• TABLE OF CONTENTS
HIDE
 Acknowledgement
 Cover
 Title Page
 Table of Contents
 List of Tables
 Preface
 Abstract
 Introduction
 Figure 1: Water supply areas within...
 Figure 2: Structure and information...
 MAIN and PARAM programs
 Contents, arrays, and variables...
 Table 1: Defenitions of contents,...
 Data and information flows in MAIN...
 Subroutine control in MAIN
 Predicting sugarcane yield...
 Contents, arrays, and variables...
 Operation of SCYIELD
 Estimation of the sugarcane ET...
 SOILK for non-organic soils
 CROPK for citrus, pasture, golf...
 Table 2: Crop coefficient (CROPK)...
 Soil water relations and data...
 Figure 3: Desorption moisture-tension...
 Table 3: Example calculations from...
 Figure 4: Water gain and release...
 Table 4: Water storage capacity,...
 Soil water relations and data source...
 Empirical/statistical model
 Table 5: Definitions of constants,...
 Operation of WTPRED continued
 Characteristics and nature...
 Table 6: Definitions of constants,...
 Table 7: Saturated water content...
 Definition of a "day"
 ET estimates for each day
 Definition and removal of drainage...
 Generalized "pot" soil model...
 Table 8: Definitions of constants,...
 Defenition of a "day"
 End of "day" update of available...
 Table 9: Defintions of constants,...
 Water year vs. crop year calcu...
 Constants, arrays, and variables...
 Table 10: Definitions of constants,...
 Table 11: Agricultural and other...
 Table 12: Estimated growing season...
 Vegetable acreage production...
 Table 13: Estimated vegetable production...
 Sugarcane acreage and producti...
 Acreage of vegetable crops in planting...
 Pasture and turf acreage
 Table 15: Goal water table depths...
 Water balance projections
 Figure 5: Historical and predicted...
 Figure 7: Historical and predicted...
 Table 16: Actual and predicted...
 Estimates of evapotranspiratio...
 Economic impact projections
 Table 17: Estimates of evapotranspiration...
 Table 18: Estimated and potential...
 Table 19: Crop acreage and return...
 Conclusion
 Bibliography
 Appendices
 Appendix A: Program flowcharts
 Appendix B: Fortran program...
 Appendix C: Statistical considerations...
 Appendix D: Statistical analysis...
 Appendix F: Data for model...






Group Title: Bulletin / Agricultural Experiment Station ;, 849
Title: Area-wide agricultural water demand projection model for south Florida
CITATION PAGE IMAGE ZOOMABLE PAGE TEXT
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00026758/00001
 Material Information
Title: Area-wide agricultural water demand projection model for south Florida technical documentation, version 2.2
Series Title: Bulletin Agricultural Experiment Station
Physical Description: ix, 191 p. : ill. ; 23 cm.
Language: English
Creator: Lynne, Gary D
University of Florida -- Food and Resource Economics Dept
Publisher: Food and Resource Economics Dept., Institute of Food and Agricultural Sciences, University of Florida
Place of Publication: Gainesville Fla
Publication Date: 1987
 Subjects
Subject: Water consumption -- Computer simulation -- Florida   ( lcsh )
Water-supply -- Computer simulation -- Florida   ( lcsh )
Agriculture -- Water-supply -- Florida   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Bibliography: Bibliography: p. 61-63.
Statement of Responsibility: G.D. Lynne ... et al..
General Note: "December 1987"--Cover.
General Note: "Version 2.2 was designed for the seven water supply areas south of Lake Okeechobee, Florida, encompassing a large vegetable and sugarcane producing area"--Abstract.
Funding: Florida Historical Agriculture and Rural Life
 Record Information
Bibliographic ID: UF00026758
Volume ID: VID00001
Source Institution: Marston Science Library, George A. Smathers Libraries, University of Florida
Holding Location: Florida Agricultural Experiment Station, Florida Cooperative Extension Service, Florida Department of Agriculture and Consumer Services, and the Engineering and Industrial Experiment Station; Institute for Food and Agricultural Services (IFAS), University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: aleph - 001492020
oclc - 20366425
notis - AHA4279

Table of Contents
    Acknowledgement
        Acknowledgement
    Cover
        Cover
    Title Page
        Title Page
    Table of Contents
        iii
        iv
    List of Tables
        v
        vi
        vii
    Preface
        viii
    Abstract
        ix
    Introduction
        Page 1
        Page 2
    Figure 1: Water supply areas within the study area
        Page 3
    Figure 2: Structure and information flow of Fortran program, version 2.2
        Page 4
    MAIN and PARAM programs
        Page 5
    Contents, arrays, and variables (CAV) in MAIN and PARAM
        Page 6
    Table 1: Defenitions of contents, arrays, and variables in MAIN and PARAM
        Page 7
        Page 8
    Data and information flows in MAIN and PARAM continued
        Page 9
        Page 10
    Subroutine control in MAIN
        Page 11
    Predicting sugarcane yield (SCYIELD)
        Page 12 (MULTIPLE)
    Contents, arrays, and variables (CAV) in SCYIELD
        Page 13
    Operation of SCYIELD
        Page 14 (MULTIPLE)
    Estimation of the sugarcane ET equations
        Page 15
    SOILK for non-organic soils
        Page 16 (MULTIPLE)
    CROPK for citrus, pasture, golf courses turf
        Page 17 (MULTIPLE)
    Table 2: Crop coefficient (CROPK) values for citrus, pasture, golf courses, and turf, South Florida
        Page 18
    Soil water relations and data source
        Page 19
    Figure 3: Desorption moisture-tension relationships in Everglades peaty muck
        Page 20
    Table 3: Example calculations from desorption relations, 24-inch water table in organic muck soil, Belle Glade, Florida
        Page 21
    Figure 4: Water gain and release characteristics in Everglades peaty muck
        Page 22
    Table 4: Water storage capacity, or water released, from Everglades peaty muck for a receding water table
        Page 23
    Soil water relations and data source continued
        Page 24
    Empirical/statistical model
        Page 25 (MULTIPLE)
    Table 5: Definitions of constants, arrays, and variables in WTPRED
        Page 26
    Operation of WTPRED continued
        Page 27
    Characteristics and nature of ORGSOIL
        Page 28 (MULTIPLE)
    Table 6: Definitions of constants, arrays, and variables in ORGSOIL
        Page 29
    Table 7: Saturated water content of Everglades peaty muck soils at various depths
        Page 30
    Definition of a "day"
        Page 31
    ET estimates for each day
        Page 32 (MULTIPLE)
    Definition and removal of drainage water
        Page 33 (MULTIPLE)
    Generalized "pot" soil model (POTSOIL)
        Page 34 (MULTIPLE)
    Table 8: Definitions of constants, arrays, and variables in POTSOIL
        Page 35
    Defenition of a "day"
        Page 36 (MULTIPLE)
    End of "day" update of available soil water
        Page 37 (MULTIPLE)
    Table 9: Defintions of constants, arrays, and variables in PROFIT
        Page 38
        Page 39
    Water year vs. crop year calculations
        Page 40 (MULTIPLE)
    Constants, arrays, and variables (CAV) in TABWAT
        Page 41 (MULTIPLE)
    Table 10: Definitions of constants, arrays, and variables in TABWAT
        Page 42
    Table 11: Agricultural and other land cover in S-5A basin, South Florida Water Management District, 1979
        Page 43
    Table 12: Estimated growing season periods, season length, acreage, and turn-over coefficients for vegetable crops in the Everglades agricultural area, 1978-79
        Page 44
    Vegetable acreage production continued
        Page 45
    Table 13: Estimated vegetable production and proportion harvested each month in Everglades agricultural area, 1978-79
        Page 46
    Sugarcane acreage and productions
        Page 47
    Acreage of vegetable crops in planting date for validation of water balances in area S-5A, Everglades agricultural area, 1978-79
        Page 48
    Pasture and turf acreage
        Page 49 (MULTIPLE)
    Table 15: Goal water table depths in validation runs
        Page 50
    Water balance projections
        Page 51
    Figure 5: Historical and predicted irrigation in thousands of acre feet for actual water table depths in area S-5A
        Page 52 (MULTIPLE)
    Figure 7: Historical and predicted drainage in thousands of acre feet for actual water table depths in area S-5A
        Page 53 (MULTIPLE)
    Table 16: Actual and predicted irrigation and drainage requirements, S-5A area, South Florida, 1979-78
        Page 54
    Estimates of evapotranspiration
        Page 55
    Economic impact projections
        Page 56
    Table 17: Estimates of evapotranspiration (ET) for celery at various planting dates
        Page 57
    Table 18: Estimated and potential evapotranspiration (ET) for ratoon sugarcane
        Page 58
    Table 19: Crop acreage and return summary for S-5A in Everglades area for 1978-79
        Page 59
    Conclusion
        Page 60
    Bibliography
        Page 61
        Page 62
        Page 63
    Appendices
        Page 65
    Appendix A: Program flowcharts
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
    Appendix B: Fortran program code
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
    Appendix C: Statistical considerations in selecting the sugarcane yield model
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
    Appendix D: Statistical analysis of the sugarcane evapotranspiration relationship
        Page 175
        Page 176
        Page 177
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
    Appendix F: Data for model validation
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
Full Text





HISTORIC NOTE


The publications in this collection do
not reflect current scientific knowledge
or recommendations. These texts
represent the historic publishing
record of the Institute for Food and
Agricultural Sciences and should be
used only to trace the historic work of
the Institute and its staff. Current IFAS
research may be found on the
Electronic Data Information Source
(EDIS)

site maintained by the Florida
Cooperative Extension Service.






Copyright 2005, Board of Trustees, University
of Florida





3Dember co "19
December 1987


Area-Wide Agricultural Water Demand
Projection Model for South Florida:
Technical Documentation, Version 2.2


Gary D. Lynne, Philip J. d'Almada, Wayne C. Martin, and
Robert S. Mansell


















Agricultural Experiment Station
Institute of Food and Agricultural Sciences
University of Florida, Gainesville
J. M. Davidson, Dean for Research


Bulletin 849



















AREA-WIDE AGRICULTURAL WATER DEMAND PROJECTION MODEL FOR SOUTH
FLORIDA: TECHNICAL DOCUMENTATION, VERSION 2.2





G. D. Lynne, P. J. d'Almada, W. C. Martin, and R. S. Mansell





















Food and Resource Economics Department
Institute of Food and Agricultural Sciences
University of Florida
Gainesville, Florida 32611







TABLE OF CONTENTS


Page

Acknowledgements............................................. vii
Preface ...................................................... viii
Abstract................. .................................. ix
INTRODUCTION................................................ 1
Study Area ..................... ... ...................... 2
Overall Organization Of The Computer Program.............. 2
MAIN AND PARAM PROGRAM.............................. ......... 5

Constants, Arrays, and Variables (CAV) in MAIN and PARAM... 6
Data and Information Flows in MAIN and PARAM............... 6
Subroutine Control in MAIN.................................. 11
PREDICTING SUGARCANE YIELD (SCYIELD).......................... 12
Data, Model, and Results of Statistical Yield Estimation... 12
Constants, Arrays, and Variables (CAV) in SCYIELD.......... 13
Operation of SCYIBLD........................................ 14
DETERMINATION OF EVAPOTRANSPIRATION (ET) MODELS.............. 14

Estimation of the Sugarcane ET Equations................... 15
SOILK for Non-Organic Soils................................. 16
CROPK for Vegetable Crops................................... 16
S CROPK for Citrus, Pasture, Golf Courses and Turf........... 17
ORGANIC SOIL MODEL (ORGSOIL AND WTPRED)...................... 17
Estimation and Use of the Water Release Prediction
Equation................... ................................ 17

Soil Water Relations and Data Source.................... 19
Empirical/Statistical Model ............................ 25
Constants, Arrays, and Variables (CAV) in WTPRED......... 25
Operation of WTPRED.................................... 25

Characteristics and Nature of ORGSOIL ...................... 28

Coefficients, Arrays and Variables (CAV) in ORGSOIL..... 28
Initial Conditions..................................... 28
Definition of a "Day"................................... 31
ET Estimates for Each Day .............................. 32
Definition and Removal of Flood Water.................... 32
Definition and Removal of Drainage Water ................ 33
Addition of Irrigation Water............................ 33
End of Year Calculation.................................. 33
GENERALIZED "POT" SOIL MODEL (POTSOIL)........................... 34

Coefficients, Arrays, and Variables (CAV) in POTSOIL....... 34
Initial Conditions ........................................ 34
Definition of a "Day"...................................... 36
ET Estimates and Adjustment (SOILK and CROPK) Factors...... 36
System Capacity and Addition of Irrigation Water............ 36


iii

















Drainage, Percolation, Flood, and Runoff Water.............. 36
End of "Day" Update of Available Soil Water.................. 37
Values Provided by POTSOIL ................................ 37

CALCULATING PROFITS AND ECONOMIC IMPACT (PROFIT)............. 37
Constants, Arrays, and Variables (CAV) in PROFIT............. 37
Water Year vs. Crop Year Calculations...................... 40
Operation of PROFIT ....................................... 40
WATER BALANCE SUMMARIES (TABWAT)............................. 40
Constants, Arrays, and Variables (CAV) in TABWAT........... 41
Operation of TABWAT ........................................ 41
MODEL VALIDATION ............................................ 41
Vegetable Acreage and Production............................ 41
Sugarcane Acreage and Production............................ 47
Pasture and Turf Acreage ................................... 49
Choice of Water Management Strategies.......................... 49
Water Balance Projections.................................. 51
Estimates of Evapotranspiration............................ 56
Economic Impact Projections................................. 57
CONCLUSIONS AND RECOMMENDATIONS............................ 60
REFERENCES................................................... 61
APPENDICES.................................................. 65
A PROGRAM FLOWCHARTS..................................... 67
B FORTRAN PROGRAM CODE.................................... 105
Main Control Program................................... 105
Subroutine PARAM................. ...................... 125
Subroutine CALJUL ..................................... 128
Subroutine JULCAL ..................................... 130
Subroutine ORGSOIL .......... ....................... .. 132
Subroutine WTPRED........................................... 143
Subroutine POTSOIL.................................. 147
Subroutine TABWAT.................................... 154
Subroutine SCYIELD ..................................... 161
Subroutine PROFIT...................................... 162
C STATISTICAL CONSIDERATIONS IN SELECTING THE SUGARCANE
YIELD MODEL............................................ 167
D STATISTICAL ANALYSIS OF THE SUGARCANE
EVAPOTRANSPIRATION RELATIONSHIP ......................... 175
E CROPS, COSTS, YIELDS, PRICES, AND SEASON LENGTH IN
EACH WATER SUPPLY AREA................................. 179
F DATA FOR MODEL VALIDATION............................... 187


iv







LIST OF TABLES


Table Page

1 Definitions of constants, arrays, and variables
in MAIN and PARAM................................... 7
2 Crop coefficient (CROPK) values for citrus,
pasture, golf courses, and turf, South Florida....... 18
3 Example calculations from desorption relations,
24-inch water table in organic muck soil, Belle
Glade, Florida..................................... 21
4 Water storage capacity, or water released, from
Everglades peaty muck for a receding water table..... 23
5 Definitions of constants, arrays, and variables
in WTPRED........................................... 26
6 Definitions of constants, arrays, and variables
in ORGSOIL........................................ ..... 29
7 Saturated water content of Everglades peaty muck
soils at various depths............................. 30
8 Definitions of constants, arrays, and variables
in POTSOIL......................................... 35
9 Definitions of constants, arrays, and variables
in PROFIT.......................................... 38
10 Definitions of constants, arrays, and variables
in TABWAT ........................................... 42
11 Agricultural and other land cover in S-5A Basin,
South Florida Water Management District, 1979........ 43
12 Estimated growing season periods, season lengths,
acreage, and turn-over coefficients for vegetable
crops in the Everglades Agricultural Area, 1978-79... 44
13 Estimated vegetable production and proportion
harvested each month in Everglades Agricultural
Area, 1978-79...................................... 46
14 Acreage of vegetable crops by planting date for vali-
dation of water balances in area S-5A, Everglades
Agricultural Area, 1978-79........................... 47
15 Goal water table depths in validation runs........... 50
16 Actual and predicted irrigation and drainage
requirements, S-5A area, South Florida, 1978-79...... 54
17 Estimates of evapotranspiration (ET) for celery at
various planting dates............................... 57
18 Estimated and potential evapotranspiration (ET) for
ratoon sugarcane.................................. 58
19 Crop acreage and return summary for S-5A in
Everglades area for 1978-79......................... 59


v

















LIST OF APPENDIX TABLES


Table Page

C.1 ANOVA for testing sensitivities of sampling errors
sugarcane experimental data, Belle Glade, Florida..... 168
C.2 ANOVA with pooled experimental and sampling errors,
testing split-split plot effects, sugarcane experi-
mental data, Belle Glade, Florida...................... 168
C.3 ANOVA showing tests for linear and quadratic water-
table main effects and interactions with variety,
plant cane, Belle Glade, Florida...................... 169
C.4 ANOVA showing tests for linear and quadratic water-
table main effects and interactions with variety,
ratoon cane, Belle Glade, Florida..................... 169
C.5 Coefficients and calculated t-statistics, plant and
ratoon sugarcane, Belle Glade, Florida, 1977-79....... 170
D.1 Statistical estimates of evapotranspiration to water
table relations, Belle Glade, Florida................. 176
E.1 Crops, costs, prices, yields, and growing season
length in Caloosahatchee River water supply area...... 179
E.2 Crops, costs, prices, yields, and growing season
length in Rim Basins water supply area................. 180
E.3 Crops, costs, prices, yields, and growing season
length in St. Lucie Canal water supply area........... 181
E.4 Crops, costs, prices, yields, and growing season
length in Everglades water supply area................. 182
E.5 Crops, costs, prices, yields, and growing season
length in Lower East Coast 1 water supply area........ 183
E.6 Crops, costs, prices, yields, and growing season
length in Lower East Coast 2 water supply area........ 184
E.7 Crops, costs, prices, yields, and growing season
length in Lower East Coast 3 water supply area........ 185


vi















LIST OF FIGURES


Figure Page

1 Water supply areas within the study area............... 3

2 Structure and information flow of FORTRAN
program Version 2.2.................................. 4

3 Desorption moisture-tension relationships
in Everglades peaty muck .............................. 20

4 Water gain and release characteristics in
Everglades peaty muck ................................. 22

5 Historical and predicted irrigation in
thousands of acre feet for actual water table
depths in area S-5A ................................... 52

6 Historical and predicted irrigation in
thousands of acre feet for recommended water
table depths in S-5A.................................. 52

7 Historical and predicted drainage in thousands
of acre feet for actual water table depths in
area S-5A............................................ 53

8 Historical and predicted drainage in thousands
of acre feet for recommended water table depths
in area S-5A......................................... 53

A.1 MAIN Flowchart........................................ 67

A.2 PARAM Flowchart...................................... 75

A.3 CALJUL Flowchart..................................... 77

A.4 JULCAL Flowchart...................................... 79

A.5 WTPRED Flowchart...................................... 80

A.6 ORGSOIL Flowchart..................................... 83

A.7 POTSOIL Flowchart.................................... 94

A.8 SCYIELD Flowchart.................................... 98

A.9 PROFIT Flowchart..................................... 99

A.10 TABWAT Flowchart...................................... 102

D.1 Evapotranspiration (ET) by month at varying
water table (WT) depths for plant cane at
Belle Glade, Florida................................. 177

D.2 Evapotranspiration (ET) by month at varying
water table (WT) depths for ratoon cane at
Belle Glade, Florida................................. 177


vii













PREFACE


Version 2.2 of the AGRICULTURAL WATER DEMAND PROJECTION MODEL
FOR SOUTH FLORIDA represents a "first generation" attempt at such
projection for larger areas. The user is advised to proceed with
caution in its application, until further validation and testing
can be completed. In fact, the user is encouraged to contact the
authors with comments/suggestions for improvements. It is
intended the structure of this model be built upon and improved as
better information becomes available. This is not the last ver-
sion, but rather, the start of several being made available
through time.









I
ACKNOWLEDGEMENTS


The funds for this modeling effort were provided in part
under contract 4-FCD-22 8002-302 with the South Florida Water
Management District, West Palm Beach, Florida. Completion of the
model also relied heavily on resources of the Institute of Food
and Agricultural Sciences, and many faculty in that organization,
including Professors J. Bennett (Agronomy), D. Harrison (Ag.
Eng.), J. Jones (Ag. Eng.), R. Koo (Horticulture, Lake Alfred), K.
Shuler (Extension, Belle Glade), S.P. Kovach (Bradenton), and
others. This does not mean the model is without error. This
model does reflect the interpretation by the authors of how these
complex relations operate, and all others are absolved from any
responsibility.









I


viii

















ABSTRACT


An agricultural water demand projection capability is needed
during water shortage periods. The computer simulation model
documented herein operates at the field level on a daily time
step, aggregates water demands over larger regions for months, and
generates totals for an entire year. A financial impact projec-
tion is also provided, for any mix of planting dates, irrigation/-
drainage strategies, price and cost conditions, and rainfall
quantity. Daily rainfall data for a water year of August 1 July
31 "drives" the model. The water balance is simulated in the root
zone for organic soils and a general soil/water model is used for
all other soil types. Version 2.2 was designed for the seven
water supply areas south of Lake Okeechobee, Florida, encompassing
a large vegetable and sugarcane producing area.


Keywords: Water demand, agricultural water use, water use projec-
tions


ix







AREA-WIDE AGRICULTURAL WATER DEMAND PROJECTION MODEL FOR SOUTH
FLORIDA: TECHNICAL DOCUMENTATION, VERSION 2.2

G. D. Lynne, P. J. d'Almada, W. C. Martin, and R. S. Mansell


INTRODUCTION

The Florida water management districts are charged with the
efficient and equitable allocation of limited water supplies
during water shortage periods. Shortages may result through
natural drought occurrences, in which case the problem of water
supply is generally of a short-run nature. There may also be
"man-caused" shortages, however, when the quantities demanded
exceed the available water in some supply area, even over periods
of plentiful rainfall. Under both situations, difficult decisions
must be made concerning which types of economic activity to be
reduced and to be supported. Some means of predicting the impact
of various types of water allocation strategies in these situa-
tions can provide useful information to both the district and to
those affected by water allocation and reallocation decisions.

The purpose of the water demand modeling effort described
herein was to provide a means for generating information on water
demand in the basic, "on-farm" agricultural sector. This comput-
erized model was organized to provide estimates of future water
demands over periods of less than one year in South Florida. One
very real concern in Florida water allocation is the expected
* impact of water shortages within crop seasons. The user of the
model can select any length of crop season, and make water use
projections for that season and for the entire water year for any
tract of land. In addition, with specification of all the various
crop types, planting and harvesting dates, acreages, and location
of production, the user can predict the water demand over larger
water supply areas. Also, it is possible to obtain a financial
prediction of the expected profit associated with each level and
timing of plantings for both tracts and areas.

More detail on the capabilities of the model are presented in
the User's Manual (Lynne, 1984). It is the purpose of this bulle-
tin to describe and document the logic entering the modeling
effort of Version 2.2, completed in December 1983. This includes
a discussion of the sources and integration of information, the
computer algorithms, and the framework built into the model.
Further documentation of the actual computer code is also provided
for those cases where it was felt necessary to elaborate over and
above the text included in the actual FORTRAN programs. Flow



G.D. LYNNE is associate professor, P.J. D'ALMADA is research
assistant, and W.C. MARTIN is former research assistant, all in
the Food and Resource Economics Department; R.S. MANSELL is
* professor in the Soil Science Department. All are in the
Institute of Food and Agricultural Sciences, University of
Florida.


1








charts are provided in Appendix A, and the computer code is in
Appendix B, for reference during study of this documentation. It
will also be helpful if the reader is generally familiar with the
material of the User's Manual, which will aid in placing this more
technical discussion in context.


Study Area

The focus of the model is the water management area south of
Lake Okeechobee, within the boundaries of the South Florida Water
Management District (Figure 1). There are seven water supply
areas (WSA's) within the larger area, namely

1. Caloosahatchee River
2. Rim Basins
3. St. Lucie Canal
4. Everglades Agricultural Area
5. Lower East Coast 1
6. Lower East Coast 2
7. Lower East Coast 3

The subbasin S-5A within the Everglades Agricultural Area is also
identified because the land use in that area in 1979 was employed
in the validation of the model.
This is a diverse agricultural area, with at least three
major soil types, including the organic, flatwoods, and rock
soils. The major crop in an acreage sense is sugarcane, repre-
senting nearly 350,000 of the 2.2 million acres (U.S. Bureau of
the Census, 1978) devoted to agriculture in the area. All of the
sugarcane and nearly all of the vegetables are irrigated, making
them major water users. Over 28 different vegetable crops can be
grown, although not all are grown in each of the WSA's. Pasture,
turf, and golf courses also use significant amounts of water. The
pasture enterprises are mainly of the beef cow-calf variety,
although there are some dairy enterprises as well. There is also
citrus in the area, including the limes and lemons of Dade
county. In addition, there are many specialty crops, including
many "Caribbean" vegetables grown in Dade county, as well as the
avocado and mangoes produced there.


Overall Organization Of the Computer Program

The computer model is composed of a control program and nine
subroutines, as illustrated in Figure 2. The FORTRAN programs
are:



IThe model is quite general in its applicability, providing a
structure useful wherever water is a significant input into agri-
cultural processes. However, the model has been specifically
designed for the study area outlined in Figure 1, and will require
some modification for use in other areas. The interested reader
should contact the authors for further detail.


2





















Lucie Canal


Lower East
Coast 1







Lower East
Coast 2







Lower East
Coast 3


Figure 1. Water supply areas within the study area.







































PARAM ORGSOIL POTSOIL TABWAT H SCYIELD PROFIT





WTPRED Tables Tables
1,2,3 4,5




Figure 2. Structure and information flow of FORTRAN program,
Version 2.2.










4


4








MAIN the control program, which directs action and the flow
of information throughout the entire simulator;
PARAM used to read input files containing some of the param-
Seters necessary to the operation of the model;
CALJUL calculates the Julian date, given input of the calendar
date;
JULCAL calculates the calendar date, given the Julian date;
ORGSOIL organic muck soil water balance program;
WTPRED water table prediction program for the organic muck
soils;
POTSOIL soil water balance program for all other soil types;
SCYIELD sugarcane yield prediction program;
PROFIT calculates costs and net returns for each crop in the
simulation, and generates the cost/return summary
tables;
TABWAT tabulates the water balance for each tract and for the
area, and generates the water summary tables.

The overall program is modular in form. This is a convenient
and useful programming style in that it allows for maximum flexi-
bility in program operation and for changes in the program. The
display in Figure 2 shows this arrangement. The user can operate
in either interactive or "batch" mode. If the latter, the input
data are all accepted from an input file, and MAIN proceeds to
control the simulation. If not, the user is prompted to provide
information on the items listed in Figure 2. Some of these data
are read from files, such as the crop type in each water supply
area from WATCOS.DAT, and made available through PARAM for inspec-
tion and change by the user.

Once the data are provided to MAIN, the planting, harvest,
and water year calendar dates are converted to Julian days, using
CALJUL. There is also some need to convert from Julian days to
calendar dates, which is accomplished by JULCAL. MAIN then calls
either the organic soil model (ORGSOIL) or the model provided for
all other soil types (POTSOIL). ORGSOIL uses the water table
prediction capabilities of WTPRED to establish the depth to water
table on any given day. The soil models provide the daily water
balance calculations, which are transferred to TABWAT for conver-
sion to monthly estimates, and yearly summaries for the water year
SAugust 1 July 31. The computer generated Tables 1, 2, 3 (See
User's Manual) come from TABWAT. If sugar cane is being analyzed,
control is next transferred to SCYIELD to estimate yield as asso-
ciated with a particular water table on the organic soils. PROFIT
is then called by MAIN to estimate the water costs, for drainage
and irrigation, and the net profits for each harvest month and the
year. The computer generated Tables 4 and 5 (See User's Manual)
come from PROFIT.


MAIN AND PARAM PROGRAMS

MAIN and PARAM serve to control the other programs and read
various user-supplied files and/or collect other information
interactively while a projection is being made, as described in
detail in the User's Manual (Lynne, 1984). In fact, nearly all of


5








MAIN is used to set various default values, and facilitate
interactive changes by the user, or to read a user provided data
file in the "batch" mode. The first task assigned to MAIN is to
prompt the user as to whether this particular run is to be done
interactively, or is to be based on a data file and operated in
batch mode. If the latter, data must be provided on crop type,
acreage, water management strategy, and other parameters
(COEFPARA.DAT, format described in the User's Manual). MAIN reads
in a value for INTER, which is used as a "switch" throughout the
program to insure the correct values are inserted where needed.
All of the steps discussed in the following are performed by MAIN
in either mode.


Constants, Arrays, and Variables (CAV) in MAIN and PARAM

The CAV names utilized in MAIN and PARAM (meaning "param-
eter") are illustrated in Table 1. Many of these are initialized
in MAIN/PARAM and made available to other subroutines. The reader
can follow the information flow discussed in the next section by
referencing Figures A.1 and A.2 in Appendix A or the actual com-
puter code in Appendix B.


Data and Information Flows in MAIN and PARAM

Other than the determination of operating mode, the first
major task in MAIN is to accept the month and area namesets (into
AMONTH and AREA) and to convert the January to December rainfall
year into the water year. The water year is defined from August 1
to July 31, which was chosen because the agricultural crop season
starts, in the study area, in mid-August. Also, water allocation
decisions often have to be made in the early fall months for the
balance of the year because the water supply is well-defined by
early fall. That is, the "wet season" ends by mid-October, mean-
ing the water available in the system (lakes, canals, aquifers) at
that time must usually form the upper limit. Little rainfall is
(generally) received during the main irrigation season.

The user must also select the rainfall year to be simu-
lated. It is presumed the user has already decided what type of
rainfall year is expected to occur, based on whatever criteria are
desired. There is no "algorithm," or any programmed means in the
model for selection of the type of rainfall year. Version 2.2
(V2.2) allows the user to select a year from the 55-year record
available from the Belle Glade weather station, located at the
Agricultural Research and Education Center. These data were
available through the Northeast Regional Data Center (NERDC) at
the University of Florida, through the Hydrologic Information
Storage and Retrieval System (HISARS). The data were "cleaned"
and average daily values were inserted for missing data days,
giving a continuous record over the 1925-1979 period. The com-
puter file used in V2.2 and containing these data is BGRAIN; it
can be replaced with other data, or expanded and updated as
needed, as described in the User's Manual.


6








Table 1. Definitions of constants, arrays, and variables in MAIN
and PARAM.

In MAIN
AMONTH NAME(12) Name of month, starting with August
RAIN(60,365) Daily rainfall, starting January 1 December
31, for each of up to 60 years, read from a user
supplied file, inches
INTER "Switch," causing "batch" operation if INTER =
2, or interactive operation if INTER = 1
AREA NAME Name of area
MTABLE "Switch," yielding daily water use estimates only
if = 1 monthly only if = 2, and both if = 3
NYRSTA Year simulation is to start; e.g., 1978 means a
water year of August 1, 1978 through July 31,
1979.
RAINWY(365) Daily rainfall, starting August 1 July 31,
inches
NRAIN "Switch," listing the daily rainfall values if =
1
ENAME(19,7) Crop name, for up to 19 different crops and 7
water supply areas, names provided by the user
in the file WATCOS
ECON(19,7,7) Costs, price, average yield, and growing season
length for 19 crops, 7 different values, and 7
areas, values provided by the user in the file
WATCOS
CSTRET(12,8) Cost and return summary array for an area, for
each of 12 months and 8 different items
WATSUM(12,5) Water balance summary array for an area, for
each of 12 months and 5 different water measures
FCAP Field capacity of non-organic soils, in inches
of available water
PFCAP Proportion of available water at field capacity
in non-organic soils at which an irrigation is
to be started
RATE Depth of application per irrigation in non-
organic soils, inches
MIRRD Minimum number of days between irrigations, in
non-organic soils
GWT Goal water table, organic muck soils, in inches
to water
NTRACT The identification number for the tract, an
integer 1 to 9999
WCI Water costs for all irrigation taking place
before August 1, thousands of dollars
WCD Water costs for all drainage taking place before
August 1, thousands of dollars
MSUM "Switch," set to 1 will result in saving the
CSTRET() and WATSUM() arrays on file
WATUSE.RES File name to which the computer generated Tables
1 5 and all other computer generated output is
* written
MCROP Number of crop selected for the area


7








IDAYSTART

MONSTART
IDAYHARV

MHARV

IDAYPL,IDAYHA

SEASON
ISTARTDAY

IENDDAY

MOS, MOE

IDAYS, IDAYE

TAC
PROPH

ACRES(12)
MSOIL

EFFIRR
ORGPCD

ORGPCI

TDDRY, TIDRY


TDWET, TIWET


DEPTHS
WRULE(12)

WRATE(12)


MY



Y

PRICE
AVCWI

YIELD

SEASON


- Planting date, a calendar day, like 10 for
September 10
- Planting date, month, August 1, ..., July = 12
- Harvest date, calendar day, like 10 for
September 10
- Month of crop harvest, August = 1, ...,
July = 12
- Planting and harvest dates, in Julian days,
August 1 = 1, ..., July 31 = 365
- Length of growing season, in days
- Day to start the water simulation period,
calendar day
- Day to end the water simulation period, calendar
day
- Month of start and end of water season, August =
1, ..., July = 12
- Day to start and end the water season, in Julian
days
- Total acres in tract of concern
- Proportion of acreage planted that is actually
harvested
- Acreage in each month, acres
- Soil type, if = 1 have organic soil, if = 2,
other soil
- Efficiency of irrigation, in percentage terms
- Pumping capacity for drainage of organic soils,
inches/day
- Pumping capacity for irrigation of organic
soils, inches/day
- "Trigger" for start of drainage and irrigation
during dry period of year, organic soils, inches
distance from GWT
- "Trigger" for start of drainage and irrigation
during wet period of year, organic soils, inches
distance from GWT
- Depth of organic muck soil to rock, inches
- Water rule for irrigation of other soils, for
each month, inches of water
- Depth of irrigation application, for other
soils, inches per irrigation
In PARAM
- Type of sugarcane yield model, 1 = average
yields; 2 = plant cane (PC), variety 3(V3),
CL54-378; 3 = PC, V4 (CL54-336); 4 = PC, V5
(CP65-357); 5 = ratoon cane, V5 (CP65-357)
- "Switch," used to affect ET calculated for
sugarcane; 1 = ratoon cane, 0 = plant cane
- Price per unit of crop harvested
- Average variable cost of water applied for
irrigation dollars/acre inch
- Yield/acre, measured in units typically used for
each crop (tons, crates, cwt, etc.)
- Growing season length in days


8


C


4








The next major task performed by MAIN is to read the data
file WATCOS (Appendix B), into an array called ENAME(19,7), which
includes the crop names, and an array called ECON(19,7,7), which
S contains cost, price, and growing season information for each crop
in each area. The crop types included in this file, and thus
available in V2.2, were selected on the basis of a review of
various sources. The "Vegetable Summary," "Field Crop Summary,"
"Citrus Summary," and "Livestock Summary" (Florida Crop and Live-
stock Reporting Service, 1979, 1980) were all consulted. Also,
soil surveys for each of the respective counties were reviewed,
for soil suitability for the respective crops, as well as publica-
tions specific to commodity types (Marlowe, 1982). Crop budgets,
average yields and prices were derived from Brooke (1980); Brooke
and Hooks (1980); Lopez et al. (1979); Abbitt and Muraro (1979);
Anderson and Hipp (1974); and Walker (1973). Some of the budgets
were not available at the time of completion of V2.2; thus, WATCOS
contains several "zeroed out" rows. Also, budgets for the latest
crop year will need to be provided in practice (see User's
Manual).

Crop season length, measured in days, was derived from basic
information provided by extension personnel at Bradenton and from
Marlowe (1982). V2.2 has provision for supplying only one growing
season length for each crop in each area. This may be a problem
in some areas, as the same crop (e.g., corn) is grown both in the
fall and spring periods, and sometimes throughout the winter
period as well (e.g., radishes). The growing season length will
generally be longer during the cooler periods, and appropriate
S adjustments will need to be made. The default value is in the
"middle," reflecting an average.

Each area can have as many as 19 crops specified, based on
array sizes in V2.2. The user must provide a 19 by 7 array for
each of the 7 possible areas. The User's Manual describes the
elements in the array (also, see Appendix B). The user can either
change the elements of ECON(I,J,K) by changes to the file WATCOS,
or can make "interactive changes" when PARAM prompts the user, as
described in the User's Manual.

The default values provided in the code were selected based
S on various sources. The FCAP value of 4 inches reflects a typical
field capacity in the top 4 feet of deep, well-drained sands, such
as Lakeland fine sand (Reitz, 1977). Permanent wilting point was
selected at PWP = 1.0 inch, based on citrus capability to remove
water from Lakeland fine sand. This suggests that 3 inches of





2Personal communication with S. P. Kovach, Agricultural
Research and Education Center, Bradenton, Florida.

S3Personal communication with R. C. J. Koo, Lake Alfred
Agricultural Research and Education Center.


9








water are assumed available for plant use. For other crops and
soils in the study areas, this value will have to be changed
accordingly. The PFCAP value of 60 percent was found to be the
most profitable choice in a study of optimal irrigation strategies
for citrus irrigation (Anaman, 1981) and around 70 percent in
soybean production (Boggess and Lynne, 1983). That is, allowing
some soil storage for rainfall, such that more of the rainfall is
effective, appears to pay under Florida conditions. This also
implies that some plant stress may also be economic. The default
value of 0.60 can be changed in any given application of the
model, but empirical evidence suggests the range of 0.60 to 0.70
may be a reasonable place to start for most crops. However, if
something less than 1.0 is used, the related yield response must
also be known.

Irrigation system capacity is represented by either RATE
(depth of application, in acre-inches per day) and MIRRD (minimum
number of days before the system can be used to irrigate again)
for sprinkler systems or GWT (goal water table), ORGPCD (pumping
capacity for drainage), and ORGPCI (pumping capacity for irriga-
tion) for the water-table maintenance soils. The depth of appli-
cation at 1.0 inch per day reflects a typical design value for
many sprinkler systems in the citrus area as well as for some
vegetable crops. Also, it is often decided that the system should
be able to irrigate every tract on any given farm within MIRRD = 7
days. Stated somewhat differently, these two values translate
into a design criterion of 1 inch every week for a farm. That is,
it is assumed an inch can be applied to each acre on the farm
every seven days. These are the default values for all the
systems except the ground-water maintenance method used for the
organic muck soils in the Everglades Agricultural Area.

The rule of GWT at 24 inches will have to be changed for each
crop, running as high as 12 inches for some vegetables, and as low
as 36 inches during some parts of the year for sugarcane and
pasture. The 24-inch rule is offered as a starting point, based
on recommendations of Snyder et al. (1978). These recommendations
reflect a judgment based on many factors, including an appropriate
level of soil subsidence. The pumping capacities are assumed to
vary between that for drainage (ORGPCD = 1.5 inches/day) and
irrigation (ORGPCI = 0.4 inches/day), based on judgments of exten-
sion personnel knowledgeable of conditions in the area.

Irrigation efficiency will vary considerably, as across
system types. The default values are set at EFFIRR = 1.00 for the
organic muck, water-table maintenance soils, and EFFIRR = 0.75 for
all other types. It is fully recognized that EFFIRR is usually
less than 1; these can be changed by the user as the model is
applied. The value of one is but a starting point, based on the
assumption that little water escapes from the canals.




4
Personal communication with Dalton Harrison, Extension
Agricultural Engineer, IFAS, University of Florida.


10








The values of FCAP, PFCAP, and RATE are used in MAIN to
establish the irrigation rule, WRULE(I), and application rate,
WRATE(I), for each of the I = 1, 2, ..., 12 months. V2.2 sets
S these decision rules equal across all months of the water year.

The potential evapotranspiration (ET) for all crops was
derived from the publication by Jones, et al. ( 1984 ). The
monthly estimates are provided to the program in PARAM; the aver-
age ET per day is then calculated. These values represent the
amount of water that will be removed by evapotranspiration "of an
extended surface of an 8- to 15- cm tall, green grass cover,
actively growing, completely shading the ground and not short of
water" (Doorenboos and Kassam, 1979, p. 15). The values provided
through PARAM are for the Belle Glade weather station (Jones, et
al., 1984 ), and are expected to be representative of the
potential ET values for the entire study region.


Subroutine Control in MAIN

MAIN calls several subroutines. The first of these are the
CALJUL and JULCAL routines, which are used to determine the Julian
days from calendar dates provided by the user (CALJUL) or vice
versa (JULCAL). The calendar month is always specified as August
= 1,..., July = 12, and the Julian day count starts on August 1.

After the various date and other default data are organized
and/or changed, MAIN transfers control to one of the soil water
subroutines. This may be either ORGSOIL (and WTPRED) to calculate
the soil water balance relations on a daily time step for an
organic muck soil, or to POTSOIL for all other soil types. The
net result is a set of daily estimates of the water balance for
one acre of the crop selected.

The daily data on rainfall, irrigation, drainage, and evapo-
transpiration are then used in TABWAT, which is the next sub-
routine called by MAIN, to generate monthly and yearly water
summaries for the tract and area. That is, the aggregation from
the per-acre estimates takes place in TABWAT. Further, TABWAT
provides a listing of the parameter values in a table, identified
) as "Table 1" by the computer, followed by a tract-level water
balance summary in "Table 2," and an area summary of the overall
water balance in "Table 3." Examples of these are provided in the
User's Manual.

If the crop is sugarcane, MAIN next calls SCYIELD to cal-
culate yield, based on the average water-table depth calculated in
the ORGSOIL subroutine. If the current simulation run is for some
other crop, the corresponding average yield made available through
PARAM (from the file WATCOS, or as supplied by the user interac-
tively) is used. MAIN then makes that yield value available to
PROFIT, where the net return associated with each of the
enterprises can be determined. PROFIT provides a financial sum-
mary in computer-generated "Table 4" and "Table 5." If in inter-
active mode, the computer then prompts the user regarding whether
further runs are to be made and, if so, the process starts over.


11








PREDICTING SUGARCANE YIELD (SCYIELD)


Statistical yield models were developed for sugarcane. This
was in contrast to the other crops where only average yields can
be used in V2.2. Also, the yield models developed for sugarcane
are limited in that the experiment extended for only two years.
Strictly speaking, the yield models discussed below can be used to
project yield only for the following conditions:

cane planted in late spring (March), and harvested in
December;
first-year ratoon5 cane, growing from January to harvest
in December;
only certain varieties, as detailed below.

As a result, the program is also configured to allow the user to
insert an average yield appropriate to the particular case being
examined, such as for third- or fourth-year ratoons.


Data, Model, and Results of Statistical Yield Estimation6

Lysimeter data were available on plant and ratoon cane, in
the Everglades Agricultural Area (EAA) for three water-table
levels and two years (Shih and Gascho, 1980). The experimental
design was a 3 x 5 x 2 split-split plot arrangement of treatments
in a completely randomized design. Water table (W) was on the
main plot in three levels. Sugarcane variety (V), in five levels,
formed the split plot, while plant cane and first ratoon cane
crops were the two levels of the year (Y) split-split-plot treat-
ment factor. Two lysimeters per level of water table accounted
for the complete randomness of the design, while three replicates
(R) of variety per lysimeter provided a test of variation between
lysimeters within water-table level.

The split-split plot model with sampling error accounted for
90.88 percent of the variation. Since no significant sampling
variation was detected, sampling and random errors were pooled at
each level of the model. A significant second-order interaction
(YVW) was detected (P < 0.05). Observing a highly significant
year (Y) main effect (P < 0.01), it was decided to examine the
main and split plot effects at each level of year (i.e., for plant
cane and ratoon cane). The main effect of water table (W) and the
variety-water table interaction (VW) were separated into their
linear and quadratic components.





5"Plant" cane refers to the first year the cane is
harvested. Each subsequent year (cane is perennial) is referred
to as "ratoon."

A more detailed discussion of the statistical analysis is
provided in Appendix D.


12







The hypothesized model, for each year (or cane crop), was
represented as:

SY = 0 + B1W + 2 + 3V + 4V2+ 85V3 + 86V4

+8w v + sWv2 + BWV3 + 0W v + BWVi (I)
+7 i 1 8 i2 9 i3 10 4 1 q 1 (1)

+ 12 q 2 + 13 q 3 + 14qV4 +

where Y = biomass, in pounds per stool, converted to tons by
multiplying by 2 (based on 4000 stools per acre and
2000 pounds per ton);
W =30, 60, 90 cm to the free water table, the linear
component of water table;
Wq = 302, 602, 902, for the quadratic component of the
water table;
Vi = varieties, 1=CP63-588; 2=CP56-59; 3=CL54-378; 4=CL54-
336; 5=CP65-357. Variety 5 effects will be repre-
sented in the 80 parameter.

For each variety, this model simplifies to:

Y = -O + aI + a2Wq + (2)

S where ai = Bi + j+41, i = 0, 1, 2; j = 3, 4, 5, 6. (3)

(For variety 5, 8j+4i = 0 for all i, j in Equation 1. Hence, a =
Bi, in Equation 3.)

The coefficients, standard errors, and calculated t-statis-
tics are shown in Appendix C, Table C.5. The linear water-table
variable was shown statistically significant at the 0.05 level for
varieties V3, V4, and V5 for plant cane, and V5 for ratoon cane.
The quadratic effects were significant for only two varieties.


Constants. Arrays, and Variables (CAV) in SCYIELD

SCYIELD has very few CAV, as follows:

MY "Switch," defining the model selected by the user. The
choices given the user are:
1. MY = 1, for average yields entered by the user;
2. MY = 2, for plant cane, variety CL54-378;
3. MY = 3, for plant cane, variety CL54-336;
4. MY 4, for plant cane, variety CP65-357;
5. MY = 5, for ratoon cane, first year, variety CP65-357.
YIELD Yield in tons per acre.
S CALGWT Calculated goal water table, the average water table the
ORGSOIL program was able to maintain for the season, in
inches.


13








Operation of SCYIELD


SCYIELD uses the regression equations from Appendix C, Table
C.5 to predict the yield. The value of CALGWT is brought, via
COMMON, to SCYIELD from ORGSOIL for this purpose. If the user has
specified average yields, SCYIELD simply returns control to MAIN
immediately, and the average yield supplied by the user is util-
ized. The flow chart is in Appendix A, Figure A.8.

Yield is set to zero if the GWT averages less than 4
inches. This facilitates examining fallow conditions in the cane
fields.


DETERMINATION OF EVAPOTRANSPIRATION (ET) MODELS

There are three general factors that affect evapotranspira-
tion (ET), broadly categorized into the plant, atmospheric and
soil water conditions. Radiation is the most critical. In addi-
tion, temperature, wind, and humidity all affect the plant condi-
tion and the amount of water in demand. The amount of water in
the soil, available for evapotranspiration by the plant, also
helps determine the upper limit on how much water can be used by
the plant on any given day. Jones, et al. ( 1984 ) provide a
succinct discussion of the nature of ET; the reader is referred to
that publication for further detail.

As a result of these three factors, some means was sought to
integrate across all the atmospheric conditions separately from 1
the soil water effects. This was accomplished in accordance with
the crop and soil type. The general procedure was to calculate a
CROPK and/or SOILK, both of which are the ratio of the actual to
the potential ET (AET/PET). The former serves to integrate across
the atmospheric conditions, while the latter proxies the impact of
soil water conditions for a given crop. Thus, when these factors
are multiplied by the adjusted Penman potential evapotranspiration
ET, the result is an estimate of the actual ET (See Jones, et al.,
1984, for a definition of the Penman approach and adjustment pro-
cedure).

For the organic soil, no SOILK is calculated, at least not
directly. That is, for sugar cane the SOILK is implicit in the ET
equation estimated from the Shih and Gascho experiments, and cane
is allowed only on the organic soils. For the vegetable crops, it
was assumed that growers would either choose the "no stress" water
strategy, or not grow vegetables. Thus, the SOILK is unity. For
pasture and turf, the SOILK was also assumed equal to 1, which may
lead to over-estimation of ET in these crops. That is, the 0.6 to
0.70 guideline cannot be used unless there is also a yield
response estimate associated with that level of the factor. This
was necessary in V2.2 because of the lack of data on yield
response for these crops. This will be especially the case for
pasture, as water tables are allowed to decline (irrigation is
reduced) and economic conditions in the livestock industry fluc-
tuate.


14







Estimation of the Sugarcane ET Equations

As noted, sugar cane was handled differently from other
S crops. The integration across atmospheric and soil water copdi-
tions was accomplished by an ordinary least squares regression of
the actual ET as a function of the depth to water table during two
years of an experiment at the Belle Glade experimental station
(Shih and Gascho, 1980). There was no attempt to separate out the
effects of each different influence.

It was reasoned that the ET would be related to the depth of
the maintained water table. This response was expected to be
curvilinear in nature, as plant growth is first increased as the
water table is lowered. It is expected there is a decreased crop
response, and a related decrease in the water demanded, as the
water table (WT) is lowered further. Thus, it is expected ET
would first increase, and then decrease, as the water table is
lowered from the surface. Also, it was anticipated that plant
cane would use more water than ratoon cane because of the very
rapid and lush growth of newly planted cane. Thus, the following
model form was hypothesized

AET = B0 + WT + 2WT2+ B3Y (4)

where 6I = the population parameters;
WT = depth to water table, in inches;
Y = a zero/one "dummy" shifter, reflecting a value of Y =
0 for plant cane, and Y = 1 for ratoon cane.

The results are presented in Appendix D, Table D.1

All the R2 values were high, except for June and December,
indicating reasonable "fits" to the data for each month. However,
the response during any given month was varied with respect to the
WT variable. In some months, the ET was shown to increase at an
increasing rate as the WT was lowered (January, March, April,
September, December). For others, the ET increased at a decreasing
rate, and then declined (May, June, July, August, October), as was
hypothesized. For still others, the response gave ET declining as
the WT deepened, at an increasing rate (February, November).
However, in the majority of the cases, the t-statistics for the
non-linear WT components were not significant even at the 10
percent level. The general hypothesis that ET will first
increase, reach some peak, and decline must be rejected for most
months, while it does hold in a few periods. The only apparently
defensible conclusion appropriate to all the months is that ET
generally increased as the WT was lowered toward 24 inches.

Plant cane is expected to use more water than ratoon cane.
This would be predicted with a negative coefficient on the Y



This estimation process is discussed in more detail in
Appendix D.


15







variable. Of course, these results must be tempered by the nature
of the experiment. There was a killing freeze in mid-January of
the year when the cane was first planted. The cane was replanted
in March, which probably explains why the "dummy" variable has a
positive sign in the earlier period, and a negative sign as the
season progresses.


SOILK for Non-organic Soils

Soil water conditions were included by calculating a rela-
tionship between the ratio of AET/PET and the soil moisture on any
given day. In particular, a linear relationship of the following
form was used

SOILK = BSLOPE SM (5)

where BSLOPE is the slope and SM = available soil water. This
function is used for 0 level of SM at which plant stress starts. TSTRES is calculated as
STRESS AVWAT, where STRESS is the percentage of the water
holding capacity of the soil (AVWAT) at which stress starts. The
SOILK values are calculated by using the amount of water in the
soil at field capacity and at permanent wilting point. For the
default case of FCAP = 4.0, PWP = 1.0, and STRESS = 0.70, the
function becomes

SOILK 0.3333 SM

which reaches a value of 1.0 at SM = 3.0 inches of available water
(FCAP PWP). SOILK is assumed to stay at 1.0 for all values of
SM > TSTRES

The nature of the SOILK function will vary depending on the
crop and soil conditions. The choice of a linear function is
justified only by its simplicity, and the lack of specific infor-
mation on the appropriate functional forms for the crops consid-
ered in the model.


CROPK for Vegetable Crops

Version 2.2 uses a function representing the impact of the
time distance through the crop season on the CROPK for vegetable
crops. The basic functional relation was derived from basic data
on tomatoes, as represented by Jones, et al. ( 1984 ). To the
extent that other vegetable crops have different responses to
time, there will be a bias in the estimates.

The CROPK function for all vegetable crops was

CROPK = -0.8701 + 8.0756PG 10.5759PG2 + 3.9775PG3 (6)
(0.1347) (0.7849) (1.3802) (0.7481)
where PG = proportion of the way through the growing season, and
0.20 < PG < 1.00


16







For PG < 0.20 or PG > 1.00, the CROPK was set to 0.4. That is, it
is assumed the CROPK is 0.4 early in the season, until about 20
percent of the way through that season, and then returns to 0.4
after harvest. It is assumed this CROPK is suitable for explain-
ing the ET on bare ground. The 20-percent break point is also
based on the results for tomatoes (Jones, et al., 1984).

The standard errors are in parentheses under the parameter
estimates, all indicating statistics significance at least at the
0.0002 level of probability. The R value was 0.98, suggesting a
high degree of reliability for estimating the CROPK for tomatoes,
based on the data available.

Notice this functional form allows the CROPK > 1.0 during
mid-season, around 50 to 60 percent of the time distance from
planting. This was the peak water-using period for tomatoes in
S the experiment.


CROPK for Citrus, Pasture, Golf Courses and Turf

The CROPK estimating approach for citrus, pasture, golf
courses, and turf were all handled similarly. Data were available
from various experiments carried out in Florida, as reported by
Jones, et al. ( 1984 ). These reported actual ET values were
simply divided by the Penman potential ET estimates for Belle
Glade from the same publication. More directly, the monthly AET
values of Table 9 by Jones, et al. ( 1984) were divided by the
first column of Table 7 in the same publication, which is the
Penman estimated potential ET (free water surface ET multiplied by
0.70). The citrus estimates are based on the values obtained at
Ft. Pierce. The CROPK values estimated using this approach are
shown in Table 2.


ORGANIC SOIL MODEL (ORGSOIL AND WTPRED)

The organic soil model is composed of two subroutines. One
is called WTPRED, meaning "water table prediction model," which is
used to update the current, daily water table based on rainfall,
S irrigation, drainage, and evapotranspiration levels. The other is
called ORGSOIL,which incorporates the logic and algorithms chosen
as regards the decisions to irrigate and drain. ORGSOIL is used
to control WTPRED. The overall logic is derived fror the fact
that farm/firm managers attempt to control various irrigation/-
drainage/water table parameters. Thus, the program is developed
such that the user can control the same variables and parameters
in the simulation.


Estimation and Use of the Water Release Prediction Equation

The basis for the organic model is a soil water release
equation. This equation was developed statistically, using exper-
imental data, as described in the following.


17







Table 2. Crop coefficient (CROPK) values for citrus, pasture,
golf courses, and turf, South Florida.

Citrus Pasture/turf

Month PETa AETb AET/PET AETb AET/PET

Inches Inches Inches
Aug 4.33 4.64 1.07 4.80 1.11
Sept 3.78 4.13 1.09 3.86 1.02
Oct 3.19 3.66 1.15 3.42 1.07
Nov 2.20 2.48 1.12 2.48 1.13
Dec 1.81 2.20 1.22 1.93 1.06
Jan 1.89 2.20 1.17 2.01 1.06
Feb 2.44 2.20 0.90 2.52 1.03
Mar 3.38 3.19 0.94 3.35 0.99
Apr 4.21 3.54 0.84 4.21 1.00
May 4.60 4.53 0.98 5.20 1.13
June 4.45 5.08 1.14 4.25 0.96
July 4.45 5.00 1.12 4.80 1.08

aPotential evapotranspiration from Jones, et al. (in press), Table 7.
Actual evapotranspiration, from Jones, et al. (in press), Table 9.


18








Soil Water Relations and Data Source


Experiments were conducted at the Agricultural Research and
S Education Center, Belle Glade, Florida in the late 1950s as
regards the soil water release features of Everglades peaty muck
soils (Harrison, 1959; Weaver and Speir, 1960). These studies
formed the basis for the statistical estimation of a water release
relation, and the eventual development of WTPRED. The following
is a discussion of some of the fundamental soil-water relations
forming the basis for WTPRED.

The basic soil water desorption moisture-tension curves are
depicted in Figure 3, showing the relation between the pressure
exerted on a profile of this soil and the resulting water con-
tent. The intercept of any one curve with the horizontal axis
represents the percentage of that soil layer that is water.
SMovements away from that value show what the water content becomes
after suction, shown on the vertical axis, is exerted for a cer-
tain interval of time (Weaver and Speir, 1960, p. 2); these data
represent hydrostatic equilibrium. The calculations for the 24-
inch water table are shown in Table 3 to illustrate the use and
interpretation of these functions.

The average suction exerted on the 0-to 3-inch layer of soil,
for example, with a 24-inch water table is 22.5 inches, or 57.15
centimeters (Table 3). The log10 (57.15) = 1.76, which is the
scale of the vertical axis in Figure 3. Reading across from the
1.76 value yields an estimated water content of 58 percent in that
layer, or 1.74 inches. The saturated water content, in turn, is
about 80 percent, or 2.38 inches. Therefore, the amount released
from that layer, or drained by natural forces, assuming a 24-
inch water table, is about 0.64 inches (Table 3).


The suction exerted is less on each subsequent layer. For
example, at the 12-to 24-inch layer, there is only 6.0 inches of
suction, because that is the distance of the WT from that layer
(on the average). This is one reason there is more water per inch
at the deeper locations, although soil compaction at the surface
may also influence the differences. Following the same procedure
Sfor each of the other soil layers gives the overall predicted
release of 3.30 inches.

The release of about 3.30 inches is shown in Figure 4 for the
"complete water release curve" (Weaver and Speir, 1960, p. 6),
starting with full saturation. This value also represents the
total storage capacity in t e top 24 inches, given a 24-inch water
table, as shown in Table 4.




8The value in Table 4 is 3.27 inches. The difference is due
to the error in reading off the curves of Figure 3. The raw data
were not included in the Weaver and Speir paper, necessitating
such approximations.


19


















5 .5
S3.0---------




S. --- 0-3 "".

----- 12-24
1.- .......-..... 24
.o 3 .\ -.,


\IL \V.
0.50-------------------------------------------------- -- \ t------ -

05- ---------------- ----.'

10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Water Content (percent by volume)

Figure 3. Desorption moisture-tension relationships in Everglades peaty muck.














Table 3. Example calculations from desorption relations, 24-inch water table in organic muck soil, Belle
Glade, Florida.


Soil layer

Inches

0- 3
N 3-6
6-9
9-12
12-24


Inches

22.5
19.5
16.5
12.5
6.0


Pressure exerted

Cm log10 (cm)

57.15 1.76
49.53 1.69
41.91 1.62
31.75 1.50
15.24 1.18


Total


Water content in
top 24 inches

Percent Inches

58 1.74
64 1.92
64 1.92
71 2.13
88 10.56

18.27


Water content
at saturation

Percent Inches

80 2.38
85 2.55
82 2.48
86 2.58
96 11.58

98 21.57


Difference

Inches

0.64
0.63
0.56
0.45
1.02

3.30


--









7.00


.C
o 4.00



0
.- 3.00


I


0 3 6 9 12 15 18 21 24 27 30 33 36
Depth to rising or falling water tables (inches)


Figure 4. Water gain and release characteristics in Everglades
peaty muck.


22


4Y


I --







Table 4. Water storage capacity, or water released, from
Everglades peaty muck for a receding water table.

Depth to Total water
receding storage capacity, or Predictedb
water table water released

-----------------Inches----------------


0
3
6
9
12
15
18
21
24
27
30
33
36
39
42
45
48


0.00
0.09
0.30
0.63
1.08
1.56
2.10
2.76
3.27
3.84
4.38


0.00
0.22
0.49
0.81
1.17
1.58
2.03
2.53
3.08
3.67
4.31
4.99
5.72
6.50
7.32
8.19
9.11


aSource: Weaver and Speir (1960), Table 4, p. 8, for the 0 to 30
inch depths.
b Predictions from Equation (7).


23








The other release curves in Figure 4 were derived by Weaver
and Speir from the "complete water release curve." Basically,
then, each of these curves simply represents a horizontal shift to
the right of the one curve found for releases from full satura-
tion. The procedure followed in this study, then, was to simply
fit a curve to the "complete water release curve." The one func-
tion is used to predict the response from all depths of the water
table.

One potential problem with using this approach is the data on
water release is for a hydrostatic equilibrium, which may not be
reached in one day. There is no exact information on what the
response would be in one day, the time step used in this simula-
tion model. Thus, there will be a certain bias in the predic-
tions. It is expected the model will "over-predict" the pumping
needs to maintain the goal water table selected in each case.
More directly, the WT change predicted to occur in one day with
the model really occurs in hydrostatic equilibrium. Thus, there
may be more irrigation and drainage activity shown by the model
than will actually occur in the "real world." The bias may not be
great, however. Harrison and Weaver (1958), in another paper
concerned with the same experiment, noted about 85 percent of the
water was actually removed within the first 24 hours (p. 187).
This seems to imply there could 0be about a 15 percent "over-
reaction" predicted by this model.

There were no data on water gain. Thus, it was assumed
herein that gain in the water table would be the "reverse plot,"
or mirror image of the release curves. This is partially justi- *
fied by the fact that the water release curves are developed from
an initial wet profile, from ground level to the starting water
table. Weaver and Speir did not develop dry gain curves, but did
indicate they would be only slightly "flatter." Thus, hysteresis
was ignored in this modeling effort because of the lack of data.
This may have led to some "over-responsiveness" of the model, with
regard to the irrigation decision. That is, for any given amount
of water added as irrigation, the water table will probably
respond less than V2.2 predicts; how much less is not currently
known. This could contribute to a prediction of more drainage
than actually occurs in actual field conditions.









9They also note: "this remaining water (the 15 percent) will
require at least an additional seven days to completely drain down
to the water table, unless first depleted by evapotranspiration
(p. 187)." Thus, ET could affect the rate of drawdown, and also
remove some of the bias expected in this model.

10See the model validation section.


24







Empirical/Statistical Model

Data for the regression analysis were taken directly from the
S "complete water release curve" in Figure 4. The curve demon-
strated an apparent quadratic relationship, thus, the form hypo-
thesized was

WINEST 0.00666(WT) + 0.0026(WT)2 (7)
(0.0041) (0.0001)
where WINEST = the water inflow or outflow in inches/day;
WT = the water table depth, in inches below the sur-
face.

The model was fit without an intercept, reflecting the principle
that there is no water release (or soil water storage capacity)
when the profile is saturated. The t-statistics were all signifi-
cant at least at the 0.01 probability level; the R2 is not pre-
sented because of the forced fit through the origin.1


Constants, Arrays, and Variables (CAV) in WTPRED

The CAV used in WTPRED are defined in Table 5. There are
several other indices, counters, and intermediate variables in the
actual code. These can be understood by referring to the flow
chart in Appendix A, Figure A.5.


Operation of WTPRED

Equation (7) formed the basis for the development of
WTPRED. However, the model required the ability to predict the
water table depth given the inflow/outflow of water, the opposite
of the hypothesized statistical model. That is, in order to
preserve the statistical properties of the data set, it was neces-
sary to fit the equation as noted above. At the same time, it was
necessary to predict the water table, given the inflow. Rather
than using the inverse of the equation, an "iteration" process was
developed whereby the program approximates the WT, given an esti-
mate of the water inflow or outflow for the day.

The current inflow or outflow for the day, XWIN, is brought
in from ORGSOIL. It is then adjusted by 1) adding the value of
the total inflow or outflow to date (TOTWINL) and 2) subtracting
the water storage capacity of the soil when the water table is at
the goal level, called RELEASE. This facilitates always basing



11However, another estimate was accomplished with the inter-
cept term, which gave R = 0.9997 from the same quadratic form,
suggesting an appropriate equation type. The intercept was sig-
nificantly different from zero at the 0.05 probability level.
However, the term was small (-0.33), and logic dictated the func-
tion must pass through the origin.


25












I


Table 5. Definitions of constants, arrays, and variables in
WTPRED.

SOILCP(11) Inched of water release, or total soil storage
capacity, at various depths of water table
ZEROWT(11) Depth in inches to the water table, as
associated with the SOILCP() values
XWIN After substitution of RELEASE and addition of
TOTWIN, this is the net water entering or
leaving the soil profile since the first day,
in inches
RELEASE Inches of water released from saturation to the
goal water table
TOTWINL Net water entering (> 0) or drained from the
soil (<0) to date, at the end of "yesterday"
TOTWIN(365) Same as TOTWINL, but for each day of the
simulation
WINNET(366) Value of XWIN each day, after adjusted for
RELEASE and TOTWINL
WINF Absolute value of the net water inflow (XWIN)
to the soil profile, in inches of water
IWINF Intermediate value of WINF, used to "round-off"
WINF to the nearest 1/1000 inch
WINEST Estimated water inflow, associated with the
current level of the water table, updated
iteratively until it matches the rounded WINF
XWT The incoming (to WTPRED from ORGSOIL) current
level and outgoing (back to WTPRED) predicted
water table level
WINSTD Used to calculate the amount of water that will
fill the profile, in the case there are flood
waters
FLDW(365) Flood waters on each day


26







the current predictions of the water table as if starting at
saturation conditions. This is what is measured in the water
release curve.

XWIN is then assigned to WINNET(J), an array that simply
stores the XWIN values for the entire year. The user can request
these as output. The water table is then set to XWT = 0, because
every prediction assumes a start at saturation.

The next step is to round XWIN to the nearest 1/1000 inch.
This new value (WINF) is compared iteratively to the values of
soil storage, until the first value of soil storage, just greater
than WINF, is found. If this level of storage is within 1/10000
inch of WINF, then the inches of change in the water table level
(X), corresponding to this soil storage capacity, is added to or
subtracted from the initial water table level. The change is
added if the inflow is strictly negative, and subtracted if posi-
tive, the latter case resulting in a rise in the water table.

In the iteration above, the last unacceptable level of soil
storage capacity provided for a change in water table at a level
of 3 inches higher (mathematically lower) than the changed water
table (X). The latter corresponds to the accepted level of soil
storage capacity for corresponding 3-inch increments of change in
water table level. Consequently, if the accepted level of soil
storage capacity, discussed above, is not within 1/10000 inch of
WINF, then 1.0 inch is (iteratively) subtracted from the corre-
sponding level of change (X) in water table. This new level of
S change in water table (X 1.0) is used in Equation (7) to arrive
at an estimate of water inflow (or release), WINEST. This value
is rounded to the nearest 1/1000 inch, and then compared to
WINF. If WINEST exceeds WINF, the iteration is repeated. Another
inch is subtracted and the new level of change in water table (X -
2.0) is used to predict a new value for water inflow. The new
WINEST will be compared with WINF, but clearly, a third iteration
will not occur since a change in water table of (X 3.0) is the
same as the last unaccepted change in water table.

If WINEST does not exceed WINF, either at the first or second
iteration, and if WINEST is within 1/10000 inch of WINF, then the
inches of change in the water table level (X 1.0 or X 2.0),
are added or subtracted. Again, this is with respect to a
strictly negative, or nonnegative inflow. If, on the other hand,
WINEST does not exceed WINF, but WINEST is not within 1/10000 inch
of WINF, the set of iterative changes in water table level is
repeated. The difference, however, is that a 1/10 inch increment
is added to the change in water table (X 1.0 or X 2.0) from
the second set of iterations. Nine iterations are allowed at this
stage; then, a value of WINEST within 1/10000 inch of WINF is
accepted. Or, if WINEST is not within 1/10000 inch of WINF, a
fourth set of iterative changes in water table level is entered.

If this final stage is reached, a 1/100 inch increment is
subtracted from the change in water table level. The final result
may look like: (X 1.0 + 0.5 0.09), where X is determined in
the first set of iterations, 1.0 is the first iterative value in


27








the second set of iterations, 5 iterations in the third set of
iterations result in 0.5, and for 0.09, all nine iterations were
necessary in the fourth set of iterations.

At this juncture, there will be an estimate of the new water
table (XWT). If XWT is at or below the surface, WTPRED is
exited. If not, there are flood waters, and the total water into
the soil to date is set to the RELEASE value. This simply means
the profile is full to saturation.


Characteristics and Nature of ORGSOIL

ORGSOIL is a detailed program designed to monitor and sim-
ulate the movement of water in the top 48 inches of an organic
muck soil. V2.2 uses the water release curve for the soil in the
Everglades Agricultural Area; however, the same program can be
used for other areas if the basic soil release equation and the
various storage and pumping coefficients are changed. The major
features of ORGSOIL are discussed in the following.


Coefficients, Arrays and Variables (CAV) in ORGSOIL

The CAV used in ORGSOIL are defined in Table 6 in the order
the entities are found in the program. Some were defined in MAIN
as well, because of the initialization process. Those definitions
are included again to facilitate ease of interpreting the content
of ORGSOIL.


Initial Conditions

MAIN supplies values for several of the parameters and con-
trol variables, as discussed earlier. In addition, each time
ORGSOIL is called from MAIN, several variables are initialized.
These include all the daily water balance arrays, namely
FLDW(365), WPUMPI(365), WPUMPD(365), SET(365), TOTWIN(365),
TSW(366), and BWT(366), all of which are defined in Table 6. All
of these are set at 0.0 before the start of the first day.

The water table is set to the goal water table before the
first day, by BWT(J) = GWT (Appendix A, Figure A.6). Then, the
total soil water down to bedrock is calculated and the amount
released for the GWT is determined. The net result is that the
initial total amount of water in the soil is that amount down to
bedrock, but associated with the GWT specified by the user. The
total water in the soil to bedrock was determined from Figure 2.
The intercept of the curves in Figure 2 with the horizontal axis
represents the percentage moisture content in each depth of soil
when fully saturated. For example, the estimate is about 79
percent for the 0- to 3-inch layer, meaning there is about (0.79 x
3) = 2.37 inches in that layer. The estimates are shown in Table
7. Note it was assumed the percentage for the 24-to 36-inch layer
was appropriate for all layers deeper than 36 inches.


28








Table 6. Definitions of constants, arrays, and variables in
ORGSOIL.


b FLDW(365)

WPUMPI(365)
WPUMPD(365)
WIN(365)

SET(365)
TOTWIN(365)

RELEASE

BWT(366)

GWT

DEPTHS

TSW(366)






SWSTAR

IDAYS, IDAYE

TDWT, TIWT


TDWET, TIWET



TDDRY, TIDRY

NAREA
MCROP
ADJET


ET

PERGS

SEASON
CROPK

RAINWY(365)

XWT


- Flood water in inches, on each day, defined as
water above the surface of the ground
Water pumped for irrigation each day, in inches
Water pumped for drainage each day, in inches
Water inflow (WIN(J) > 0) or outflow (WIN(J) <
0), in inches per day
Evapotranspiration each day, in inches
Net water in or out to date, updated at the end
of the day
Inches of water released from saturation to the
goal water table
Beginning water table, at the start of each
day, in inches
Goal water table, in inches below the surface
of the ground
Remaining depth to bedrock, or depth of organic
muck
Total soil water each day, in inches, defined
as the amount of water the soil can hold to the
depth DEPTHS less the amount released in
lowering the water table to the GWT or BWT(J);
defined for 366 days only because the model
updates the "next day," or J+1, necessitating
one more day in each "year"
Saturated soil water capacity from the surface
to the depth DEPTHS, in inches of water
Julian day, to start and end the soil water
balance simulation
"Triggers," measured by the depth of the water
table at which to start drainage (TDWT) and
irrigation (TIWT), in inches
Distance above the GWT to start drainage
(TDWET) or below to start irrigation (TIWET),
during the "wet" season, defined as April 15 -
October 14, measured in inches
- Same as above, but for the dry season, defined
as October 15 April 14
- Number of the water supply area
- Number code of the crop
- "Adjustment" factor, used to vary the level of
potential evapotranspiration (PET) set as the
maximum on any given day
- Evapotranspiration, in inches, each day (same
as SET(365))
- Percentage of the way through the growing
season, based on the current Julian day
- Length of the growing season, in days
- The crop coefficient, based on Penman potential
ET as the base for calculating ET
- Daily rainfall in the water year, starting
August 1 and ending July 31, in inches
- "Interim" water table level, transferred to
WTPRED, for updating, in inches to water table


29








XWIN, FWIN


ORGPCI


ORGPCD

DVOL


- Most recently calculated water inflow or
outflow, in inches of water, transferred to
WTPRED to use in updating the water table depth
- Pumping capacity for irrigation in the organic
soils, measured in inches pumped per 24- hour
period
- Pumping capacity for drainage in the organic
soils, in inches pumped per 24- hour period
- Volume of irrigation water needed to raise the
water table to the GWT or the volume of
drainage water that must be pumped out to lower
the water table to the GWT, in inches


Table 7. Saturated water content of Everglades peaty muck soils
at various depths.

Depth of Saturated Water
soil content Summation

Inches

0- 3 2.38 2.38
3- 6 2.55 4.93
6- 9 2.48 7.41
9-12 2.58 9.69
12-24 11.58 21.27
24-36 11.70 32.97
36-48 11.70 44.67

Note: Water content is calculated from Weaver and Speir (1960,
Figure 1, p. 3) where the curves intercept the horizontal axis
(see Figure 2).


30








The next step in initialization is to define the "triggers"
for irrigation (TIWT) and drainage (TDWT). This is accomplished
by adding or subtracting distances to and from the GWT, depending
on the current season. It is hypothesized that managers would
keep the WT lower during wet periods. This expectation is based
on the fact that it costs the producer to pump water, and there is
always the possibility of water damage to crops. That is, the
drainage or flood water removal process requires time because of
pump capacity limitations. It is reasoned there may be an eco-
nomic benefit from lower water tables during wet periods and
higher water tables during the dry season. The initial values
reflect this general argument, with drainage assumed to start when
the GWT is reached during wet periods, and when the water table
rises to 6 inches above the GWT during dry periods. Similarly,
irrigation is assumed to start when the water table drops to 6
inches below GWT during wet periods, but at 0.5 inches below GWT
in dry periods. The model is sensitive to the selection of these
parameters. Thus, the user may wish to try various vaues, which
can be easily changed in the interactive mode in MAIN.

After all the initial conditions are set, the program enters
the Day "loop." Note that all of the parameters, water table
depths, total soil water variables, and other variables discussed
above are re-initialized every time ORGSOIL is called from MAIN
(Appendix A, Figure A.6). Thus, the user can define different
initial conditions on every tract of land in the water supply area
under analysis.


Definition of a "Day"

Various activities which occur simultaneously during a "day"
in the "real world," must be sequential in the computer model.
Thus, some assumptions had to be made regarding the sequencing of
calculations. It is assumed to rain first (if it rains at all),
followed by the occurrence of evapotranspiration. The model then
proceeds to check if there is water above the ground surface,
meaning there is a flood. If there is, flood waters are removed
at the rate allowed by the ORGPCD constraint. The program then
proceeds to the next day.

If there is no flood water, the program checks for the need
to drain or irrigate. If the water table is within the boundaries
of the "triggers," the program again proceeds to the next day. If



12The choice of these values was not completely arbitrary,
but was based on discussions with D. Harrison, Extension Agricul-
tural Engineer, IFAS, University of Florida. Prof. Harrison indi-
cated the "0-inch" rule for drainage and the "6-inch" rule for
irrigation was most typically followed. However, this rule seemed
to give too much pumping, so the seasonal breakdown was devel-
oped. The seasons were decided on the basis of Casselman (1970,
p. 13) as "wet" for the period May I October 30, and "dry" for
November 1 April 30.


31








it finds, instead, a need to drain (water table is above TDWT),
then drainage occurs (the amount pumped is again constrained by
ORGPCD), and then the program loops to the next day. If the water
table is, instead, below the "trigger" for irrigation (TIWT), then
water is added, bounded by the limit on irrigation pumping
ORGPCI. Then, the program loops to the next day.


ET Estimates for Each Day

At the beginning of each "day," the simulator selects the ET
appropriate to either flood or non-flood conditions. If there is
water above the surface, the free-water-surface Penman value is
used, but it is adjusted to account for the effect of plant shad-
ing (Jones, et al., 1984 ). The adjustment factor in V2.2 is
0.90, chosen somewhat arbitrarily, but reflecting the fact that
evaporation would be reduced. It is also assumed this value will
be the same for all crop types, which may also be questionable.
That is, it might be expected that it would be higher for shorter,
thinner crops and lower for taller, heavier crops, because of the
shading effect. However, there was no apparent empirical basis
for this contention, so the 0.90 factor was used for all crops.

If there are no flood conditions, the program selects the ET
calculation section appropriate to the crop currently under con-
sideration. In those cases where there may be several days before
the crop emerges (after planting), or for those days included in
the simulation after harvest, the Penman ET is adjusted accord-
ingly. This may occur for the non-continuously growing vegetable
crops, where the crop season is shorter than the water year. For
these cases, the potential ET is multiplied by 0.40 (CROPK), to
represent essentially "bare" ground conditions. If the ground is
not bare, this factor must be changed. In V2.2, this requires a
change in the actual program code of ORGSOIL. This value was
derived from the results for bare ground shown for the tomato
experimental data in the Jones, et al. (1984) publication.


Definition and Removal of Flood Water

Flood water is defined simply as all water found above the
surface of the ground on any given day. It is removed as quickly
as possible, using the pumping capacity defined by ORGPCD, with
the default value set at 1.5 inches/24-hour period.

If the quantity of flood water is more than ORGPCD, the model
simply removes ORGPCD and proceeds to the next day. During the
next day, the model will again pump as much flood water as needed
or allowed.





13Personal communication with D. Harrison, Extension Agricul-
tural Engineer, IFAS, University of Florida.


32







If the flood water on any given day is less than ORGPCD,
several adjustments may be made. First, all the flood water is
removed. Then, an adjustment is made to the WT, to reflect the
S fact that some water was removed from the soil. For example, if
there were 0.5 inches above the surface, the program would also
remove 1.0 inch from the available ground water. This would place
the WT at about the 12-inch level (Table 4). If the goal water
table for this example were 9 inches, the movement would have been
too far. Thus, the program calculates the amount that must be
removed to reach the GWT = 9.0 (0.63 inches of water, Table 4),
and the total flood water pumped would be 0.5 + 0.63 = 1.13
inches. If the GWT was greater than 12 inches, then WT = 12 and
flood water removed is 1.5 inches, and the program loops to the
next day. Note that flood water is included in the water-pumped-
for-drainage array (WPUMPD(365)).


Definition and Removal of Drainage Water

Drainage water is that water in the soil below the surface
and above the "trigger" drainage depth. Using the preceding
discussion of flood water above, we see that sometimes flood water
and drainage water overlap in their definitions. That is, the
flood water removal subroutine is actually into the drainage zone
for those cases where flood water is less than ORGPCD.

The drainage algorithm is activated whenever 0 < WT < TDWT.
Water is then removed at the rate of ORGPCD, unless such removal
would lower the WT below the GWT. The program calculates the
volume of water removal necessary (DVOL) to return the WT to the
GWT. If DVOL is less than the ORGPCD, DVOL is pumped and the WT
is returned to the GWT. If DVOL is greater than or equal to the
ORGPCD, the WT is lowered in accordance with the outflow of
ORGPCD.


Addition of Irrigation Water

If the check on the WT indicates a depth below the "trigger"
for the start of irrigation, the program makes several adjust-
S ments. The first task is to calculate the volume of water needed
to raise the water table to the goal water table.

Then adjustments are made for the pumping capacity. If the
volume requirement is less than ORGPCI, that volume is pumped. If
the necessary volume is more than or equal to ORGPCI, then the
ORGPCI value is added. In both cases, the WT is adjusted to the
new depth before looping to the next day.


End of Year Calculation

After all the daily calculations, ORGSOIL calculates the
average water table the model was able to maintain throughout the
year. This value is used in the projection of sugar cane yield.


33








GENERALIZED "POT" SOIL MODEL (POTSOIL)


A generalized soil water model was created to simulate all
the soil types other than the organic mucks. The model operates
under a simple principle of monitoring the water inflow and out-
flow from a defined system, such as a "pot," and thus the name
POTSOIL. As with ORGSOIL, this model is based on the overall
logic that farm managers attempt to control the amount of water
available in the root zone. Control is exercised through choice
of values for the number of days between irrigations, application
rate, and the irrigation rule. The flow chart for POTSOIL is in
Appendix A, Figure A.7.


Coefficients, Arrays, and Variables (CAV) in POTSOIL

The CAV used in POTSOIL are defined in Table 8 in the order
the entities are found in the program. Notice that some, while
having the same name, must be interpreted slightly differently
than when the same variable is used in ORGSOIL. For example,
WPUMPD(365) in POTSOIL refers to water going to deep percolation
and/or flooding, in contrast to ORGSOIL, where the same variable
does not include any deep percolation. The reader is cautioned to
review the variable definitions in each subroutine.


Initial Conditions

MAIN supplies most of the parameters and constants used in
POTSOIL, and some of the initial values of variables. This ini-
tialization process is discussed in the section of this report
concerned with MAIN and PARAM. In addition, each time POTSOIL is
called from MAIN, several initialization processes are imple-
mented.

First, the daily water arrays are set at 0 (Appendix A,
Figure A.7). These include the WPUMPD(365), ESM(365),
WPUMPI(365), SET(365), and WAVAIL(365) arrays, all defined in
Table 8.

Secondly, IRRD is set equal to negative MIRRD, which insures
that an irrigation would be allowed the first through the MIRRD
day of the simulation. That is, it is assumed the last irrigation
took place at least MIRRD days before the start of the water
season for any particular simulation.

Third, the beginning available soil moisture (AVWAT) is set
at field capacity less water in the soil at permanent wilting
point (the maximum available soil water, by definition). This
simply means it is assumed the irrigation manager starts out with
a full "pot" of water on Day 1. The quantity of water available
at the point where plant stress starts is then calculated
(TSTRES). The slope of the soil coefficient is then set before
entering the Day loop; these are discussed below as related to the
ET adjustment factors.


34








Table 8. Definitions of constants, arrays, and variables in
POTSOIL.


WPUMPD(365)


WPUMPI(365)
ESM(365)




SET(365)
WAVAIL(365)




IRRD
MIRRD
AVWAT

TSTRES

BSLOPE
FCAP

PWP

XESM

IDAYS, IDAYE

DWRULE(12)


DWRATE
IRCHEC



SM

PERGS

CROPK


SOILK



S STRESS


- Water removed because of drainage needs, and by
deep percolation and runoff; may or may not be
pumped, in acre inches
- Water added for irrigation, in acre inches
- Soil moisture at end of day, in acre inches,
representing water available for plant use plus
water remaining at permanent wilting point.
This is field capacity when the profile is
full.
- Evapotranspiration on each day, in acre inches
- Measure of water available for plant growth on
any given day, calculated as the sum of
rainfall plus water added for irrigation less
deep percolation and runoff each day, in acre
inches
- Julian day of the last irrigation
- Number of days between irrigation
- Maximum water holding capacity of the soil
available for plant use
- Available water in the soil at the point where
plant stress is started
- Slope coefficient for the SOILK equation
- Field capacity, amount of water in the root
zone, in acre inches
- Water level in the soil at which the plant
suffers permanent wilt, in acre inches
- Soil moisture, same as ESM(), but the amount at
the beginning of the "day"
- The day simulation is started and ended
respectively, in Julian days
- The proportion of available water at field
capacity when an irrigation is to start, in
acre inches, one rate for each month
- Rate of water application per irrigation
- Difference between the time of last irrigation
and the system capacity constant, in Julian
days, used to determine if an irrigation can
take place
- Soil moisture, same as ESM() and XESM, but an
intermediary value, calculated during the "day"
- Percentage of the way through the growing
SEASON
- Crop coefficient, the ratio of actual ET to
potential ET, as affected by atmospheric and
crop features
- Soil coefficient, the ratio of actual ET to
potential ET, as affected by soil moisture
conditions, and crop susceptibility to water
stress
- Percentage of available water at which plant
stress starts


35








Definition of a "Day"

As mentioned in the discussion of ORGSOIL, evapotranspira-
tion, deep percolation, and rainfall processes all occur simul-
taneously in actual operating conditions. Again, a sequence of
calculations are necessary in order to simulate these conditions
with the computer.

The program first checks for the need to irrigate, based on
the "trigger" level selected by the user. If water levels are
low, irrigation proceeds as bounded by the pumping constraint.
The soil water is then updated, and the evapotranspiration (ET) is
subtracted. The water going to deep percolation or runoff is
calculated. The program then updates the soil moisture level and
loops to the next day.


ET Estimates and Adjustment (SOILK and CROPK) Factors

After water has been added for irrigation, if appropriate,
the ET is calculated for the day. As noted in the section on ET
estimation, the Penman free-water-surface estimate is multiplied
by 0.70, with this serving as the base (except for the empirical
sugarcane functions). The SOILK and CROPK values are updated each
day, and multiplied by this base value.


System Capacity and Addition of Irrigation Water

The irrigation system capacity variable is brought in under
the notion that each field in the farm cannot be irrigated until
MIRRD days have passed. For example, a typical design standard
for citrus in Florida is an application depth of 2 inches every
MIRRD =7 to 8 .

An irrigation can take place every MIRRD days, then, at
varying levels of RATE. The smaller the MIRRD and the larger the
RATE, the more water will be used in any given year. Raising RATE
also generally causes more water to go to deep percolation and
runoff in any given year.

The "trigger" for irrigation is based on the percentage of
available water at field capacity. It is calculated by indicating
the percentage; the program uses the calculated soil moisture
level as the trigger. For example, with a FCAP = 4, PWP = 1, and
PFCAP = 0.60, the trigger soil moisture level is SM = 1.8
inches. An irrigation will occur, with RATE applied, whenever SM
is less than 1.8 on any given day.


Drainage. Percolation, Flood, and Runoff Water

All excess water (defined as SM > soil capacity for available
water AVWAT) is added and reported as WPUMPD(365) water on any
given day. The value is calculated by difference, simply by
assuming a water balance between inflows and outflows for the


36








day. For example, if SM = 1.0 inches in "the morning," an irriga-
tion would be triggered with the 1.8-inch rule cited above. If
RATE = 1.0, the SM = 2.0; however, if RAINWY(365) is, for example,
S2.0, then there would be some excess water (SM > AVWAT), or 1.0
inch of WPUMPD(365). It is assumed to rain "after" the irrigation
has occurred for the day. This also assumes that all rainfall
received on a day after the soil is refilled to capacity with
irrigation will be lost to deep percolation and/or runoff, except
for what may have gone to ET on that day.

Notice there is no attempt to separate the different types of
excess water. Such separation is possible only if the soil model
is designed to account for the amount of water at each soil depth,
as is done in ORGSOIL.


SEnd of "Day" Update of Available Soil Water

After all irrigation, drainage, and ET effects are estimated,
the "end-of-day" values of SM are calculated using the starting
soil moisture "this morning" as the base. The water balance
relation is statement 50 in POTSOIL (Appendix A, Figure A.7). The
water available during the day is first calculated by subtracting
deep percolation/ runoff estimates from the starting plus inflow
values. Then, the soil moisture at the end of the day is cal-
culated by subtracting the evapotranspiration for the day. The
program then "loops" to the next day.


Values Provided by POTSOIL

The net result of all the iterations through the Day loop is
the daily water balance values for the tract being simulated.
These are then made available to TABWAT and PROFIT for further
processing. All calculations are in inches/acre/day.


CALCULATING PROFITS AND ECONOMIC IMPACT (PROFIT)

The PROFIT subroutine accepts input from various other sub-
S routines to calculate the costs and returns for the various crops
being considered in the simulation. The program simply serves to
calculate profits and maintain a summary file of the current level
of profits for the area (Appendix A, Figure A.9).


Constants, Arrays, and Variables (CAV) in PROFIT

The various constants, arrays, and variables (CAV) in PROFIT
are defined in Table 9. The current values of each of the vari-
ables are initialized in MAIN and transferred to PROFIT, via
COMMON statements, on the first run through the program. Only the
variables used to get totals of the various magnitudes continually
updated in PROFIT are initialized, and this is accomplished on
each calling of PROFIT from MAIN.


37








Table 9. Definitions of constants, arrays, and variables in
PROFIT.


TCI

TCD
AVCWI

AVCWD

WIRRM(12)

WDPROM(12)

ACRES(12)
TPROD

YIELD
MHARV
PROPH
AVCH

WCI, WCD



FC

Cl
C2

C3

C4


C5
C6

C8

CSTRET(12,8)



TACRES

THARVC

TWATC

TOCOST

TTCOST


- Total costs of irrigation water, dollars per
acre
- Total costs of drainage water, dollars per acre
- Average variable costs of irrigation water,
dollars per acre inch
- Average variable costs of drainage water,
dollars per acre inch
- Water pumped for irrigation in month I, acre
inches
- Water pumped for drainage in month I, acre
inches
- Acres of crop irrigated in month I
- Total production for the entire tract, in
thousands of crates, tones, etc.
- Yield per acre
- Month of harvest, August = 1, ..., July = 12
- Proportion of acreage harvested
- Average variable costs of harvest, per unit
harvested per acre
- Water costs for irrigation and drainage,
incurred before the start of the water year
chosen by the user, in thousands of dollars
expended for the entire tract
- Fixed or other costs of production, in dollars
per acre
- Thousands of acres harvested for the tract
- Thousands of dollars expended for harvest on
the tract (variable costs)
- Thousands of dollars expended for irrigation
and drainage (variable cost)
- Thousands of dollars for other costs on the
tract, including all the fixed costs and other
variable costs
- Sum of all the costs
- Thousands of pounds, crates, tons (unit varies
by crop type) produced on the tract
- Thousands of dollars in returns from the sale
of commodities from the tract
- Array summarizing the items in Cl C8 across
all tracts; the summary values are inserted
into the row corresponding to the month of
harvest
- Total acreage across all tracts for all 12
months
- Total harvest costs for all tracts for all 12
months
- Total water costs for all tracts for all 12
months
- Total other costs for all tracts for all 12
months
- Total all costs for all tracts for all 12
months


38


I


~I








TTPROD Total production across all tracts for all 12
months (meaningful if and only if there is only
one crop type in the simulation, see text)
TPROF Total profit in thousands of dollars across for
all tracts for all 12 months
INTER "Switch," set at value of 2 for batch
processing, and 1 for interactive processing
AREA NAME Name of the area
NYRSTA Year the water simulation was started
MSUM "Switch," if set to 1, the profit summaries are
written to a file; if set to 2, the program
gives only CRT output


39







Water Year vs. Crop Year Calculations


Cost, return, and profit calculations are accomplished for
the month of planting through the month of harvest. For example,
if the crop was planted August 15 (month 1) and harvested December
15 (month 5), costs and returns are calculated over the period
August 1 December 31. As a result there will be an upward bias
in costs in those cases where the water year is specified as
longer than the crop growing season. If the user specified, for
example, the water year from August 15 to December 15, water costs
are calculated only over that period. If the user specified the
water year as August 1 to January 31, water costs would be cal-
culated from August 1 to December 31.


Operation of PROFIT

Calculations are based on the economic notion that there are
variable and fixed costs within any given season. Version 2.2
allows only the harvest costs and the water costs to vary within
any given season, truly a short-run type of return calculation
process (Appendix A, Figure A.9). PROFIT generates tract esti-
mates and area-wide summaries in the computer-generated "Tables 4
and 5," respectively (see Lynne, 1984).

Caution is advised in the interpretation of the CSTRET(12,8)
summary table which is provided as output in "Table 5" from the
computer. First, the costs, returns, and profits are all reported
in the month of harvest specified by the user, and not in the 4
month when such costs may have been incurred. For example, if a
crop received irrigation water during the months of September -
December, and was harvested in December, all the water costs are
reported in the December row of the computer-generated "Table
5." The logic here is that one wishes to have the net profit for
a tract reported during the month the tract is harvested, such as
to establish the overall and general economic impact.

A second caution relates to the calculation of the total crop
production from the entire area in which production is taking
place. The computer has no way of discerning differences in the
units in which yield is measured per acre for each crop. As a
result the total production estimates for the area will be errone-
ous, unless only one crop is being considered. Otherwise, PROFIT
is adding such diverse measures as pounds, crates, and tons per
tract together to give a meaningless total production estimate for
the area.


WATER BALANCE SUMMARIES (TABWAT)

The subroutine TABWAT is used for only one purpose--to output
summaries of water used per acre, per tract, and for the entire
area. TABWAT provides the computer-generated "Table 1," "Table
2," and "Table 3." These are, respectively, the coefficient/-
parameter summary for the current tract, the per-acre water bal-
ance for the tract, and the current values of the water balance
for the area.


40








Constants, Arrays, and Variables (CAV) in TABWAT

The CAV in TABWAT are defined in Table 10. Most of the CAV
) values refer to monthly and yearly water balance summaries.


Operation of TABWAT

TABWAT is called from MAIN after the water simulation for
each tract is completed. The subroutine is then used to total the
water values (from ORGSOIL or POTSOIL) over the months of the
water year, also specified in MAIN (Appendix A, Figure A.10).
These values are then provided in computer-generated Tables 1-3 in
the program output (see Lynne, 1984).


MODEL VALIDATION

Model validation is a difficult process in the case where
little experimental or "real world" data are available on the
basic processes at work, as is the case in this study. There has
been considerable work on various dimensions of the agricultural
irrigation and water-use process in the laboratory and at smaller
levels of aggregation, such as an acre. However, there has been
little research effort in Florida, at trying to understand the
physical, economic, and biological relations at a tract level, and
even less at a regional or area-wide level. Thus, validation of
the model will have to continue as it is used. Some progress was
made before release of the model, as described in the following.

Data were available on regional water movements into and out
of various parts of the study area from records of the South
Florida Water Management District. These data were obtained for a
portion of the Everglades Agricultural Area (EAA) for validation
purposes. The area designated as S-5A (Figure 1) was selected, as
it is entirely within the EAA, which facilitated use of Florida
Crop and Livestock Reporting Service statistics on acreage and
harvest dates. The latter are especially crucial because of the
complications offered with regard to the vegetable crops.

) Land use data for 1979 were provided for S-5A by the Dis-
trict, as displayed in Table 11. The total acreage for pasture
and turf were used directly in the validation process. However,
several adjustments were necessary for both the sugar cane and the
vegetables. All cost and price coefficients are in Appendix E,
Table E.4.


Vegetable Acreage and Production

The first task for vegetables was to determine the acreage
planted in the EAA. These estimates are shown in Table 12 for
1978-79, indicating nearly 90,000 acres planted that season. This
compares to a land-use estimate for the same period of about
22,000 acres (So. Fla. Water Management District, 1981), veryify-
ing the fact that much of the acreage is multiple cropped through-


41








Table 10. Definitions of constants, arrays, and variables in
TABWAT.


RAINM(12)
WIRRM(12)
ETM(12)
WDPROM(12)

WBAL(12)


MOS, MOE
JS, JE
RAINWY(365)
WPUMPI(365)
WPUMPD(365)
SET(365)
EFFIRR
TRAIN
TWI

TET

TWD
TWBAL
INTER

MSUM

WATSUM(12,6)












AREA NAME
NYRSTA
MONTH NAME


- Rain in month I, inches
- Water for irrigation in month I, inches
- Evapotranspiration in month I, inches
- Water going to deep percolation and/or runoff
in month I, inches
- Water balance for month I, calculated from the
sum of inflows minus outflows for the month,
inches
- Month of start and end of the water year
- Day of starting and ending point for each month
- Rain received on each day, inches
- Water pumped for irrigation each day, inches
- Water pumped for drainage each day, inches
- Evapotranspiration for each day, inches
- Efficiency of irrigation, percentage
- Total rain over all month, inches
- Total water for irrigation over all months,
inches
- Total evapotranspiration over all months,
inches
- Total water drained over all months, inches
- Total water balance over all months, inches
- "Switch," set to 2 for batch processing, and to
1 for interactive processing
- "Switch," set to 1 to store computer Tables 1 -
3 on a file
- Array which includes the water summary values
for each month, as follows for each column:
J=1, thousands of acres
J=2, thousands of acre-feet of water from
rainfall
J=3, thousands of acre-feet of irrigation
water
J=4, thousands of acre-feet of
evapotranspiration
J=5, thousands of acre-feet of water for deep
percolation and runoff
J=6, thousands of acre-feet, water balance
- Area name
- Year the water simulation was started
- Month


*


42








Table 11. Agricultural and other land cover* in S-5A Basin, South
Florida Water Management District, 1979.

Type Acres

Agricultural 121,437
Sugarcane 111,514
Truck crops 6,103
Sod farms 142
Improved pasture 3,678
Other 1,902

Total 123,339

*Estimates from So. Fla. Water Management District (1981). The
sugarcane and truck crop (vegetables) acreage represents land
devoted to this activity, and will be different from planted
acreage.














Table 12. Estimated growing season periods, season lengths, acreage, and turn-over coefficients for vegetable
crops in the Everglades Agricultural Area, 1978-79.

Season Land Percent
Planted Growing period length Turn- used of total
acres (months) (days) (days) over (acres) land

Celery 10,550 Aug-Mar 243 85 2.86 3,689 13.6

Sweet corn 31,900 Fall and Spring 2.00 15,950 58.8

Escarole 5,550 Oct-May 243 98 2.48 2,238 8.2

Lettuce 12,200 Oct-May 243 40 6.08 2,007 7.4

Radish 29,600 Sept-May 272 30 9.07 3,263 12.0

Total 89,800 3.31 27,147 100.0










out the season. Some means was needed to determine how much was
planted of each crop, and when these plantings took place. Unfor-
tunately, there are currently no records maintained for such
action, indicating an area where some resources may have to be
placed to effectively implement a water-use projection capability.

The procedure used herein required several assumptions.
First, a "turnover coefficient" was determined by dividing the
number of days in the growing season by the typical season length
for the crop type. The estimates are shown in Table 12. For
example, celery is usually grown from August to March, or for 243
days. The growing season length is 85 days, giving (243/85) =
2.86. The coefficients vary from 2.00 to 9.07 depending on the
crop (Table 12). The second step was to divide the planted acre-
age by the turnover coefficient, giving an estimate of the land
devoted to each crop (Table 13). Corn was handled somewhat
differently, in that there are only two seasons. The total
planted acreage was simply divided by 2, to yield an estimated
15,950 acres devoted to sweet corn production for the entire
vegetable growing season.

The total land used was then summed, and the proportions
devoted to each crop were calculated. The vegetable land devoted
to each crop is shown to vary from 7.4 percent for lettuce to 58.8
percent for sweet corn (Table 12).

Next, harvested acreage in each month was derived from the
Florida Crop and Livestock Reporting Service (1980, "Vegetable
Summary") for 1978-79. These data were then adjusted, based on
conversations with individuals familiar with the production and
cultural practices in the area. The primary source of produc-
tion during the winter months, other than from the EAA, is that
flowing from the Zellwood (Seminole County) and Dade county
areas. Thus, some means was sought to separate out the production
from these areas to isolate that from the EAA.

Nearly all the celery comes from the organic muck soil areas
at both Zellwood and the EAA. It was assumed the flow was steady
from Zellwood, for the period of December through June. Further,
it was assumed that all of the production from "Central" Florida
(Florida Crop and Livestock Reporting Service, 1980, "Vegetable
Summary") came from the Zellwood area. This gave an estimate of
142,000 crates a month out of that area in 1978-79, which was
subtracted from that in the EAA. Then, the percentage sold each
month was applied against the total figures in the EAA to yield
the estimates of Table 13. For sweet corn, it was assumed all of
the October through December, all of the April, and 50 percent of
the June harvest came out of the EAA. This assumed that Dade



1More specifically, much of the following was developed
based on conversations with William Llewellyn (Seminole County)
and Kenneth D. Shuler (Palm Beach County), both vegetable exten-
sion specialists for the county extension offices in their
respective county.












Table 13. Estimated vegetable production
1978-79.


Crop


Celery 1000 crates
%

Sweet corn 1000 crates
%

Escarole 1000 crates
%

Lettuce 1000 cwt
%

Radish 1,000 cartons
%


and proportion harvested


Oct. Nov.


283
4.0

888 867
15.4 15.1

24 190
1.0 7.0

4 38
<1 1.7

447 1,685
5.0 19.0


Dec.


995
13.9

412
7.2

406
14.9

310
13.6

1,151
13.0


Jan.


1,018
14.3




494
18.2

283
12.4

975
11.0


each month in Everglades Agricultural Area,


Month
Feb. Mar. Apr. May June


980 1,316 1,138 1,226 185
13.7 18.4 15.9 17.2 2.6


390
14.3

283
12.4

975
11.0


562
20.7

605
26.5

1,239
14.0


1,990
34.6

450
16.5

649
28.4

1,420
16.0


1593
27.7

202
7.4

111
4.9

799
9.0


176
2.0


--


--







County producers supplied the corn during January through March,
and Zellwood produced the other 50 percent during May.

Zellwood was assumed to produce 21.5 percent of the escarole,
spread equally across the time period November through June. This
gave an estimate from that area of 90,120 crates per month, which
was subtracted from the EAA production estimates. This percentage
was based on the production in the "Central" area in 1978-79, as
for the celery (Florida Crop and Livestock Reporting Service,
1980, "Vegetable Summary," p. 29).

Lettuce production was selected in an identical manner. The
Central region produced 322,000 cwt. of lettuce in 1978-79
(Florida Crop and Livestock Reporting Service, 1980, "Vegetable
Summary," p. 31). This was apportioned equally across the seven-
month period from November through May, and subtracted from pro-
duction in the EAA.

Radish production was assumed distributed identically in both
Zellwood and the EAA, with 12 percent coming from Zellwood. This
was based on the assumption that about 4,000 acres were harvested
in the Zellwood area. This compares to the 29,600 acres harvested
for all of Florida in 1978-79 (Florida Crop and Livestock Report-
ing Service, 1980, "Vegetable Summary," p. 37), or 12 percent of
the acreage.

The next step was to apply the percentages of Table 12 to the
estimated land cover data for S-5A. The resulting acreage for
S each vegetable crop is shown in Table 14, which also displays the
estimated planting dates. The planting dates were determined by
"working backward" from the acreage harvested in each month (Table
13) as adjusted for production coming out of other areas. The
growing season length data were subtracted from an assumed harvest
date in the middle of the month, to yield the estimates of Table
14. The total acreage devoted to vegetable production was esti-
mated by the District at 6,103 acres; these were allocated such as
to yield a range of 452 acres for lettuce to 3,588 acres for sweet
corn (Table 14).


Sugarcane Acreage and Production

Sugarcane estimates also required special handling, as the
111,514 acres in area S-5A (Table 11) were not all harvested each
year. In fact, according to estimates by Alvarez and Snyder
(1984), at any given time about 20 to 25 percent of the cane is
being fallowed. It was assumed that 20 percent was representa-
tive. This implies that 20 percent are each devoted to first,
second, and third ratoons, with a replanting at the end of every
fourth year or beginning of the fifth year.

As noted elsewhere in this manual, the sugarcane yield and ET
models are suitable for estimates only over the periods of January
to December of each year, with cane harvest always in December.
Thus, a feasible alternative was to assume that in any given year
the land devoted to sugarcane production would be allocated with:


47













Table 14. Acreage of vegetable crops by planting date for validation of water balances in area S-5A, Ever-
glades Agricultural Area, 1978-79.

Planting date
Crop Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May Total

Celery 33 116 118 114 153 132 142 22 830

Sweet corn 554 542 257 1241 994 3588

Escarole 40 75 91 72 103 83 37 500

Lettuce 69 56 56 120 128 22 452

Radish 37 139 95 80 80 102 117 '66 15 732

Total 627 770 605 350 392 351 1642 1261 88 15 6103







22,302 acres in fallow;
22,302 acres in plant cane;
22,302 acres in first year ratoon;
22,302 acres in second year ratoon;
22,302 acres in third year ratoon.

ET estimates were available only for plant cane and first
year ratoon cane. Thus, this was further adjusted to consider the
entire 66,906 acres as first year ratoon. That is, it is assumed
the ET would not be radically different among the various aged
cane fields, after the plant cane year. It was also assumed the
fallow acreage would be flooded for the entire year; this was
simulated by requiring a goal water table of 0 inches below the
surface, meaning the goal is to keep the soil saturated at all
times.

As noted in the discussion of SCYIELD, yield is set to zero
whenever the average goal water table for the year is higher than
4 inches. Also, fixed costs are set to zero under these condi-
tions.

It is assumed for this validation process that sugarcane
yield will be the same for all ratoons. Thus, the predicted
profit from cane will be higher than can be normally expected as
yields are usually lower as more ratoons are attempted.


Pasture and Turf Acreage

The 1979 land cover data available from the District on
pasture and turf showed 3,678 acres in improved pasture and 142
acres in turf grass in S-5A (Table 11). These were incorporated
directly into the validation data set, with the improved pasture
assumed used for beef cattle (cow-calf) operations.


Choice of Water Management Strategies

The actual strategies used in field conditions were not known
for each of the tracts in the validation area. Conversations with
farmers/ managers in the area suggested the goal water table
values shown in Table 15 may be appropriate. That is, it appears
water tables are kept deeper than recommended for most crops. The
recommendations (Snyder, et al., 1978) are shown in the second
column of Table 15. Both water table strategies are evaluated in
the following.

"Triggers" for the start of irrigation and drainage processes
were also selected, using the limited knowledge of actual field
conditions. Early runs of the simulator showed the model quite
sensitive to the choice of triggers. This suggests an area where
effort must be placed to determine what water managers are ac-
tually doing in individual tracts.

As a base point, it was assumed the appropriate triggers were


49







Table 15. Goal water table depths in validation runs.

Goal water table

Crop Actuall Recommended2

---------Inches--------

Sugarcane 36 30

Celery 18 18

Sweet corn 33 27

Escarole 33 27

Lettuce 33 33

Radish 18 15

Pasture/turf 36 24

Based on conversations with extension personnel and
farmers/managers in the area.
2Recommended goal water table levels, from Snyder, et al. (1978).


i


50


1








TDDRY = 0.5
TIDRY =0.5
TDWET = 0.5
STIWET = 6.0

where TD and TI refer to triggers for drainage and irrigation,
respectively. The DRY and WET designations refer to the dry and
wet seasons, as defined earlier. Thus, a manager is assumed to
start drainage whenever WT, the water table, reaches within 0.5
inches of the goal water table (GWT) during all seasons. For
irrigation, water application is started when WT drops at least 6
inches below the GWT during the wet season, and to 0.5 inches
during the dry period The delay for start of an irrigation during
the wet period allows for more storage of rainfall, which is more
frequent in this time frame.

Under the recommended (R) water table/triggers, it was
assumed the triggers during wet season for irrigation would vary
with the GWT. It was assumed the appropriate TIWET was 12 inches
for a 30-inch GWT, which would allow for the storage of a 3-inch
rainfall and reduce irrigation substantially during the wet summer
months. This was thought a reasonable first approximation because
of the small amount of irrigation that is accomplished during the
summer period. With an 18-inch depth, for example, the trigger
becomes TIWET = 7 inches. With the actual (A) strategy, it was
assumed that TIWET = 6 for all depths considered. In both A and
R, the 0.5 inch triggers were used for all depths during the other
periods, simply reflecting an assumption that farmers/managers
S would drain and irrigate as soon as the WT moved even slightly
away from the GWT.

It was also assumed that irrigation was 100 percent effi-
cient, meaning that all releases reached the tracts. Thus, esti-
mated irrigation levels will have to be increased to account for
actual evaporation losses from canal surfaces, plus lateral move-
ment and deep seepage. Pumping capacities were assumed at 1.5 and
0.4 inches per 24-hour period, for drainage and irrigation,
respectively. These assumptions and those regarding the acreage
and planting dates are all illustrated in Appendix F for the
"actual" (A) case. The data of Appendix F were used in the batch
S operational mode of the model to generate the estimates for the A
case in Table 16.


Water Balance Projections

The water use and drainage results for the 1978-79 period are
displayed in Figures 5 8 from data in Table 16. Version 2.2 is
shown to function moderately well in predicting both drainage and
irrigation during most months, with the exception of September -
November. The historical irrigation demand during this period
indicated very little water being moved into area S-5A. In fact,
the primary activity is drainage, with (40.8 + 30.3 + 12.7) = 83.8
thousand acre feet moved out of the area (Table 16). This repre-
sents 77 percent of the rainfall plus irrigation water available
to the area, based on rainfall records at Belle Glade, suggesting


51










4


25 \

6 20-

* 15 -,
is- / \ ,

10- \

5


Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May Jun. Jul.
Month

Figure 5. Historical and predicted irrigation in thousands of
acre feet for actual water table depths in area S-5A.



40

35

30 g

o 25





S/~,,I,
20




5-


Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May Jun. Jul.
Month

Figure 6. Historical and predicted irrigation in thousands of
acre feet for recommended water table depths in S-5A.


52
















70


60-


I 50- 0

40-
@1
30-
0 Predicted


S\ \/ \
20-





Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May Jun.

Month

Figure 7. Historical and predicted drainage in thousands of
acre feet for actual water table depths in area
S-5a.


E




C
I-
O


70-

60-

50-

40-

30-

20-

10-


Historical











l


Jul.


Predicted

/ IF
..,' /
...,* ..', /


-


Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May Jun. Jul.
Month


Figure 8. Historical and predicted drainage in thousands of
acre feet for recommended water table depths in area
S-5A.


53


i








Table 16. Actual and predicted irrigation and drainage require-
ments, S-5A area, South Florida, 1978-79.

Irrigation Drainage

Predicted Predicted
Month Historical A1 R2 Historical A1 R2

-----------------Acre feet (thousands)----------------


0.0 0.0 0.0


0.0

0.5

5.4

8.4

10.5



8.7

26.6

23.0

10.6

35.9

0.0

129.6


3.0

18.1

17.5

11.5

6.1



5.2

12.6

18.6

0.1

28.8

0.0

121.5


0.0

6.0

30.1

11.7

6.4



7.4

17.8

24.0

0.0

19.6

0.0

123.0


68.0 49.2 49.2


40.8

30.3

12.7

28.8

24.4



0.0

1.6

11.5

7.5

0.0

0.0

225.6


4.7

0.3

14.7

28.4

24.6



0.0

6.2

12.1

22.4

3.8

23.4

189.6


1.8

0.3

15.1

28.6

24.4



0.0

6.1

10.9

16.4

0.0

17.5

170.2


1Actual goal water table, based on conversations with extension
personnel and farmers/managers in the area.
2Recommended goal water table levels, from Snyder, et al. (1978),
with proportional triggers for irrigation in the wet period.


54 )


I4


August

September

October

November

December

January



February

March

April

May

June

July

Total


I








there is some behavioral aspect of the growers and/or water man-
agement personnel not included in the simulation. More directly,
the large amount of actual drainage in consort with virtually no
S irrigation demand during this period suggests there was only 24.9
thousand acre feet going to ET during this 3-month period, or only
0.85 inches per acre per month. This is virtually impossible,
suggesting a problem in the assumed acreage, crop, and management
scenario simulated. This problem can be resolved only with more
data collection, as regards characterizations of management in the
area.

Both the actual and recommended WT simulations follow the
general trends of irrigation water demands during the December -
July intervals (Figures 5 and 6). This suggests the assumptions
for the winter vegetable crop are closer to reality than possibly
for sugar cane, turf, and pasture. The predictions from both the
S (A) and the (R) management sets are very similar.

Drainage predictions reflect the general trends in the his-
torical record from December through June (Figures 7 and 8). In
fact, the estimates for December February are exact for the (A)
strategy, and April is well within 10 percent. Estimates are
similar for the (R) strategy for this same period (Figure 8).
Again, this suggests the assumptions for the winter period are
presumably "reasonable." The large jump in the model's prediction
for May is currently unexplainable, again probably reflecting some
management decision at either the farm or water district level
that was not simulated by the model.

The drainage predictions for the wet/summer period are in the
correct range. The historical total for June, July, and August
was 68 thousand acre feet, all occurring in August. The model
with the (A) data predicted 76.2 thousand, and that for the (R)
data predicted 66.7 thousand. Thus, the substantial difference
regards the distribution of the pumping, and not the total amount.

It was also not known what impact, if any, the upward move-
ment of water from areas below the root zone may have had in S-
5A. V2.2 models only the water balance associated with rainfall
and irrigation as inflows, and drainage and ET as outflows. Any
upward movement of water, from such sources as seepage from Lake
Okeechobee (Meyer, 1971) are not included. If this were a sub-
stantial influence in S-5A, then the irrigation predictions would
be higher than the historical pumping.


Estimates of Evapotranspiration

As noted in the earlier descriptions of the ET estimation
process, estimates for all the vegetable crops were based on the
relative ET for tomatoes. The accuracy of the estimates may be
subject to question because of this procedure, and as a result are
given special attention in the following discussion.

The estimates of vegetable ET were quite different for the
vegetable crops as compared to earlier estimates by Rogers and


55








Marlowe (1976). The predicted actual versus potential ET for a crop
of celery planted at various dates is shown in Table 17. The
estimated ET is considerably lower than predicted by Rogers and
Marlowe using the Blaney-Criddle approach. Yet, the estimated ET
values are bounded by the PET predicted by the application of the
Penman method, as documented by Jones, et al. (1984). In fact, the
Rogers and Marlowe predictions of actual ET using Blaney- Criddle
exceed the potential evapotranspiration (PET) for the area as
predicted by Jones, et al. (1984). Given that the Penman equation
is superior, as argued by Jones, et al. (1984), the estimates used
in V2.2 may be more accurate than the Blaney-Criddle estimates,
although only experimentation would provide sufficient evidence to
verify or deny this contention.

Estimates of ET for ratoon sugarcane are shown in Table 18.
Again, they are compared to the PET from Jones, et al. (1984). The
fact that cane was harvested in December is demonstrated in Table
18, where the ET is shown to be considerably less than the PET in
the months immediately after harvest. This suggests a shortcoming
in the cane ET estimating procedures. That is, V2.2 requires that
the cane be harvested in December. This also leads to distortion
in the irrigation predictions, as cane is actually harvested
regularly throughout the winter period, and ET estimating equations
for later versions of the model must reflect these differences.

Estimates of ET for the other crops considered are not dis-
played here, as they were taken directly from Jones, et al. (1984).
These are the most accurate and well documented, primarily because
the harvest dates are always approximately the same for the con-
tinuously growing pasture and citrus crops. There is a large void
in the knowledge base of how ET varies through time for each of the
crop types having different planting dates.


Economic Impact Projections

The projected costs, returns, and profits for the validation
run using the first set of GWT levels are displayed in Table 19.
The agricultural industry is shown to lose money that year, espe-
cially in the production of sugar cane. Unfortunately, there were
no estimates of actual returns for area S-5A of the Everglades
Agricultural Area, making validation impossible. This suggests
another area in which resources must be directed, to collect
information on returns.

In addition, there is a severe need for information on price
response to varying production levels. This is an especially severe
problem with respect to the production of vegetables. The activity
reported in Table 19 reflects an average price, which may be totally
inappropriate for the production levels actually generated that
year.

I


56








Table 17. Estimates of evapotranspiration (ET) for celery at
various planting dates.

S Planting Estimated Rogers/Marlowe
date ETa PETb Estimatesc

------------------ Inches----------------

Aug. 15(15) 7.3 9.8 12.2

Sept. 15(46) 5.7 7.8 10.6

Oct. 15(76) 4.5 6.2 8.6

Nov. 15(107) 4.1 5.5 7.9

S Dec. 15(137) 4.9 6.4 7.8

Jan. 15(168) 6.5 8.2 8.1

Feb. 15(199) 8.1 10.4 9.1

May 15(227) 9.2 11.8 10.2

aBased on tomato relative ET curve (see text), and a 25-day
growing season.
potential ET estimates for Belle Glade, based on Jones, et al.
* (in press).
CAssumed planted on the Ist of each month and harvested 80 days
later (Rogers and Marlowe, 1976, p. 2) in the Everglades and Lower
East Coast Areas.


57








Table 18. Estimated and potential evapotranspiration (ET) for
ratoon sugarcane.

Estimated Potential
Month ETa ETb


August

September

October

November

December

January

February

March

April

May

June

July


aFrom statistical equation,
text)
bFrom Jones, et al. (in press


6.2

5.4

4.6

3.1

2.6

1.6

0.7

1.5

1.5

3.6

6.4

6.4

43.5

based on

is).


4.3

3.8

3.2

2.2

1.8

1.9

2.4

3.4

4.4

4.6

4.4

4.4

40.9

Shih Gascho experiments (see


4


58


I


I







Table 19. Crop acreage and return summary for S-5A in Everglades area
for 1978-79.

Harvested Harvest Water Other Total
Month acreage cost cost cost cost Profit


0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
0.6 270.8 0.0 257.3 528.1 -147.8
0.8 394.6 0.0 353.1 747.7 -116.5
112.1 26522.5 58068.1 41518.9 126109.6 -18224.9
0.3 385.3 0.0 305.3 690.7 105.9
0.3 361.4 0.0 285.7 647.1 99.7
0.5 525.9 0.0 418.3 944.3 143.4
0.5 477.4 0.0 380.1 857.4 126.7
2.5 1412.1 0.0 1288.1 2700.2 -488.0
0.1 67.9 0.0 51.4 119.3 13.6
3.8 0.0 0.0 341.2 341.2 154.3
121.5 30418.0 58068.1 45199.5 133685.6 -18333.7


59


Aug
Sept
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
S Jun
Jul
Total







CONCLUSIONS AND RECOMMENDATIONS


Version 2.2 of the area-wide water demand projection model is
truly a "first generation" model. It finds its greatest utility
in the fact that it provides an overall structure for integrating
and organizing the information necessary 1) to provide a water use
prediction and 2) to predict the primary economic impact of some
water allocation. In fact, V2.2 did integrate across the avail-
able information on agricultural water use relations in the South
Florida area. Improvements can be made as resources and new
research information become available. There were problems with
the ability to predict overall water use and drainage. These did
not appear due to any fundamental problem with the organic soil
water model. Rather, the problem appeared due primarily to a lack
of "ground truth" information regarding 1) the acreage of various
crops and when planted, 2) the costs and returns associated with
those planted acres, and 3) the water management strategies being
used under actual field conditions, at the tract level. Also, the
"POT" model for all other soils was not tested because of this
lack of information.

There was no detailed, published information of when plant-
ings actually occurred in the study area. This must be known
before a prediction capability can be implemented.

There was little information on the costs and returns asso-
ciated with particular types of crops. "Up-to-date" budgets are
needed for this type of prediction effort. Also, price prediction
models for vegetables were not available. Any water shortage
could have a substantial impact on prices, and as a result, a
substantial impact on the local economy.

The degree to which validation could be accomplished was
severely hampered by the availability of information on the nature
of the aggregates. The basic soil water models were developed at
the field level. Aggregation to the tract and area levels
requires information that is currently unavailable. Development
of an area-wide projection capability requires that resources be
allocated to such data collection. Also, further field level
experimentation is necessary to 1) better test the soil/water/-
plant relations used in the computer algorithms, and 2) obtain an
indication of yield response to varying distributions and quanti-
ties of water.









U


60







REFERENCES


Abbitt, B. and R.P. Muraro. Budgeting Costs and Returns: Central
Florida Citrus Production, 1978-79. Food and Res. Econ.
Dept. Econ. Info. Rpt. 113. Gainesville: University of
Florida, 1979.

Anaman, J.A. Optimal Irrigation Strategies for the "Valencia"
Citrus Crop in Florida. Unpublished MS Thesis. Gainesville:
University of Florida, 1981.

Alvarez, J. and F.M. Pate. The Economics of Growing Field Corn in
the Everglades Agricultural Area and of Transporting and
Feeding to Beef Cattle. Food and Res. Econ. Dept. Econ.
Info. Rpt. 102. Gainesville: University of Florida, 1978.

) Alvarez, J. and G.H. Snyder. "Effect of Prior Rice Culture on Sugar-
cane Yields in Florida." Field Crops Research 9, 314 (Nov.
1984): 305-321.

Anderson, C.L. and T.S. Hipp. Requirements and Returns for 1000-
Cow Beef Herds on Flatwood Soils in Florida. Coop. Ext.
Serv. Circ. 385, Gainesville: Institute of Food and Agricul-
tural Sciences, University of Florida, 1974.

Bean, D. Steven. Costs and Returns from Vegetable Crops in
Florida. Season 1979-80 with Comparisons. Food and Res.
Econ. Dept. Econ. Info. Rpt. 143. Gainesville: University of
Florida, 1981.

Boggess, W.G. and G.D. Lynne. "Risk-Return Assessment of Irriga-
tion Decisions in Humid Regions." Southern Journal of Agri-
cultural Economics 15(July 1983):135-144.

Brooke, D.L. Cost of Producing Sugarcane and Processing Raw Sugar
in South Florida. 1975-76. Food and Res. Econ. Dept. Econ.
Rpt. 84. Gainesville: University of Florida, 1977.

Brooke, D.L. Costs and Returns from Vegetable Crops in Florida,
Season 1977-78 with Comparisons. Food and Res. Econ. Dept.
I Econ. Info. Rpt. 110. Gainesville: University of Florida,
1979.

Brooke, D.L. Costs and Returns from Vegetable Crops in Florida,
Season 1978-79 with Comparisons. Food and Res. Econ. Dept.
Econ. Info. Rpt. 127. Gainesville: University of Florida,
1980.

Brook, D.L. and R. Clegg Hooks. Citrus Costs and Returns in
Florida, Season 1978-79 with Comparisons. Food and Res.
Econ. Dept. Econ. Info. Rpt. 138. Gainesville: University of
Florida, 1980.

Casselman, T.W. The Climate of the Belle Glade Area, Fla. Ag.
Exp. Stat. Circ. S-205. Gainesville: Institute of Food and
Ag. Sciences, University of Florida, 1970.


61








Doorenbos, J. and A.H. Kassam. Yield Response to Water. FAO
Irrigation and Drainage Paper 33. Rome: Food and Agriculture
Organization of the United Nations. Second Printing, 1981.

Florida Crop and Livestock Reporting Service. Citrus Summary,
1979. Orlando, FL: 1979.

Florida Crop and Livestock Reporting Service. Field Crops Sum-
mary, 1979. Orlando, FL: 1980.

Florida Crop and Livestock Reporting Service. Livestock Summary,
1979. Orlando, FL: 1980.

Florida Crop and Livestock Reporting Service. Vegetable Summary,
1979. Orlando, FL: 1980.

Harrison, D.S. "Water Table Fluctuations in Organic Soils of
South Florida During Periods of Drainage and Irrigation."
Proc. of the Soil and Crop Science Society of Florida,
19(1959): 353-356.

Harrison, D.A. and H.A. Weaver. "Some Drainage Characteristics of
a Cultivated Organic Soil in the Everglades." Soil and Crop
Science Society, 18(1959): 184-192.

Jones, J.W., L.H. Allen, S.F. Shih, J.S. Rogers, L.C. Hammond,
A.G. Smajstrla, and J.D. Martsolf. Estimated and Measured
Evapotranspiration for Florida Climate, Crops, and Soils.
Fla. Ag. Exp. Sta. Bul. 840. Gainesville: Institute of Food
and Ag. Sciences, University of Florida, December, 1984.

Lopez, R., J. Alvarez, and G. Kidder. Enterprise Budget for
Sugarcane Production in South Florida, 1978-79. Food and
Resource Economics Dept. Info. Report 119. Gainesville:
University of Florida, 1979.

Lucas, Robert E. Organic Soils (Histosols), Formation, Distribu-
tion, Physical and Chemical Properties and Management for
Crop Production. Ag. Exp. Sta. Res. Rpt. 435. East Lansing,
Michigan: Michigan State University, 1982.

Lynne, G.D. Area-Wide Agricultural Water Demand Projection Model
for South Florida: User's Manual, Version 2.1. Food and
Resource Econ. Info. Rpt. 194. Gainesville: Institute of
Food and Agricultural Sciences, University of Florida,
February 1984.

Marlowe, G.A., Jr. Commercial Vegetable Varieties for Florida.
Coop. Ext. Serv. Circular 530. Gainesville: Institute of
Food and Agricultural Sciences, University of Florida, 1982.

Meyer, F.W. Seepage Beneath Hoover Dike. Southern Shore of Lake
Okeechobee, Florida. Report of Investigations No. 58.
Tallahassee, FL: U.S. Geological Survey, 1971.


62

















Reitz, Herman J. Water Requirements for Citrus. IFAS Water
Resource Council WRC-4. Gainesville: University of Florida,
1977.

Rogers, J.S. and D.S. Harrison. Irrigation Water Requirements for
Agronomic Crops in Florida. IFAS Water Resources Council
WRC-5. Gainesville: University of Florida, 1977.

Rogers, J.S. and G.A. Marlowe, Jr. Water Needs of Florida Vege-
table Crops. IFAS Water Resources Council WRC-2. Gaines-
ville: University of Florida, 1976.

Shih, S.F. and G.J. Gascho. "Water Requirement for Sugarcane
Production." Transactions of the ASAE 23(1980): 934-937.

Snyder, George H., H.W. Burdine, J.R. Crockett, G.J. Gascho, D.S.
Harrison, G. Kidder, J.W. Mishoe, D.L. Myhre, F.M. Pate, and
S.F. Shih. Water Table Management for Organic Soil Conserva-
tion and Crop Production in the Florida Everglades. Fla. Ag.
Exp. Sta. Bul. 801. Gainesville: Institute of Food and
Agricultural Sciences, University of Florida, 1978.

South Florida Water Management District. 1979 Land Cover Data.
Mimeo. West Palm Beach, FL: 1981.

U.S. Bureau of the Census. Census of Agriculture. State and
County Data. Florida, Vol. 1, Part 9. Washington, D.C.:
U.S. Dept of Commerce, 1978.

Walker, C. An Analysis of Costs and Returns for a Cow-Calf Opera-
tion on the Organic Soils of South Florida. Food and
Resource Econ. Dept. Report 47. Gainesville: University of
Florida, 1973.

Weaver, H.A. and W.H. Speir. Applying Basic Soil Water Data to
Water Control Problems in Everglades Peaty Muck. Ag. Res.
Serv. ARS 41-40. Washington, D.C.: Soil and Water Conserva-
tion Research Division, U.S. Dept. of Agriculture, 1960.


63






























APPENDICES


65












Appendix A


PROGRAM FLOWCHARTS


monthnh name(i)),l=1, 12),
(((rain(i,j)),j=1,365),i=1,55),
inter


f


f


Figure A.1. MAIN Flowchart.


67













nrain,((ename(irow,karea),
(econ(irow, jcol area)
jcol=1,7),irow=1,19),karea=1<7)
nans


f


fcap rate, pfcap,
rate,pwp, mirrd,
gwt, ntract, wci,
wcd, msum


Figure A.l. MAIN Flowchart (continued).







*


68


(((cstreet(i. k)=0.O), k=1, 8), i=1, 12)
(((watsum(i, I)=0.0), 1=1,2 ), 1=1, 12)













f


t


Figure A.1. MAIN Flowchart (continued).


69


t













































Figure A.1. MAIN Flowchart (continued).


*


70


4

























/ if t
inter-2

il msoil, ef f iri
orgpci,tddr.
tdwet tiwet


if t
msoil1-



effirr=0.75


orgpcd=1.50
orgpci=0.40
tddry-6.0
tidry-6.0
tdwet-0.0
tiwet-12.0
depths-48.0



Figure A.I. MAIN Flowchart (continued),


71










4


f


t


.r
Sif t
inter=2



_name(mharv), 7060







((acres(i)=0.O) ,i:1,12)



tac, proph,mos,istartday






Figure A.I. MAIN Flowchart (continued).









4


72














f


t


allow user to assign different
values to following variables:
mirrd, fcap.pwppfcap rate.effirr


Figure A.1. MAIN Flowchart (continued).


73


allow user to assign different
values to following variables;








4


f t
msoi1--


ll potsoil call orgsoil
;:--5---- ------- all- |
call tabwat(inter)











call profit(inter)


Snans



Sif t r
nans=l1




nans=2


~op 5


Figure A.1. MAIN Flowchart (continued).

74













f


t


f t
intTe-=2


I my y-



1100 my my,

my111














Figure A.2. PARAM Flowchart.


75











4


iff t
inter=2

111



altered values for
price, avcui. avcwd, avch, -
fc yield,season




stop



Figure A.2. PARAM Flowchart (continued).








(I


76


price=Econ(mcrop, 1.narea)
avcui=Econ(mcrop, 2. area)
avcwd=Econ(mcrop3, area)


fc=Econ(mcrop 4, area)
avch=Econ(mcrop 5. area)
qield=Econ(mcrop 6, narea)
season=Econ(mcrop.7,narea)




























f 1l t
itemp month



to of days = i = 1 montl
Luday lday + 1 no ofdays




1 month=

110
if
1 _month=
iu month














1+ no
i = month




f if t
i <=iu month
-1

i=i+1
no of days = max
10 + no



Figure A.3. CALJUL Flowchart.


77











4


f if t
11 month 1

Sistart days = istartdays +
1 day

if
1 <= 11 f
month-i
Sif t
iu month
=1

days istartdays +
Limit(i)
L + i iend days = iend Days +
i uday









t if f
i -1


L+1
days = ienddays = max-lmitt(i)


Sendchat



Figure A.3. CALJUL Flowchart (continued).
I


78




























































Figure A.4. JULCAL Flowchart.








79




ure A.5. WTPRED Flowchart


f


t


iwinf=(winf+00.0005)*10000
winf=iwinf/1000
i=0


f


t


x = x-1.0
winest=0.06662402*x+0.00256465*
x**2
iwinest+(winest+0.0005)*1000
winest=iwinest/1000


f


80


t












Figure A.5 (continued).


100




x=x+O.1
winest=0.06662402*x +
0.00256465*x**2
iwinest=(winest+0.0005)*1000
winest=iwinest/1000


f


x=x-0.01
winest0=.06662402*x +
0.00256465*x**2
iwinest=(winest+0.0005)*1000


81











4


t


4


82


f






Figure A.6. ORGSOIL Flowchart



start



J=1



f if t
<=365


fldw(j)
upump(j
J=1 uwpumpd(
uin(j)=
set(j)-
twin(j)
J=J+l


f if t
j<=366

tsw(J)
|but(idays)=gwt bwt(j)



f t
depths<= t
12. O
stara'


f t i


83






Figure A.6. (continued)


100



tsw(idays)=swstar-0.06662402)*gwt+0.00256465
*gwt**2
j=idays




f i t
j<=idaye

600
6if
f J<=92 or, t
1j<=273

tdwt-gwt-tddry tdwt=gwt-tdwet
tiwt=gwt+tidry r tiwt=gwt+tiwet



f tif t
fldw(j)>
0.0


z=ename(mcrop,narea)9007





f if 2= ^ t
'sugarcane'




2000

if
z=pasture or
z=pasture (cow/calf) or
f z=pasture (dairy) or z=turf or t
z=golf course or z=oranges or
z=grapefruit or
z limes


84

















Figure A.6. (continued)


dvol=0. 06662402*bwt( J+1) +
0. 00256465*bwt ( J+1)**2 -
release


t


85










4


Figure A.6. (continued)


f


86















Figure A.6. (continued)


f


f


87


t










4


Figure A.6. (continued)








2100



pergs=flowat(-idayp1 )/season


f if t
pergs<=0.20



f if t cropk=0. 40
pergs<=
1. 0
cropk=-0. 8701+8. 0757*pergs-10.5759
*opk=0.40 *pergs**2+3.9775*pergs**3




6600













(I


88









Figure A.6. (continued)


89









4


Figure A.6. (continued)





6000


k=j






fldw (j ) =fldw(k-l)+win(j)

Lif t
: ::: fldw(j)>0.0


fwin=fldw(j)
xwt=0.0 6001
wtpred(fwin,xwt,j)
bst(j+l)=xwt
fldwj=0.0

4000


90














Figure A.6. (continued)


f


bwt(J+l)=xwt
wpumpd(j) fldwj
fldw(j)- 0.0


bwt(j+l)-gwt
wpumpd(j)- release + fldw(j)
fldw(j)m 0.0


91


t










4


Figure A.6. (continued)






8502










dvol= release (0. 06662402*bwt(j+1) +
0. 00256464*bwt(j+1 )**2)



f t
dvol>=orgpcd


pumpd(J)i dvol wpumpd(j)> orgpcd
win- -dvol w xwin- -orgpcd


wtpred(xwin xwt J)
bwt(j+1)- xwt




4000









*


92




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs