
33. Koo, R. C. J. 1979. The influence of N, K, and irrigation on tree size and fruit
production of 'Valencia' orange. Proc. Fla. State Hort. Soc. 92:10-13.
34. Koo, R. C. J., and J. W. Sites. 1955. Results of research and response of
citrus to supplemental irrigation. Proc. Soil Sci. Soc. Fla. 15:180-190.
35. Leach, S. D. 1978. Source, use, and disposition of water in Florida. Water
Resources Investigation 78-17, U.S. Geological Survey, Water Resources
Division, Tallahassee, Florida. 90 pp.
36. Leggett, G. E. 1959. Relationships between wheat yield, available soil mois-
ture, and available nitrogen in eastern Washington dryland areas. Bull. 609.
Washington Agricultural Experiment Station.
37. Linsley, R. K., M. A. Kohler, and J. L. H. Paulhus. 1975. Hydrology for
Engineers. McGraw-Hill Book Co., Inc., New York.
38. Mierau, R. 1974. Supplemental water use in the Everglades Agricultural
Area. Technical Publication No. 74-4, South Florida Water Management
District, West Palm Beach, Florida.
39. Monteith, J. L. 1964. Evaporation and environment. Pp. 205-234 in The
State and Movement of Water in Living Organisms, Symp. Soc. Exp. Biol.,
40. NOAA. 1978. Climatological data for Florida, Vol. 82.
41. Palmer, R. L., Z. Hanscom, and W. M. Duggar. 1977. Studies of high
temperature effects on fruit drop from Washington Navel Orange: 1. Interac-
tion of temperature and leaf water potential. Proc. Int. Soc. Citriculture
42. Penman, H. L. 1948. Natural evaporation from open water, bare soil and
grass. Proceedings of the Royal Society, Series A. 193:120-145.
43. Penman, H. L. 1956. Evaporation: An introduction survey. Neth. J. Agr.
44. Penman, H. L. 1963. Vegetation and Hydrology. Tech. Comm. No. 53,
Commonwealth Bureau of Soils. Harpenden, England. 125 pp.
45. Phung, T., and J. F. Bartholic. 1976. Water balance in a peach orchard. In
Disposition of water from fruit crops and approaches to increase water use
efficiency. Water Resources Research Center, Publication No. 33, OWRT
Project No. B-014-FLA.
46. Powers, J. F., D. L. Grimes, and B. A. Reichman. 1961. The influence of
phosphorus fertilization and moisture on growth and moisture absorption by
spring wheat. In Plant growth, uptake, and moisture use. SSSA Proc. 25:209-
47. Pride, R. W., F. A. Meyer, and R. N. Cherry. 1966. Hydrology of Green
Swamp Area in central Florida. Report of investigations No. 42, Florida
Geological Survey and U.S. Geological Survey. 137 pp.
48. Ritchie, J. T. 1972. Model for predicting evaporation from a row crop with
incomplete cover. Water Resour. Res. 8:1204-1213.
49. Rogers, J. S., L. H. Allen, and D. V. Calvert. 1983. Evapotranspiration from
a developing citrus grove in a humid climate. Trans. Am. Soc. Agr. Eng.
50. Saxena, G. K., L. C. Hammond, and W. H. Lundy. 1971. Effect of an asphalt
barrier on soil water and on yields and water use by tomato and cabbage. J.
Amer. Soc. Hort. Sci. 96:218-222.
51. Saxton, K. E. 1982. Mathematical modeling of evapotranspiration on agri-
cultural watersheds. Proc. Int. Symp on Rainfall-Runoff Modeling, May
18-21, 1981, Mississippi State, Mississippi, (In press).
52. Shih, S. F., D. L. Myhre, J. W. Mishoe, and G. Kidder. 1977. Water
management for sugarcane production in Florida Everglades. Proc. Int. Soc.
of Sugar Cane Technologists, 16th Congress (Sao Paulo, Brazil) 2:995-1010.
53. Shih, S. F., and G. Gascho. 1980. Water requirement for sugar cane produc-
tion. Trans. Am. Soc. Agr. Eng. 23(4):934-937.
54. Shih, S. F., L. H. Allen, Jr., L. C. Hammond, J. W. Jones, J. S. Rogers and
A. G. Smajstrla. 1983. Basinwide water requirement estimation in Southern
Florida. Trans. Am. Soc. Agr. Eng. 26(3):760-766.
55. Shih, S. F., G. S. Rahi, and D. S. Harrison. 1982. Evapotranspiration studies
on rice in relation to water use efficiency. Trans. Am. Soc. Agr. Eng.
56. Soil Conservation Service. 1967. Irrigation water requirements. Technical
Release No. 21. USDA Soil Conservation Service, Engineering Division.
57. Spier, W. H. 1962. Installation and operation of nonweighing lysimeters.
Proc. Soil Crop Sci. Soc. Fla. 22:167-176.
58. Stanhill, G. 1960. The relationship between climate and the transpiration and
growth of pastures. In Proceedings, Eighth International Grassland Con-
59. Stansell, J. R., J. L. Shepherd, J. E. Pallas, R. R. Bruce, N. A. Minton, D.
K. Bell, and L. W. Morgan. 1976. Peanut responses to soil water variables in
the southeast. Peanut Sci. 3:44-48.
60. Staple, W. J. and J. J. Lehane. 1955. The influence of field shelterbelts on
wind velocity, evaporation, soil moisture, and crop yield. Canadian J. Agr.
61. Stegman, E. C., J. T. Musick, and J. I. Stewart. 1980. Irrigation water
management. Pp. 763-816 in Design and operation of farm irrigation systems
(edited by M. E. Jensen) ASAE Monograph 3. Amer. Soc. Agric. Eng., St.
62. Stephens, J. C. 1965. Estimating evaporation from insolation. J. Hydr. Div.,
Am. Soc. Civ. Eng. 91(HY5): 171-182.
63. Stephens, J. C., and E. H. Stewart. 1963. A comparison of procedures for
computing evaporation and evapotranspiration. Publ. No. 62, Trans. Intern.
Union of Geodesy and Geophysics, Berkeley, Calif. Pp. 123-133.
64. Stewart, E. H., D. P. Powell, and L. C. Hammond. 1963. Moisture charac-
teristics of some representative soils of Florida. ARS 41-63. USDA, Agr.
65. Stewart, E. H. and W. C. Mills. 1967. Effect of depth to water table and plant
density on evapotranspiration rate in Southern Florida. Trans. Am. Soc.
Agr. Eng. 10:746-747.
66. Stewart, E. H., J. E. Browning, andE. O. Burt. 1969. Evapotranspiration as
affected by plant density and water-table depth. Trans. Am. Soc. Agr. Eng.
67. Stewart, J. B. 1980. Areas of natural recharge to the Floridan Aquifer in
Florida. Map Series 98. U.S. Geol. Surv. in Coop. with Fla. Dept. Environ.
Reg., Bureau Water Mgt., Tallahassee, Florida.
68. Sudar, R. A., K. E. Saxton, and R. G. Spomer. 1981. A predictive model of
water stress in corn and soybeans. Trans. Am. Soc. Agr. Eng. 24:97-102.
69. Tanner, C. B., and W. L. Pelton. 1960. Potential evapotranspiration esti-
mates by the approximate energy balance method of Penman. J. Geophys.
70. Tanner, C. B., and M. Fuchs. 1968. Evaporation from unsaturated surfaces:
A generalized combination method. J. Geophys. Res. 73:1299-1304.
71. Tanner, C. B. 1981. Transpiration efficiency of potato. Agron. J. 73(1):59-
72. Tanner, C. B. and T. R. Sinclair. 1983. Efficient water use in crop produc-
tion: Resarch or re-research. In Limitations to Efficient Water Use in Crop
Production (H. M. Taylor, W. R. Jordan, T. R. Sinclair, eds.). Am. Soc.
Agron. Monograph, pp. 1-27.
73. Thornthwaite, C. W. 1944. Report of the Committee on Transpiration and
Evaporation. 1943-1944. Trans. Am. Geophys. Un. 25:683-693.
74. van Bavel, C. H. M., and F. J. Verlinden. 1956. Agricultural drought in
North Carolina. Tech Bull. No. 122. North Carolina Agricultural Experi-
ment Station, Raleigh, North Carolina.
75. Viets, F. G. 1962. Fertilizers and the efficient use of water. Advances in
76. Whittlesly, N. K., and L. R. Colzar. 1968. Decision making under conditions
of weather uncertainty in the summer fallow-annual cropping area of eastern
Washington. Tech. Bull. No. 58. Washington Agricultural Experiment Sta-
77. Yates, P., L. H. Allen, W. G. Knisel, and J. M. Sheridan. 1982. Channel
Modification effects on Taylor Creek Watershed. Pp. 78-86 in Environmen-
tally Sound Water and Soil Management, Proc. of the ASCE Irrigation and
Drainage Division Specialty Conf. (E. G. Kruse, C. R. Burdick, and Y. A.
Yousef, eds.). ASCE, New York.
This publication was promulgated at an annual cost of $4943 or a
cost of $1.24 per copy to provide information on potential evapo-
transpiration for various watersheds in Florida and for various crops.
All programs and related activities sponsored or assisted by the Florida Agricultural
Experiment Stations are open to all persons regardless of race, color, national origin,
age, sex, or handicap.