PB84-155217
On-Farm Improvements to Reduce Sediment and Nutrients in Irrigation
Return Flow
Washington State University
Pullman, WA
Feb 84
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
HHS
-------
EPA-600/2-84-044
February 1984
ON-FARM IMPROVEMENTS TO REDUCE SEDIMENT
AND NUTRIENTS IN IRRIGATION RETURN FLOW
By
L. G. King
B. L. McNeal
F. A. Ziari
S. C. Matulich
Washington State University
Pullman, WA 99164-6120
Grant No. R805527
Project Officer
James P. Law, Jr.
Irrigated Agriculture Section
Robert S. Kerr Environmental Research Laboratory
Ada, OK 74820
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OK 74820
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-84-044
3. RECIPIENT'S ACCESSION-NO.
- 5 5 9 1 "?
II _ *• •"
4. TITLE AND SUBTITLE
ON-FARM IMPROVEMENTS TO REDUCE
IN IRRIGATION RETURN FLOW
SEDIMENT AND NUTRIENTS
5. REPORT DATE
February 1984
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
L.G. King, B.L.
S.C. Matulich
McNeal, F.A. Ziari, and
8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Departments of Agricultural Engineering and
Agronomy, Washington State University,
PutIman, WA 99164-6120
10. PROGRAM ELEMENT NO.
C34BIB
11. CONTRACT/GRANT NO.
R-805527
12. SPONSORING AGENCY NAME AND ADDRESS
Robert S. Kerr Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Ada, OK 74820
13. TYPE OF REPORT AND PERIOD COVERED
FINAL
14. SPONSORING AGENCY CODE
EPA/600/15
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Research on an 800-hectare irrigated tract in central Washington's Columbia
Basin Project studied the effects of on-farm improvements on reduction of the
discharge of sediment and nutrients via irrigation return flow. Technical assistance
and financial cost-sharing were provided to participating farmers. Physical
improvements included pipes to convey tail water to improved drains, sediment basins
and mini-basins, concrete lining of head ditches, gated pipe systems, and conversion
of furrow irrigated land to sprinkler systems (center pivot and solid set).
Structural improvements greatly reduced the sediment discharge from the overall
tract. Measures which controlled sediment loss were not equally effective in
reducing phosphorus loss, even though reductions in phosphorus loss were significant.
Project activities had little effect on the discharge of nitrogen during the period
of study. Results presented in the report address the problems of sediment and
nutrient discharges on three basic levels: individual furrows, individual fields,
and the total 800-ha study area. Models are presented which deal with sediment loss,
nutrient loss, and economic motivation for BMP adoption.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Irrigation, Sediment, Nutrients, Water
Quality, Water Pollution
Irrigation Return Flows
Nonpoint Sources
Columbia River Basin
Project, Washington
Best Management Practice
43F
48B
68D
98C
13. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
Unclassi f ied
21. NO. OF PAGES
20?
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
DISCLAIMER
Although the research described in this report has been funded in part
by the U.S. Environmental Protection Agency through Grant No. R-805527 to
Washington State University, it has not been subjected to the Agency's
required peer and administrative review and therefore does not necessarily
reflect the views of the Agency and no official endorsement should be
inferred.
-------
FOREWORD
EPA is charged by Congress to protect the Nation's land, air and water
systems. Under a mandate of national environmental laws focused on air and
water quality, solid waste management, and the control of toxic substances,
pesticides, noise, and radiation, the Agency strives to formulate and imple-
ment actions which lead to a compatible balance between human activities
and the ability of natural systems to support and nurture life. In partial
response to these mandates, the Robert S. Kerr Environmental Research
Laboratory, Ada, Oklahoma, is charged with the mission to manage research
programs to investigate the nature, transport, fate and management of
pollutants in ground water and to develop and demonstrate technologies for
treating wastewaters with soils and other natural systems; for controlling
pollution from irrigated crop and animal production agricultural activities;
for developing and demonstrating cost-effective land treatment systems for
the environmentally safe disposal of solid and hazardous wastes.
This report contains the major findings and recommendations for the
implementation of best management practices to control soil erosion and
sediment emanating from irrigated agriculture in the Columbia Basin Project
area of the Pacific Northwest region. The study results presented address
the problems of sediment and nutrient discharges on three basic levels:
individual furrows, individual fields, and the total 800-hectare study
area. Models are presented which deal with soil loss, nutrient loss, and
economic considerations for BMP adoption. This report should benefit
managers as they identify and implement pollution control strategies
relevant to western irrigated agriculture.
Clinton W. Hall, Director
Robert S. Kerr Environmental
Research Laboratory
-------
ABSTRACT
Research on an 800-hectare irrigated tract consisting of a hydrologic
sub-basin with well-defined surface drainage covered five complete
irrigation seasons (1977-81). This predominantly furrow-irrigated area
was located within central Washington's Columbia Basin Block 86. The
cooperative research project studied the effects of on-farm improvements
on reduction of the discharge of sediment and nutrients (nitrogen and
phosphorus) from the tract via irrigation return flow. Both technical
and financial (cost-sharing of construction) help were given to the
particip'ati'hg'farmers. Between the 1978 and 1979 irrigaton seasons,
facilities.-.on various farms were constructed with the grant providing 70
percent (up to .,a pre-determined maximum amount) and the farmer 30 percent
(plus any excess above the pre-determined maximum) of the costs. A goal
of about $125 per hectare benefited was set as a maximum cost share to be
provided by research grant funds. The constructed facilities included
pipes to convey center pivot overflow and furrow tail water to improved
drains, sediment basins, sediment mini-basins, concrete lining of head
ditches, gated pipe systems, and conversion of furrow-irrigated land to
sprinkler (both center pivot and solid set). Approximately $70,000 of
grant funds were spent on cost-sharing of construction.
Results showed that construction of proper sediment control facilities
on furrow-irrigated farms greatly reduced the discharge of sediment from
the overall tract. The 3-year average sediment discharge from the area
after construction of on-farm improvements was about twenty percent of the
2-year average discharge before construction. The irrigation return flows
decreased only about three percent following construction.
While reductions in phosphorus loss were significant, results showed
that measures which controlled sediment loss were not equally effective in
controlling phosphorus loss. The 3-year average phosphorus discharge after
construction was 51 percent of the 2-year average before construction. The
difference in effectiveness of control measures for sediment and phosphorus
was attributed to the association of phosphorus with clay-sized sediment
particles which are not easily settled once they become suspended in
irrigation tailwater. End-of-field sediment/phosphorus ratios were often
about 1,500 whereas these ratios for water in the main drain leaving the
entire study area were only about half this value.
The project activities had little effect upon the discharge of
nitrogen from this irrigated tract during the period of study. A
considerable part of the area was served by subsurface drainage systems,
which discharged into the main surface drain through the area. The water
from surface and subsurface sources was comingled in the main drain as it
IV
-------
left the irrigated tract. Discharge of nitrogen was about 20 to 30 kg per
hectare per year during the study period.
Techniques were devised to reasonably separate the effects of ashfall
deposited into the study area by the May 18, 1980, eruption of Mount St.
Helens from the effects of research project activities related to sediment
and phosphorus discharges. The effects of the ashfall were mainly confined
to a two-week period immediately following the eruption.
Results presented in this report address the problems of sediment and
nutrient discharges on three basic levels: individual furrows, individual
fields, and the total 800-hectare study area. Models are presented which
deal with sediment loss, nitrogen loss and economic motivation for BMP
adoption.
This report was submitted in fulfillment of Grant No. R805527 by
Washington State University under the partial sponsorship of the U.S.
Environmental Protection Agency. This report covers the period from
April 1, 1977 to June 30, 1982 and the work was completed as of
June 30, 1982.
-------
CONTENTS
Foreword iii
Abstract iv
Figures viii
t
Tables xi
Acknowledgements xiv
1. Introduction 1
2. Conclusions 6
3. Recommendations 9
4. Background 10
5. Study Area Description 17
6. Methods of Data Collection and Analysis 26
7. Structural Changes 41
8. Results and Discussion 46
Literature Cited 138
Appendices
A. Individual-Furrow Data 149
B. Individual-Furrow Sediment Loss Model 157
C. Nitrogen Leaching Model 166
D. Procedures for Sediment Analysis of Irrigation
Runoff Samples 191
vn
-------
FIGURES
Number Page
1 Location of the study area in Washington State 17
2 Study area farm units, flume sites, and pipe locations ..... 18
3 Monitored fields within the study area, 1977-1981 19
4 Outflow sampling device 34
5 Average weekly runoff and net sediment loss for the study
area, 1977-1981 v . . . 52
6 Net sediment and phosphorus losses from the study area,
1977-1981 54
7 Seasonal distribution of main drain discharge (runoff) and
area-wide net sediment losses, 1978 56
8 Seasonal distribution of main drain discharge (runoff) and
area-wide net sediment losses, 1979 57
9 Seasonal distribution of main drain discharge (runoff) and
area-wide net sediment losses, 1980 58
10 Seasonal distribution of main drain discharge (runoff) and
area-wide net sediment losses, 1981 59
11 Net sediment and phosphorus main-drain discharges,
spring 1980 60
12 Hydrographic record of changing runoff rate and sediment
concentrations 73
13 Sediment discharge hydrograph, flume 3, July 23, 1980 74
14 Seasonal trend for sediment loss, farm unit 53, fields
C and D, 1980 '. 75
15 Three-dimensional representation of seasonal and daily
sediment loss 76
16 Sediment basin locations in the study area, 1981 77
VI 11
-------
FIGURES (Continued)
Number Page
17 Seasonal operation of sediment basin No. 2, Block 86, 1981 . . 79
18 Seasonal operation of sediment basin No. 3, Block 86, 1981 . . 80
19 Furrow intake and runoff rate relationships, late season,
1979 90
20 Sediment discharge hydrographs for field 490, 1981 93
21 Sediment loss from farm unit 64 after cultivation, 1981 .... 94
22 Sediment losses from cutback and non-cutback irrigation
systems 97
23 Surface runoff in relation to time of cutback for a single
cutback system 101
24 The effect of percent runoff and number of irrigations on
sediment loss from a new carrot field 102
25 Sediment loss as a function of runoff percentage for several
crop types 104
26 Sediment loss from a field of beans as a function of runoff
percentage with time since cultivation 105
27 Sediment loss as a function of runoff percentage for two bean
fields of differing slope 107
28 Advance curves for ARIZONA TEST Ill
29 Advance curves for WASHINGTON test 112
30 Sediment discharge rates for WASHINGTON test 113
31 Correlation of suspended sediment with settleable solids as
measured with the Imhoff cone for tailwater from furrow-
irrigated fields in the study area, 1978 116
32 Nitrogen leaching as a function of drainage for continuously
furrow-irrigated sand, silt loam, and clay 120
33 Spatial distribution of water and nitrogen leaching below the
68 cm depth during continuous every-furrow irrigation of
sand, silt loam, and clay soils - . 122
IX
-------
FIGURES (Continued)
Number Page
34 Nitrogen leaching as a function of drainage for continuous
every-furrow irrigation of a silt loam with various
positions of fertilizer placement . 123
35 Nitrogen leaching as a function of drainage for continuous
every-furrow and every-other-furrow irrigation of a silt
loam with the fertilizer placed 18 cm deep 124
36 Nitrogen distribution for 4 days of every-other-furrow
irrigation for a silt loam soil 126
37 Nitrogen distribution for 4 days of alternating-furrow
irrigation for a silt loam soil 127
38 Nitrogen leaching as a function of drainage for furrow-
irrigated silt loam 128
39 Nitrogen leaching as a function of drainage for furrow-
irrigated sand and silt loam soils with plant uptake .... 130
-------
TABLES
Number Page
1 Study area crop history, 1977-1981 21
2 Planting and harvesting dates, Block 86 study area 22
3 Annual percentage of sprinkler and surface-irrigated lands
in the Block 86 study area 23
4 Summary of irrigation season weather data for the Block 86
study area, 1979-1981 25
5 Details of facilities constructed on participating farmer's
land 42
6 Details of sediment basins in operation during project period. . 43
7 Monthly mass balance for sediment and total phosphorus, 1977
irrigation season 47
8 Monthly mass balance for sediment and total phosphorus, 1978
irrigation season 48
9 Monthly mass balance for sediment and total phosphorus, 1979
irrigation season 49
10 Monthly mass balance for sediment and total phosphorus, 1980
irrigation season 50
11 Monthly mass balance for sediment and total phosphorus, 1981
irrigation season 51
12 Relative changes in net sediment and phosphorus losses from
1977-1978 to 1979-1981 53
13 Study area irrigation diversion summary, 1977-1981 55
14 Study area net sediment and phosphorus losses, spring 1980 ... 61
15 Sediment/phosphorus ratios in the study area runoff as
affected by the Mt. St. Helens ashfall 62
16 Comparison of sediment and phosphorus loss estimates 63
-------
TABLES (Continued)
Number Page
17 Sediment loss estimates for individual fields, 1978 64
18 Sediment and phosphorus loss estimates for individual
fields, 1979 65
19 Sediment and phosphorus loss estimates for individual
fields, 1980 66
20 Sediment and phosphorus loss estimates for individual
fields, 1981 67
21 Field slope and sediment loss, 1979 68
22' Field slope and sediment loss, 1980 69
23 Irrigation summary, flumes 4 and 19, 1979 69
24 Irrigation summary, flume 17, 1980 70
25 Field comparisons, 1978-1981 71
26 Overall seasonal sediment basin retention performance,
1977-1981 78
27 Sediment basin performance, basin No. 2, 1981 81
28 Sediment basin performance, basin No. 3, 1981 83
29 Linear regression parameters of phosphorus on area-wide
sediment losses 84
30 Phosphorus losses for monitored fields 85
31 Infiltration at 24 hours as a function of stream size 87
32 Infiltration and runoff for freshly cultivated vs. stabilized
furrows 88
33 Runoff and sediment loss relationships of some dry and
previously moist soils 91
34 Summary of soil loss results 95
35 Summary of irrigation scheduling model performance for the
irrigation season 96
36 Sediment loss and irrigation efficiency from cutback and
non-cutback furrows of field 47A (beans), during the
fourth irrigation 98
xii
-------
TABLES (Continued)
Number Page
37 Total runoff in relation to time of cutback 100
38 Particulars of comparison tests 110
39 Distribution of irrigated land in Washington1s major
irrigation districts 115
• 40 Hydraulic properties of the soils used for the simulations . . . 119
41 Nitrogen balance after three weeks of irrigating a mature
potato crop 131
i,
42 Before- and after-tax cost effectiveness of alternative
irrigation systems and crop rotations for 64.75 ha farms
by equity level 133
43 Before- and after-tax cost effectiveness of alternative
irrigation systems and crop rotations for 259 ha farms
by equity level 134
-------
ACKNOWLEDGEMENTS
Research on an 800-hectare irrigated tract in central Washington's
Columbia Basin Block 86 was initiated in April 1977 and continued through
the end of the 1981 irrigation season. The total research effort was
funded by several sources including the U.S. Environmental Protection
Agency (Grant No. R-805527), the Washington State University Agricultural
Research Center (Project No. 0402 and 0306), the Washington State
Department of Ecology (Contract No. WF-PS-78-003), the U.S. Geological
Survey, the U.S. Bureau of Reclamation, the Quincy Columbia Basin
Irrigation District, the East Columbia Basin Irrigation District, and the
.South Columbia Basin Irrigation District. Additional funding in the period
following the May 1980 eruption of Mount St. Helens was provided by the
U.S. Department of Agriculture (Special Grant No. 59-2531-1-2-008-0), the
Soil Conservation Service (Agreement No. 59-0546-1-20), and the State of
Washington Water Research Center (Grant No. 14-34-0001-1459). Special
thanks are expressed to the sixteen farmers within the study area who
allowed the research to be conducted on their land. Appreciation is also
given to W. H. Pietsch and E. P. Smith who participated in the early stages
of the research project. The authors express thanks to the many graduate
students who worked on various phases of the research.
xiv
-------
SECTION 1
INTRODUCTION
DEVELOPMENT OF PROJECT
Public Law 92-500, also known as the "Federal Water Pollution Control
Act Amendments of 1972", provided for issuance of permits for discharge of
pollutants into the nation's waters. Irrigation return flow (i.e., water
which returns to streams and rivers after being diverted and applied to the
land as irrigation) can carry many substances classified as pollutants. In
the Pacific Northwest, some of the major pollutants in irrigation return
flow have been identified as sediment, nitrates, nematodes, phosphorus
(attached to sediment), bacteria, and increased temperature (Skogerboe and
Law, 1971). In regulations promulgated pursuant to PL 92-500 much of the
irrigation return flow was classified as point sources. For a point source
discharge, a "National Pollutant Discharge Elimination System" (NPDES)
permit was required by PL 92-500.
The State of Washington Department of Ecology (DOE) was authorized by
the U.S. Environmental Protection Agency (EPA) pursuant to PL 92-500 to
issue NPDES permits for discharges into the waters within the jurisdiction
of the State. By late 1974, the DOE was well underway with the development
of a permit program. During the development of this program, it became
apparent that the greatest reduction of total sediment in irrigation return
flow would be realized by change of practices on the individual farms.
Once the return water reaches a common drain, opportunities for sediment
removal are considerably restricted.
Regulations (EPA, 1973) promulgated by EPA stated that NPDES permits
were required for discharges of irrigation return flow (such as tailwater,
tile drainage, surfaced groundwater flow, or bypass water) if there is a
point source of discharge (e.g., a pipe, ditch, or other defined or discrete
conveyance, whether natural or artificial) and the return flow is from land
areas of more than 1,214 contiguous hectares (3,000 acres), or 1,214
non-contiguous hectares which use the same drainage system.
During the time in which DOE was developing permit requirements for
irrigation return flow, the 1,214-hectare exclusion was contested in the
courts, with suit brought by the National Resources Defense Council, Inc.
(NRDC) against Russell E. Train, Administrator of the Environmental
Protection Agency, et al. On March 24, 1975, the court ruled in favor of
NRDC, (United States District Court, 1975a), i.e., that PL 92-500 did not
allow the Administrator, EPA, the latitude to exempt entire classes of
-------
point sources from the NPDES permit requirements. The effect of this
ruling was that irrigated lands of less than 1,214 hectares would not be
excluded from requiring NPDES permits. In the final judgement (United
States District Court, 1975b) filed June 10, 1975, the court ordered that
EPA within eight months publish proposed regulations and within twelve
months promulgate final regulations extending the NPDES permit system to
include all point sources of irrigation return flow. EPA appealed this
decision. However, to comply with the court order, EPA promulgated final
regulations for irrigation return flow on July 12, 1976. These regulations
eliminated the 1,214-hectare exclusion but at the same time recognized that
the NPDES permit program as then being administered was inappropriate for
dealing with the vast numbers of irrigation return flow point sources. The
regulations proposed to handle these discharges under a general permit,
details of which were to follow at a later date. Point sources of
irrigation return flow were redefined as conveyances carrying surface
irrigation return flow as a result of the controlled application of water
by any person on land used primarily for crops.
On February 4, 1977, EPA published proposed rules for the general
permit program (EPA, 1977). These proposed rules recognized that pollution
reduction from irrigation return flow is often most effectively achieved by
using best management practices (BMPs). BMPs are management techniques
that reduce the amount of pollutants at their source rather than remove
them from their point of discharge into receiving bodies of water. For
irrigation return flow, BMPs are applied to the individual farms.
The proposed rules also strongly linked the general permit program to
the 208 planning process. It was suggested that the selection and
implementation of BMPs are best achieved locally in coordination with 208
planning agencies. The EPA also stated that there was much research yet to
be completed to develop and demonstrate the effectiveness of alternative
management practices in all locations.
In 1977 the U.S. Congress passed PL 95-217 which expressly placed
irrigation return flow under section 208 and removed it from section 402,
NPDES permits, of PL 92-500. The planning activities under section 208 in
Washington resulted in a program of voluntary farmer participation to
reduce sediment carried by irrigation return flows.
Earlier (in 1974) because of the court actions and impending
decisions, the DOE (in the State of Washington) did not issue any of the
irrigation return flow permits which were then being prepared. Instead a
cooperative program to improve farming practices in order to reduce the
sediment delivered to the Yakima River via irrigation return flow was
begun. This program identified the Sulphur Creek drainage near Sunnyside,
Washington, as being a problem area which should receive high priority.
In 1975, the Sulphur Creek Demonstration Project (King et al., 1977;
King and Janke, 1983) was started. By the end of the 1976 irrigation
season, it was evident that this site had many characteristics making it
unsuitable for obtaining the desired information to fully evaluate the
benefits of on-farm changes in reducing sediment discharge to the Yakima
-------
River via irrigation return flow. What was needed was a smaller, more
compact, drainage with well-defined hydrologic boundaries having fewer
sources of inflow and points of discharge. Farmer participation had not
been as great in the Sulphur Creek project as had been anticipated and
those farmers who were participating were widely scattered throughout the
drainage. Incentives for participation apparently had not been large
enough.
Selection of a site for the work reported herein began in late 1976.
It was decided that the greatest benefit to the State could be achieved by
locating the study area within the Columbia Basin rather than the Yakima
Valley. A site within the Columbia Basin would allow the project to be
separated from the Sulphur Creek activities. By such separation the public
would not likey be confused by the different approaches of the two studies.
Criteria for selection of this site were the following:
A. Primarily surface irrigated land
B. Existing sediment problems in irrigation return flow
C. Single discharge point for return flow
D. Approximately 800 hectares
E. Well defined drainage system for surface flows
F. Several farmers
G. Range of crops
H. Range of slopes
I. Soils typical of much of the Columbia Basin
J. Indications of farmers' willingness to participate in
the study.
In the fall of 1976, the principal investigator of the research
reported herein met with the three Columbia Basin irrigation districts,
(Quincy Columbia Basin Irrigation District (QCBID), East Columbia Basin
Irrigation District (ECBID), and South Columbia Basin Irrigation District
(SCBID)) and with the help of the board of directors and district personnel
selected two alternate sites. Careful review of the sites indicated that
one site was clearly superior to the other in light of the foregoing
criteria. In early 1977, the study site was selected in Block 86 on the
Royal Slope west of Othello, Washington. The site lies within the QCBID
and Grant County.
The Geological Survey (USGS), the Bureau of Reclamation (USER), and
the three Columbia Basin irrigation districts agreed to monitor the
irrigation supply and drainage of the study area during the 1977 irrigation
season to obtain background data prior to the start of the research
project. This monitoring continued throughout the duration of the project
under joint funding by DOE, USGS and USER.
On March 8, 1977, a meeting was held in Royal at the office of the
QCBID to explain the planned program to the farmers in the study area and
to assess their willingness to participate. In addition to the local
farmers, personnel from the following agencies attended the meeting:
QCBID, ECBID, DOE, WSU, USGS, conservation districts, and the Othello Water
-------
Quality Committee (Section 208 Planning). The response of the farmers was
unanimous in support of the project.
In the project, help (both technical and financial) was given to the
individual farmers for on-farm improvements to reduce sediment and
nutrients leaving their farms in irrigation return flow. The first year of
the project (the 1978 irrigation season) was spent analyzing the operation
of all the farms in the study area, gathering data needed before changes
were made. Design proceeded as soon as possible to allow for construction
of necessary structures prior to the start of the 1979 irrigation season.
Since farmers were to be provided with financial help in construction
of facilities, participation was expected to be high. The goal was to have
all farmers in the study area participating in the project. The response
of the farmers to the project at the March 8, 1977, information meeting
indicated that this goal was attainable. The on-farm facilities were
constructed under a cost-sharing arrangement. The research grant provided
70 percent .of the cost of capital improvements up to a maximum amount. The
farmer provided, the other 30 percent and all costs exceeding the amount
agreed upon prior to the start of construction. A goal of about $125 per
hectare benefited was set as a maximum cost share to be provided by the
research grant funds.
OBJECTIVES
The overall goal of this project was to assist in developing and
implementing a program for reducing the negative impacts of irrigation
return flows on water quality. It was recognized that reduction of
pollution from these sources is most effectively achieved by improvement of
farming practices on individual farms. It was anticipated that the farming
practices to be studied under this project would include the best
management practices (BMPs) then being developed by the local water quality
committee involved in the 208 planning process for irrigated agriculture.
Implementation of the 208 plans for irrigated agriculture was a significant
component of this research project.
A second general goal of this project was to select a study area in
which a high degree of farmer participation and significant changes of
farming practices could be achieved.
Because of the importance of this work to the local area and to the
state, another goal was the timely dissemination of information via county
extension agents, periodic meetings with the three Columbia Basin
irrigation districts and conservation districts, and annual field days at
the study site.
The specific objectives of this proposed research project were:
A. To perform the research in a study area in which a high degree of
farmer participation and significant changes of farming practices
can be achieved.
-------
B. To accomplish changes in farming practices within the study area
which will eliminate or reduce sediment and nutrient discharge
from irrigated farms.
C. To quantify the reduction achieved by particular changes of
practice.
D. To identify the best farming practices on a cost-effectiveness
basis for controlling sediment and nutrient discharges from
irrigated farms.
E. To determine the effectiveness of various arrangements for
bringing about changes in practices.
G. To disseminate information concerning the project by:
(1) Periodic reports of plans and progress to the county
extension agents;
(2) Periodic meetings with irrigation districts and
conservation districts;
(3) Regular reporting of plans and results to the State of
Washington Department of Ecology (DOE) through advisory
committees and/or other means;
(4) Periodic reports of progress to U.S. Environmental
Protection Agency (EPA), Robert S. Kerr Environmental
Research Laboratory, Ada, Oklahoma, and Region X,
Seattle, Washington;
(5) Annual on-site field days.
-------
SECTION 2
CONCLUSIONS
The results of this research project show that effective reduction of
sediment discharge from individual fields and from the total study area was
accomplished by construction of on-farm facilities. The 3-year average
sediment discharge from the area after construction of on-farm improvements
was about twenty percent of the 2-year average discharge before construc-
tion. The irrigation return flows decreased only about three percent
following construction.
Funds provided for construction of on-farm facilities averaged less
than ninety dollars per hectare benefited. It is extremely difficult to
assess the value of total project activities and visibility of professional
people in accomplishing the observed reduction of sediment discharge.
Expenditures of similar capital construction assistance in different areas
in absence of these other activities may or may not produce the same effects.
Sediment basins were the most successful of measures used in this study
for reducing sediment discharge from irrigated fields. Sediment basins
constructed by farmers without technical assistance usually do not have
sufficient capacity to trap sediment for a complete irrigation season. In
1981, the sediment basins removed from 53 to 85 percent of all incoming
sediments with an average of 66 percent. When a single undersized basin
was used to attempt to retain sediment in tailwater from several farm
units, the basin filled completely after only two irrigations.
Certain problem situations, such as steep tailwater collection ditches
and lack of a conveyance channel adequate for sprinkler pond overflows,
were successfully corrected in the study area. Check dams were used in the
steep ditches and the overflow from center pivots was piped to an improved
open drain.
Reduction of sediment discharge should not be equilibrated with
reduction of erosion. Properly designed and maintained sediment basins
will reduce sediment discharge while on-field erosion may continue
unchecked. On one particular field in the study area, the soil surface
elevation at the head of the field has dropped nearly one meter since
irrigation began in the 1950's.
Reduction of sediment discharge does not necessarily accomplish
reduction of phosphorus discharge. Once the soil particles have been
suspended in the water, normal settling will be more effective in removing
sediment than phosphorus because of the association of phosphorus with
-------
clay-sized sediment particles which are not easily settled once they become
suspended in irrigation tailwater. The 3-year average phosphorus discharge
from the total study area after construction of on-farm improvements was
51 percent of the 2-year average before construction. End-of-field
sediment/phosphorus ratios were often about 1,500 whereas these ratios for
water in the main drain leaving the entire study area were only about half
this value.
Nitrogen discharge from an irrigated area is not subject to effective
control by practices used for this research project. Modeling results
indicated that considerable time may be required to observe any change in
nitrogen discharge as a result of a practice change. Measured nitrogen
discharge from the total study area was not affected by project activities
during the period of study. A considerable part of the area was served by
subsurface drainage systems which discharged into the main surface drain
through the area. The water from surface and subsurface sources was
comingled in the main drain as it left the irrigated tract. Discharge of
nitrogen was about 20 to 30 kg per hectare per year during the study
period.
The ashfall deposited onto the study area by the May 18, 1980,
eruption of Mount St. Helens did not invalidate the research findings.
Techniques were developed and demonstrated to be successful for separating
the effects of the ashfall in contrast to the effects of the research
project activities upon the discharge of sediment and phosphorus from the
area. The effects of the ashfall were mainly confined to a two-week period
immediately following the eruption.
The crop grown on a particular field has a significant effect on the
sediment loss. Row crops such as sugar beets, beans, and corn produce
much more sediment than close-growing crops such as wheat.
Scheduling of irrigation has an effect on the seasonal sediment loss
from a field. Studies showed that reducing the number of irrigations on
beans could reduce the total sediment loss for the season without lowering
the yield.
Use of methods to control the stream size in individual furrows can
reduce sediment in tailwater. There is a definite need for automated or
semi-automated furrow irrigation systems. Cutback irrigation practices
are effective in reducing sediment, but are not very popular with farmers
because of large labor requirements with irrigation systems presently on
the farms.
The Imhoff cone, using a 15-minute settling time, was tested as a
device for measuring suspended sediment in irrigation tailwater. For
sandy loam and loamy sand soil textures, the Imhoff cone reading correlated
well with suspended sediment concentration measured with standard methods.
These two textures account for approximately 31 percent of the land in
Washington which is furrow irrigated. Another 62 percent of furrow-
irrigated land is silt loam and loam. Use of the Imhoff cone may be
acceptable for these textures. Perhaps the best use of the Imhoff cone for
-------
irrigated agriculture would be as a tool for comparing the relative effec-
tiveness of various practices in reducing sediment loss from a given field.
Such use should be beneficial to a farmer, conservation district personnel,
or irrigation district personnel as an evaluation aid. The Imhoff cone
should not be used as a regulatory standard.
Working models were developed to describe the nitrogen movement and
loss from furrow-irrigated land and to describe the discharge of sediment
from individual irrigation furrows. The models are researchers' tools and
are not yet ready for more widespread use.
Economic modeling demonstrated the importance of tax considerations in
motivating BPM adoption. Immediate attention should be given to existing
institutions and programs that provide some incentive to adopt pollution
abatement technologies. A program of variable incentives depending upon
farm size and debt/equity position would be the most efficient expenditure
of funds to produce the adoption of BMP's.
-------
SECTION 3
RECOMMENDATIONS
Incentive programs must recognize the need to reduce the on-field
erosion as well as sediment discharge for furrow-irrigated land. Sediment
basins should be used in conjunction with improved water management. A
full program of farmer assistance is necessary to obtain use of proper
water management primarily including appropriate furrow stream size,
irrigation set time, and length of run. Development of systems for
automation or semi-automation of furrow-irrigation will greatly assist
efforts to obtain adoption of the changes in on-farm water management
needed for effective control of on-field erosion and of sediment and
phosphorus discharge. Development of these systems should receive federal
and state support.
Incentive programs must also address the after-tax determination of
cost effectiveness of control measures. The findings of this research
support a program of variable incentives depending upon farm size and
debt/equity position of the land owner.
Work should continue on model development of sediment discharge from
individual furrows. This model should be used to study effects of better
water management on field-wide and area-wide sediment losses. The model
should be further developed to handle erosion and deposition of sediment
along the furrow.
Farmers should be required to pipe center pivot sprinkler overflows
to an acceptable improved drain, especially for new installation of center
pivots. Steep tailwater ditches should be piped or have check dams
installed. Technical assistance should be provided to farmers for proper
sizing and design of sediment basins.
-------
SECTION 4
BACKGROUND
IRRIGATED AGRICULTURE AND WATER QUALITY
Irrigated agriculture represents this nations's single largest
consumptive use of water (Law and Skogerboe, 1972). However, irrigated
agriculture is of critical importance to our nation's agricultural economy.
The 10% of the U.S. farmland which is irrigated produces fully one-fourth
of the country's agricultural output (Federal Water Pollution Control
Administration, 1968).
Not all irrigation water is consumed, however. Law and Skogerboe
(1972) considered irrigation return flows to be comprised of four parts:
canal seepage, bypass water, deep percolation, and tailwater. Only the
last two are of consequence once water has been applied to a field.
Irrigation return flows affect downstream water quality primarily in
two ways: by introduction of sediments eroded from irrigation furrows, and
by contribution of nutrients (chiefly nitrogen and phosphorus) or salts to
downstream ponds, rivers, and tributaries. Salt effects have not been
considered in this study, because of the infrequent occurrence of
water-induced salinity problems in irrigated areas of the Pacific
Northwest.
q
Wadleigh (1968) estimated that 3.6 x 10 metric tons of sediment
(about half of this from agricultural lands) enters U.S. rivers and
waterways each year. He estimated that each metric ton of sediment
included about 0.60 kilograms of phosphorus. Larger sediment particles
settle out and fill reservoirs, streams, lakes, and tributaries. Over the
long term, eroded sediment lowers water quality for subsequent users,
degrades or destroys aquatic habitats, reduces photosynthesis through
increased turbidity, and acts as a medium of transport for a number of
potentially toxic pollutants (Robinson, 1971). Since agricultural land is
a valuable national resource, its loss or depletion through erosion is also
of serious long-term concern.
IRRIGATION AND SOIL EROSION
Carter (1976) proposed eight recommendations for reducing
irrigation-associated sediment losses:
1. Eliminate or reduce irrigation return flows.
10
-------
2. Control slope of irrigated lands via contour furrowing
or grading.
3. Control furrow stream size.
4. Shorten run length.
5. Shorten irrigation duration.
6. Limit cultivations.
7. Control tailwater via collection ditches or buffer
strips.
8. Use sediment basins to trap sediments.
Possibilities for controlling agricultural water quality have been
mentioned by others as well (Bondurant, 1971; Law and Skogerboe, 1972;
Wells, 1974; Carter and Bondurant, 1976). Among their suggestions were:
changes in prevailing cultural practices, irrigation scheduling, use of
drip irrigation where feasible, and tailwater recovery and/or re-use.
Despite wide-spread and rapid conversion to sprinkler or drip systems,
furrow (rill) irrigation remains the most widely practiced means of
irrigation in the western U.S. Under such a system, stream velocity and
volume in individual furrows decrease in more or less direct proportion to
field length. As a result, substantially greater erosion occurs near the
top of the field than near the lower (tail) portion of the run (Mech and
Smith, 1967). Consequently, eroded sediments are often redeposited at
mid-field, prior to irrigation runoff.
Unfortunately, the finer sediments are eventually redeposited even
further from their point of origin. Glymph and Storey (1967) estimate,- for
example, that the U.S. reservior capacity of about 6.2 x 107 hectare-meters
is reduced by about 1.2 x 105 ha-m each year due to sedimentation. In
addition, eroded sediments are far from inert. The large reactive surfaces
of clay-sized sediments tend to sorb (and potentially desorb) many elements
(McDowell and Grissinger, 1966; Grissinger and McDowell, 1970). In many
respects, sediments act as chemical scavengers (Robinson, 1971; Taylor and
Kunishi, 1971; Schuman et al., 1973). It is partly for this reason that
sediment and total phosphorus concentrations tend to be so strongly
correlated (Carter et al., 1974; Carter, 1976).
The mechanics of sediment transport have been studied on several
occasions (Taylor, 1940; Israelson et al., 1946; Laursen, 1958; Meyer and
Wischmier, 1969), yet much still remains unknown (Robinson, 1971). Early
work (Zingg, 1940; Gardner and Lauritzen, 1946) showed that erosion could
be expressed as an empirical function of field slope and stream size. Work
by Criddle et al. (1956) led to development of the following relationship
describing maximum stream size for surface irrigation:
Qm = 0.63/S (1)
where Q = maximum nonerosive stream size in I/sec
S = slope of field in percent.
11
-------
It appeared that many irrigators within the study area for the research
covered by this report were employing maximum stream sizes closely matching
those suggested by the above equation. Hence, by reducing the slope an
irrigator may not have to reduce water application rate (assuming that this
rate exceeds the nonerosive level).
Contour furrowing has been suggested as one means of reducing the
effective slope of irrigated fields (Carter, 1976). Land grading, i.e.
reducing the slope at the lower end of a field to decrease sediment loss
from that position, is another option. Shorter field lengths have much the
same effect. Carter (1976) reported that less erosion will occur from two
successive 150-meter fields than from one 300-meter field. Gardner and
Lauritzen (1946) and Mech (1959) showed that a smaller stream size could be
used on the shorter field, thus reducing erosion.
Sediment concentrations in surface runoff also vary with the
.irrigation and cultivation practices. For crops cultivated only in early
spring, for example, sediment concentrations tend to decrease steadily as
the season progresses (Fitzsimmons et al., 1978). For more frequently
cultivated row crops (beans or sugar beets), sediment concentrations tend
to. be highest during periods of frequent cultivation and irrigation (Mech
and Smith, 1967; Carter et al., 1974; Carter, 1976).
Because surface irrigation techniques will always cause a certain
amount of sediment transport, considerable study has focused on ways to
prevent sediment discharge from fields. Fitzsimmons et al. (1978)
researched the effects of vegetated buffer strips as sediment trapment
areas at the ends of fields in southern Idaho. Although surprisingly
effective in causing the deposition of suspended sediments, grass filter
strips quickly became inundated with sediment. Hence, their effectiveness
was short-lived. Use of small-grain buffer strips was more effective later
in the season, though their initial effectiveness was generally less than
that of grass strips (Fitzsimmons et al., 1978). Other studies using
"vegetal filters" have been conducted by Robinson and Brockway (1980).
Fitzsimmons et al. (1978) also showed that cutback irrigation could
drastically reduce sediment erosion. Cutback irrigation involves reducing
the original stream size an hour or two after it has been set — with the
exact period depending on crop, slope, and soil type. In one southern
Idaho study, corn was irrigated via both, cutback and non-cutback
techniques. Seasonal sediment losses from the cutback furrows were only 7%
of those from the traditionally irrigated furrows. Although cutback
irrigation has been endorsed by several workers (McNeal et al., 1979; Van
Nieuwkoop, 1979), growers are quick to point out that considerably more
labor is required for this method of erosion control.
Several other methods have also been suggested for reducing sediment
loss. Among these are systems involving pumpback or tail-water reuse,
gated pipes, or even conversion to center-pivot (or other forms of
sprinkler) irrigation.
12
-------
While no single conservation measure appears capable of completely
eliminating erosion during surface irrigation, perhaps the most
cost-effective method of preventing sediment loss from irrigated fields is
the use of sediment entrapment basins (Robbins and Carter, 1975). Though
sediment basins are highly effective when properly designed, heavily
sedimented irrigation runoff may continue to deposit sediment in subsequent
drainageways if the runoff moves slowly enough (Carter et al., 1974).
Sediment basin retention efficiencies often vary thoughout the
irrigation season, depending on basin size and capacity. Generally, longer
basins provide longer transit times, and hence greater retention.
Fitzsimmons et al. (1978) reported retention efficiencies of up to 82% in
southern Idaho. The present study area had several basins which, at times,
approached or exceeded this level of efficiency. Another study in southern
Idaho showed that water leaving well-designed basins was as free of sediment
as when taken originally from the Snake River (Carter et al., 1974). Other
studies also suggest the practical re-use of basin-clarified runoff waters
if adequate settling times are allowed (Law et al., 1970).
IRRIGATION AND PHOSPHORUS LOSSES
Total phosphorus contents of soils from the western United States have
been estimated at between 900 and 3000 kilograms per hectare (Parker,
1953). Most of this phosphorus is "unavailable" to higher plants, and
applied phosphorus can convert to relatively insoluble complexes within a
matter of days after application (Taylor, 1967; Law and Skogerboe, 1972).
Once fixed by the soil, large amounts of phosphorus are easily lost via
erosion, since this element tends to accumulate near the soil surface (Law
and Skogerboe, 1972). Scarseth and Chandler (1938) studied a field
fertilized for 26 years with rock phosphate and determined that 60% of the
applied phosphorus had been lost through erosion. Work by Ensminger and
Cope (1974) and Ensminger (1952) showed comparable erosional losses of
phosphorus.
The correlation between phosphorus and eroded sediments is well
established (Rogers, 1941; Carter et al., 1976; Naylor et al., 1976;
Fitzsimmons et al., 1978). Since there is such a good connection between
the two, it would appear that control of sediment loss should effectively
control phosphorus loss as well. Carter et al. (1976) further suggest
that sediments eroded from silt loam or loam soils and subsequently
retained by settling basins carry a surprisingly high amount of phosphorus,
though the clay-sized fraction of suspended sediments has a considerably
higher phosphorus/sediment ratio than do sand- or silt-sized particles.
Apparently because the dispersion of clay-sized particles in runoff waters
tends to be incomplete, much of the clay-sized fraction still settles with
the larger particles. In other cases, however, retention of phosphorus by
settling basins is much less than retention of sediments.
Carter et al. (1974) suggest that adsorbed phosphorus in lake and
stream sediments may represent a continuing source of low-level release for
water-soluble phosphorus. Romkins and Nelson (1974) recommend further
13
-------
investigation into the effects of fertilization, cultural practices,
hydrology, and geochemistry on the phosphorus levels of irrigation runoff.
PHOSPHORUS AND EUTROPHICATION
Several workers have noted the staggering annual quantity of
phosphorus introduced into U.S. waterways (Glymph and Storey, 1967; Smith,
1967; Wadleigh, 1968; U.S. Department of Interior, 1968). According to
Wadleigh (1968), about 2.2 x 109 kilograms of phosphorus are eroded with
sediments each year. Although Massey and Jackson (1952) found that
sediments from eroded soil contained higher concentrations of several
nutrients (including phosphorus), Taylor (1967) speculates that perhaps as
little as 10% of the total phosphorus present would be available to plants
or aquatic life. The amount actually solubilized should depend on
subsequent shifts in equilibrium between phosphorus adsorbed to sediments
and that-already in solution (Taylor and Kunishi, 1971). Taylor (1967)
made the..observation that phosphorus pollution from an urban community of
1200 to 1500 persons approximately equals the amount of phosphorus eroded
from 260 hectares of fertile midwestern farmland.
Only 0.60 kilograms of phosphorus dissolved in 0.12 hectare-meters (1
acre-foot) of water would provide solution phosphorus levels of 0.03 mg/1,
a level sufficient for algal growth (Sawyer, 1947). Phosphorus solubility
in soils is low, however, especially in calcareous soils (Cole and Olsen,
1959; Murrman and Peech, 1969). Carter et al. (1971) studied return flows
from an 82,030 hectare irrigated tract in southern Idaho. They found that
plants and calcareous soils in the area effectively removed about 70% of
the PO.-P present initially in the irrigation water.
Several workers and reports (President's Science Advisory Committee,
1965; Smith, 1967; Taylor, 1967; National Technical Advisory Committee of
Water Quality Criteria, 1968; Wadleigh, 1968) have speculated that
increased fertilizer use has led to elevated phosphorus levels in surface
runoff waters, thus initiating and/or stimulating algal blooms. Viets
(1970) suggested, however, that phosphorus (along with nitrogen) may have
been assigned the villain's role for water quality deterioration and
eutrophication a bit too quickly. Viets (1971) and Romkins and Nelson
(1974) claimed that a cause and effect relationship between irrigation
runoff and eutrophication has not generally been established. While
phosphorus and nitrogen may be present in rather high concentrations in
irrigation runoff or deep percolation, other work (Kuentzel, 1969; Weiss,
1969) has questioned whether availability of soluble carbon may not have
been too readily dismissed when blaming nitrogen and phosphorus for
eutrophication. Still others (Bartsch, 1972) feel that phosphorus is the
element most singly responsible for eutrophication of lakes and streams.
CALCULATING SEDIMENT LOSSES
A great deal of research has been devoted to assessment and prediction
of soil erosion under non-irrigated (i.e. rain-fed) conditions (Soil
14
-------
Conservation Society of America, 1977; De Boodt and Gabriels, 1980; Kirby
and Morgan, 1980; National Soil Erosion - Soil Productivity Research
Planning Committee, 1981). Most of this work is an outgrowth of the
development of the Universal Soil Loss Equation, USLE. Unfortunately,, the
USLE is not directly amenable to irrigated agriculture. Predicting
sediment losses from surface-irrigated fields remains in its infancy, with
little literature being available on this topic.
Gossett (1975) proposed a model in.which the sediment loss from
surface-irrigated potatoes under specified conditions was set as a baseline
level. Other crops under other conditions were then assigned multipliers
which adjusted their predicted losses relative to the baseline level.
Multipliers were developed for crop type, runoff percentage, slope, and
soil texture. McNeal et al. (1979) extended this approach using data from
the southern Idaho studies of Fitzsimmons et al. (1978). Predicted
sediment losses for other settings were provided via appropriate
multipliers, using the generalized formula:
S = (Mc) (Mf) (Ms) (Mst) B (2)
where S = predicted sediment yield, in mt/ha,
M = crop type multiplier,
M = runoff multiplier,
M = slope multiplier,
M . = soil texture multiplier,
B = baseline sediment loss rate
Development of multipliers, and their change with varying field conditions,
was based on approximate trends from the literature (McNeal et al., 1979).
Verification of this model with actual sediment loss figures from
other areas has been attempted on a limited basis (McNeal et al., 1979).
Model predictions were correlated with measured sediment loss values from
six fields studied by Carter (unpublished data, Soil Science Society of
America annual meetings, Ft. Collins, CO., 1979). While correlation was
strong (r2 = 0.99), there was not a 1:1 relationship. Predicted values
were generally lower than Carter's measured values (McNeal et al. 1979),
especially for a field with exceptionally high sediment loss.
In an effort to refine sediment loss calculations, water stage
recorders have been used in our studies. With an accurate record of runoff
volume in hand, one needs only to determine average or total sediment
content of that volume in order to calculate sediment loss. Brown et al.
(1974) studied the outflow of an 82,030 hectare irrigated tract in southern
Idaho, and concluded that bi-weekly sediment analysis was sufficiently
accurate to compute sediment losses. This approach, however, cannot
15
-------
generally be used if one is to distinguish the amount of sediment lost
from individual fields, because daily sediment loss depends on individual
irrigation patterns and field conditions -- conditions which are virtually
impossible to sort out once runoff from individual fields has merged into a
common drain. Boyle (1980), for example, relied on daily "grab" sampling
for sediment analysis. Fitzsimmons et al. (1978) sampled individual fields
much more frequently during each irrigation set; at hourly intervals
initially, and then at two- or four-hour intervals during the remainder of
the set.
There is ample evidence that the time selected for sediment analysis
of runoff samples is critical. Mech (1959) observed that the first 32
minutes of runoff from a freshly cultivated furrow carried more than 78% of
the total mass of sediment eroded during a 24-hour irrigation set.
Furthermore, essentially all sediment eroded during the 24-hour set was
lost during the first four hours of runoff. Characteristic features for
any irrigation cycle depend on several factors, including those related to
the physical condition of soil and crop, as well as those related to the
specific irrigation habits of the grower.
Kabir and King (1980) have developed a mathematical model utilizing
the concept of steady-state erosion to predict sediment discharge from
irrigated furrows. Based on dimensional analysis and multiple regression,
they analyzed sediment outflow as a function of slope, run length, median
grain diameter of streambed particles, and water intake rate. The
regression equation predicted sediment discharges only from freshly
cultivated furrows, however. Later work by Kabir and King (1981a,b,c,d)
further developed the initial model in order to predict erosion under
unsteady conditions as well. The approach was then used to predict
seasonal sediment losses from irrigated furrows. In addition, by knowing
cultivation patterns and furrow spacings, the model could be used to
determine seasonal sediment losses from a given field.
16
-------
SECTION 5
STUDY AREA DESCRIPTION
The study area is located in Block 86 of Washington's Columbia Basin
Irrigation Project. It is situated in T17N, R26E (sections 27,28,33, and
34) approximately 35 km west of Othello, and 8 km northeast of Royal City,
in Grant County, WA (Figure 1).
Figure 2 shows the study area as subdivided into its various farm
units. Of the total 800 hectares, 769 are currently irrigated. The
remainder are occupied by roads, homesites, other buildings, and irrigation
and drainage canals. In 1981 non-surface irrigation methods were used on
34.4% (265. ha.) of the irrigated area. These methods included center-pivot
systems on 212 hectares (Figure 3), solid-set irrigation on 36 hectares,
and a side-roll system on 16 hectares. The remaining 505 hectares were
surface-irrigated with siphon tubes. Shaded areas in Figure 3 represent
farm units, or portions of farm units, monitored by WSU personnel during
the period 1978 through 1981.
Columbia Basin
Project
• Royal
Royal City
Othello
Study Area
Figure 1. Location of the study area in Washington State.
17
-------
00
x = flume
o = pipe
Figure 2. Study area farm units, flume sites, and pipe locations.
-------
-/ ^v
Figure 3. Monitored fields (shaded areas) within the study area, 1977-1981.
Circles and the half circle represent center-pivot irrigation systems.
-------
SOIL TEXTURE AND TOPOGRAPHY
The predominant soil types within the study area are silt loams (66% '
of the area), fine sandy loams (32%), and fine sands (<3%) of the Kennewick
series. A caliche layer is usually present 100 to 150 cm below the soil
surface. This layer occurs at or near the present soil surface on the most
severely eroded fields of the study area. Approximately 57% of the
surface-irrigated land has slopes ranging from 0 to 3%, and 43% of the area
has slopes ranging from 3 to 5% or greater. The length of run ranges from
150 to 300 meters on 56% of the area, and from 300 to 460 meters on 35% of
the area.
CROPPING PATTERN
Major crops grown in the study area from 1977 through 1981 are shown in
Table 1. Wheat, corn, mint, beans, peas and orchard constituted a majority
of the cropped area. Small areas of additional crops such as alfalfa,
onions, dill, and asparagus were also evident during the study period.
Sugar beet production, which previously had represented 19 to 22 percent of
the study area, was discontinued due to closure of the U&I sugar beet
processing plant at Moses Lake, WA prior to the 1979 growing season.
Most crops were planted annually, with the exception of the mint and
orchard. Crop rotation was a common practice, though farm unit 53 was
planted to corn each year because of an infiltration problem and excessive
salinity. With the exception of winter wheat and carrot seed, most crops
were planted between early March and late April, and were harvested between
early August and mid-September. Planting and harvesting dates for the major
crops are summarized in Table 2.
IRRIGATION PRACTICES
Irrigation water for this area is pumped from the Columbia River at
Grand Coulee Dam, and then flows by gravity through Banks Lake and a
series of canals to the Columbia Basin Project. Irrigation water is
diverted to Block 86 by the West Canal, which flows from early spring (late
March) through fall (late October) at a volume of about 12,750 I/sec.
Seasonal irrigation patterns for the study area vary with individual
crop requirements. Winter wheat, for example, may receive only 5 to 8
irrigations per year, while corn or potatoes may receive 20 to 23
irrigations or more. Although water diversions to individual farm units
are generally highest in mid-summer, pre-irrigation demands in early spring
and late fall account for heavy water usage and disproportionate sediment
losses. Grower requests for water are made to the Watermaster's Office of
the QCBID located at Royal, Washington. Each request is relayed to a ditch
rider who adjusts a turnout wier or a constant-head orifice on the West Canal
(or a supply lateral) in order to provide the farm unit with the requested
amount of water. The ditch rider's schedule takes him past the study area,
usually at mid-morning, six days a week.
20
-------
TABLE 1. STUDY AREA CROP HISTORY, 1977-1981
r\j
Crop
alfalfa
asparagus
barley
beans
beets
carrots
corn
Christmas
trees
dill
fal low
lettuce
mint
onions
orchard
pasture
peas
potatoes
wheat
wheat and
dill
TOTALS
1977
38.5
-
-
-
168.5
17.1
86.7
-
-
35.2
4.2
66.2
-
97.6
6.3
35.9
-
199.5
-
755.7
1978
38.5
-
-
20.0
145.6
19.7
58.4
-
-
-
5.2
116.9
-
97.6
6.3
37.2
-
227.2
-
772.6
Hectares
1979
38.5
-
-
17.5
-
-
129.3
-
25.2
-
-
55.9
11.6
97.6
6.3
109.2
-
253.4
12.1
756.6
Percent of Area
1980
42.5
9.9
26.1
62.7
-
47.4
99.1
1.5
-
-
-
115.3
-
97.6
6.3
77.3
-
186.2
-
771.9
1981
30.8
9.9
-
120.0
-
18.5
136.9
1.5
-
-
-
90.7
-
97.6
6.3
43.1
64.6
146.9
-
766.8
1977
5.1
-
-
-
22.4
2.3
11.0
-
-
4.7
0.6
8.8
-
13.0
0.8
4.8
.
26.5
-
100.0
1978
5.0
-
-
2.6
18.8
2.5
7.6
-
-
-
0.7
15.1
-
12.6
0.8
4.8
-
29.4
-
99.9
1979
5.1
-
-
2.3
-
-
17.0
-
3.3
-
-
7.4
1.5
12.9
0.8
14.4
-
33.4
1.6
99.7
1980
5.5
1.3
3.5
8.1
-
6.1
12.8
0.2
-
-
-
14.9
-
12.6
0.8
10.0
-
24.1
-
99.9
1981
4.0
1.3
-
15.6
-
2.4
17.9
0.2
-
-
-
11.8
-
12.7
0.8
5.6
8.4
19.2
.-
99.9
-------
TABLE 2. PLANTING AND HARVESTING DATES, BLOCK 86 STUDY AREA
PLANTING DATE CROP HARVEST DATE
Oct.
Feb.
May
Mar.
Apri
Aug.
Mar.
1 -
1 -
1 -
15
1 15
1 -
15
Nov.
Mar.
May 25
- Apri
1
15
1 1
- June 1
Sept.
- Apri
1
1 1
Winter Wheat
Spring Wheat
Beans
Peas
Corn
Carrot Seed
Mint
Aug.
Aug.
Aug.
July
Oct.
Aug.
June
Sept.
1 -
1 -
25
15
20
15
15
1
Aug.
Aug.
- Sept
- July
- Dec.
- Sept
- June
- Sept
15
30
•
15
30
1
•
15
30 and
. 15
Mar. 15 - April 1 Sugarbeets Sept. 25 - Oct. 15
Farm units were generally comprised of several fields. This allowed a
variety of crops to be grown and irrigated individually. Diversion records
for each farm unit were provided by the QCBID. Such records were useful
when -re-constructing the probable irrigation sequence for a specific crop.
Diversion records were also helpful in portraying seasonal water usage.
Two major drainageways (DW 272A and DW 272A1), as well as one major
supply lateral (W 69.7), transport runoff and irrigation waters through the
study area (Figure 3). The USGS (Water Resources Division) maintained a
monitoring and gauging station at the confluence of the two drains
(Figure 3) throughout the study period.
Approximately 94 percent of the irrigated land in the study area was
surface-irrigated during the year prior to initiation of the study (Table
3). By the 1981 irrigation season, however, only 65.6 percent of the area
remained surface-irrigated. As Table 3 demonstrates, almost 80 percent of
the conversion was to center-pivot irrigation systems.
The most common method of surface water application in the study area
was through the use of siphon tubes and corrugation-type furrows. Depending
on soil moisture, slope, length of run, and type of stage of crop growth,
growers applied stream sizes varying from 0.12 to 0.75 I/sec (2 to 12 gpm)
at the head ditch, using 19.0 to 25.4 mm (3/4 to 1 inch diameter) siphon
tubes. Average stream sizes for the study area were 0.25 to 0.38 I/sec
(4 to 5 gpm). Average mid-season irrigation frequency was approximately
7 days, often decreasing to 5 days during periods of peak water usage.
The most common irrigation set-time was 24 hours, with occasional 12 hour
22
-------
TABLE 3. ANNUAL PERCENTAGE OF SPRINKLER AND SURFACE-IRRIGATED LANDS IN
THE BLOCK 86 STUDY AREA
Year
1976
1977
1978
1979
1980
1981
Center .
Pivot
6.1
10.1
18.5
27.5
27.5
27.5
Solid
Set
•
-
-
4.7
4.7
4.7
Side
Roll
-
2.2
2.2
2.2
2.2
2.2
Total
Sprinkler
6.1
12.3
20.7
34.4
34.4
34.4
Total
Surface Irrigation
93.9
87.7
79.3
65.6
65.6
65.6
set-times on fields with more steep slopes (e.g. Farm Unit 81, Field B).
No cutback or pumpback systems were utilized by the growers, though cutback
systems were tested extensively on an experimental basis during 1981.
Most crops were pre-irrigated before planting. To insure sufficient
water application for the pre-irrigation, and also for the first irrigation
after planting, growers generally used a block-set pattern, irrigating"
every row of the block, and a stream size as much as twice that used for
normal irrigation. Large sediment losses during pre-irrigation and/or the
first irrigation of the season can generally be attributed to this type of
practice. For the remainder of the season, growers reverted to irrigation
of every 4th, 5th or 7th row, in regular sequence. Continuous irrigation
of a given field, depending upon the crop, usually began from mid-May to
early June, and continued until mid-August to mid-September.
Water delivery per hectare varied from a low of 1.12 ha-m/ha (3.7
ac-ft/ac) in 1980 to a high of 1.43 ha-m/ha (4.7 ac-ft/ac) in 1977.
Average area-wide water delivery for the study period was 1.29 ha-m/ha (4.2
ac-ft/ac). Actual on-farm application efficiencies of surface-irrigated
fields were between 40 and 50 percent, and could be increased to 55 to 65
percent if optimal stream sizes were adopted. It was also shown that use
of cutback techniques could increase application efficiencies to the range
of 70 to 80 percent.
Most farmers in the study area have generally relied on prior
experience and on indicators such as crop color (e.g. for beans) or
leaf-roll (e.g. for corn) to determine when to irrigate. No irrigation
scheduling devices were used, with the exception of one grower who used a
portable soil moisture probe to verify proper irrigation timing.
23
-------
CULTIVATION AND TILLAGE PRACTICES
Furrows were constructed using several types of equipment; though the
most common was a standard shovel ditcher. Numbers of cultivations in the
study area varied with crops. Most row crops were cultivated at least once
prior to the first irrigation. Beans were usually cultivated 2 to 3 times
during the season, while corn and carrots were cultivated only 1 or 2 times.
Large soil losses generally resulted from excessive cultivation.
CLIMATE
The study area has warm and dry summers, averaging 27 to 29°C (80 to
85°F), and cool winters, averaging -4 to 2°C (25 to 35°F). During the
study period average maximum temperature equaled or exceeded 35°C (95°F)
during 6 to 10 days annually, with the exception of 1980 (with only 2 such
days), and typically reached 37.8°C (100°F) or higher on 2 to 3 days.
Annual precipitation is low, averaging about 185mm (7.3 inches), with
approximately 38% of the rainfall normally occurring during the early
period of the growing season. Annual evaporation from a class A pan ranged
from 1016 to 1143mm (40 to 45 inches). Average growing-season evaporation
was 965mm (38 inches), with maximum monthly evaporation (usually in July)
averaging 229mm (9.0 inches). Climatic data collected during this study
were utilized for development of an irrigation scheduling model. A summary
of data for the 1979-1981 growing seasons is presented in Table 4.
24
-------
TABLE 4. SUMMARY OF IRRIGATION SEASON WEATHER DATA FOR THE BLOCK 86 STUDY AREA, 1979-1981.
ISi
Ui
Temperature
1Q7Q
J.J/.J
APR
MAY
JUN
JUL
AUG
SEP
OCT
1980
APR
MAY
JUN
JUL
AUG
SEP
OCT
1981
APR
MAY
JUN
JUL
AUG
SEP
OCT
Avg.
Max.
16
21
24
29
28
24
18
16
19
21
27
28
23
17
17
21
23
28
32
25
16
Daily
Max.
(°C)
25
30
32
39
34
29
27
21
28
27
37
33
31
32
27
29
31
37
38
33
25
Avg.
Min.
4
7
18
12
12
12
6
4
15
9
13
12
11
4
2
7
11
13
16
10
3
Daily
Min.
(°C)
-2
-2
4
4
9
8
0
-1
1
2
4
3
4
-1
-4
1
3
6
9
3
-1
Avg.
Db(%)
(°C)
17
18
22
19
18
12
13
13
14
17
18
18
11
15
19
21
24
20
Avg.
Wb(%)
(°C)
11
13
16
15
13
8
11
10
13
14
13
11
6
10
13
16
18
14
Avg.
R.H.
(%)
48
60
53
62
58
60
70
65
85
80
58
45
58
50
54
57
53
51
Rainfall
Daily
Cum. Max.
(mm)
9.1 4.6
3.0 1.3
4.6 2.5
9.9 4.6
10.2 3.3
2.5 1.0
18.8 8.4
20.8 11.4
35.1 11.9
31.2 11.2
0.3 0.3
0.8 0.3
18.5 11.9
5.3 3.6
0.3 0.3
31.0 15.2
18.5 6.4
5.8 5.8
__
13.5 9.9
19.3 8.1
Wind
Avg.
24 hr
( km/24
13
89
113
77
55
51
42
97
63
53
76
146
126
108
296
201
232
214
161
145
156
Daily
Max.
hrs)
45
248
357
301
180
193
182
372
142
185
272
343
325
401
668
473
554
486
438
389
360
Evaporation
Cum.
(mm)
133
177
224
248
201
155
71
120
150
162
221
188
124
98
70
154
177
219
194
129
56
Avg.
(mm)
4.4
6.1
7.5
8.0
6.5
5.2
2.3
4.0
4.9
5.4
7.1
6.1
4.1
3.1
2.3
5.1
5.8
7.3
6.3
4.3
1.8
Daily
Max.
7.4
8.4
11.5
11.3
10.6
8.4
5.5
9.7
8.5
8.4
12.2
9.2
7.5
5.5
3.6
9.2
13.0
16.5
10.6
8.6
6.1
-------
SECTION 6
METHODS OF DATA COLLECTION AND ANALYSIS
Data were collected from throughout the study area each year in order
to evaluate the magnitude of pollution from irrigation return flows. These
included sampling of: (1) cumulative losses from the study area at the
main drains, (2) losses from individual fields, and (3) losses from
individual furrows. The procedures for each type of sampling will be
discussed in turn below.
SAMPLING OF THE MAIN DRAIN
Surface runoff and subsurface drainage from the study area were
collected by two open drains (DW272A and DW272A1). The U.S. Geological
Survey monitored these two drains, including installation of two automatic
samplers which were programmed to sample at two-hour intervals from
June 1977 to November 1980, at four-hour intervals from March 1981 to June
1981, and at two-hour intervals once more from July 1981 to October 1981.
No samples were collected from December 1980 to March 1981. Each automatic
sampler (manufactured by the Manning Environmental Corporation) was placed
in a refigerator to avoid nitrogen transformations in the collected
nutrient samples. From April 1, 1977 until installation of the automatic
samplers in early June of 1977, daily samples were obtained from the same
locations using a depth-integrated sampler.
Because each automatic sampler collected water from only one point in
the stream profile, a number of depth-integrated samples were taken
periodically for comparison with those taken by the automatic sampler.
These correlation samples, taken with a standard U.S. DH-48 sampler, were
used to correct data from the automatic samplers. Samples were collected
from the West Canal (the supply canal) once daily during 1977 and 1978.
From 1979 to 1981, sediment samples were collected approximately daily from
the canal, while nutrient samples were collected approximately 3-4 times
per week. The U.S. DH-59 suspended sediment sampler was used to collect
all canal samples.
Diversion samples were also collected periodically from each farm
unit, using a U.S. DH-48 sampler at outfalls over a weir, or where
turbulence was sufficient so that total constituent concentrations could
be measured. Periodic sampling of constituent concentrations for water
flowing to each farm unit was done for the purpose of determining suitable
correction factors to be applied to total constituent inflows.
26
-------
Several days throughout the year, samples were collected and analyzed
at 2-hour intervals from the two principal drains and at the daily sampling
site on the West Canal. Results from the discrete samples were used to
determine daily variability in constituent discharge. Systematic errors
due to daily variability were corrected when computing net constituent
discharge (mass balance) from the study area.
Water stages were recorded by gauges on both the DW272A and OW272A1
drains. Water discharge was measured periodically with a current meter,
and discharge-stage relations were established from these data using
standard techniques described by Carter and Davidian (1965a, 1965b). Water
discharge for individual farm units was obtained from weir formulas.
Sediment and nutrient samples collected from the main drainageways
were composited daily and nutrient samples were packed in ice and shipped
to USGS laboratories in Tacoma, WA (1977-78) or Denver, CO (1979-81) for .
analysis. These were analyzed for total phosphorous, mineral nitrogen, and
total nitrogen. All sediment samples for the project were analyzed by the
U.S. Bureau of Reclamation Agronomic Soils and Water Laboratory at
Ephrata, WA.
SAMPLING OF INDIVIDUAL FIELDS
Approximately 40% of the surface-irrigated fields within the study
area have been monitored each season since 1978 (Figure 3). Continuous
monitoring of irrigation runoff volumes was accomplished by placing
trapezoidal or Parshall flumes equipped with water-stage recorders at the
end of 15 to 20 fields- each season (Figure 2). These devices were generally
situated so that runoff from an entire field or farm unit passed through the
flume just prior to entering a sediment basin or drainageway. A majority of
the fields were monitored with 5-cm trapezoidal flumes,though 5-cm and 15-cm
Parshall flumes were also used. Instantaneous runoff volume for each flume
was related to the depth of water in the throat of the flume via an
appropriate calibration equation. Water depth measurements for each
flume were reported in terms of gauge height (GH, in meters). For a 5-cm
trapezoidal flume, runoff volume (Q, in I/sec) is given by the following
set of equations:
Q = 462.4 GH1'95 for GH < 0.12 meters
Q = 771.6 GH2'19 for 0.12 < GH < 0.18 meters (3)
Q = 2030.1 GH2-76 for GH > 0.18 meters
For 5-cm Parshall flumes, the corresponding relation is:
Q = 120.7 GH -1-55 (4)
27
-------
For 15-cm Parshall flumes the relation is:
Q = 381.4 GH 1'58 (5)
®
A Stevens type A-35 strip chart water-stage recorder was coupled with each
flume. The resultant hydrographic record not only provided an accurate
measurement of seasonal runoff, but also illustrated diurnal fluctuations
during specific irrigations. At the end of each season all strip charts
were digitized on bi-hourly intervals.
Under ideal circumstances, each flume would monitor irrigation runoff
from only one field. In some cases, however, flumes received runoff from
several fields. In these instances, unless separate efforts were made to
determine when each field was being irrigated, individual-field sediment-loss
estimates became difficult.
In addition to the flumes, thirteen pipes discharged surface runoff
into the major drainageways of the study area (Figure 2). Although surface
runoff pipe discharges were sampled on approximately a daily basis, no
method was available for recording cumulative runoff volumes. Instead,
instantaneous discharge rates were recorded each time a sample was taken.
Pipe flowrates were calculated by measuring water depth in the pipe and
applying Manning's equation for open-channel flow:
Q = (1000/n) R 2/3 A S 1/2 (6)
where Q = pipe discharge in I/sec
n = Manning's roughness coefficient
R = hydraulic radius of pipe
= area/wetted perimeter, m
A = cross sectional area, m2
S = slope
Pipes were calibrated (primarily for assignment of an appropriate roughness
coefficient) using a bucket and stopwatch technique.
Approximately 54% of the study area has subsurface drainage, which
discharges into the main drainageway via nine groundwater pipes. Throughout
the study period all subsurface drains appeared to operate satisfactorily
with the exception of one farm unit (FU 53). In this latter case the water
table rose to within 0.3 meter of the ground surface, causing problems with
infiltration and leaching. All groundwater pipes were sampled on a weekly
basis throughout the irrigation season and bi-weekly during the remainder
of the year. Sediment analysis of groundwater samples was deemed
unnecessary, but analysis for total phosphorus as well as nitrogen
(nitrate + ammonium) was carried out. Nitrogen data were utilized in a
separate study analyzing movement of this nutrient through typical soil
profiles of the study area (Schnabel, 1981).
28
-------
Sampling Schedule and Analytical Procedures
The collection of discrete runoff samples from each flume and surface
pipe provided the data base for sediment and phosphous loss estimates.
Four samples were generally taken daily at each flume. An initial sample
was taken during mid-morning, with the second and third samples taken
concurrently in early afternoon and the fourth in mid-to-late afternoon.
One of the early afternoon samples was used for nutrient analysis and the
other for sediment analysis. This sample collection schedule was evolved
in order to provide sediment and phosphorus loss values at critical times
during the irrigation day. These hours were generally characterized by
rather large variations in sediment concentration and runoff values,
variations due in large measure to the daily changing of siphon tubes for
the field being irrigated. !.
Nutrient samples from the flumes were frozen until they could be
transported to WSU for laboratory analysis of total phosphorus and mineral
nitrogen. Under most conditions, either freezing or refrigeration and
rapid analysis can be used as an effective means for preserving the chemical
integrity of a nutrient sample (Nelson and Romkins, 1972; Burton, 1973).
Sediment samples were subjected to preliminary testing in a field
laboratory located in the Royal Watermaster1s headquarters prior to
shipment for further analysis. Each sample was tested for turbidity
(measured in NTU, using a Model 2100A turbidimeter manufactured by Hach
Chemical Co.). The more heavily sedimented samples usually had to be
diluted (American Public Health Association, 1971) to bring them within
range of the instrument. The electrical conductivity (EC) of each sample
was also measured (in mmho/cm, using a Model 16300 portable conductivity
meter (manufactured by Hach Chemical Co.). The sediment content of all
samples was determined with an Imhoff cone, in which the sediment
concentration (ml/1) was recorded after 15 minutes settling time.
Sediment analysis was performed by the U.S. Bureau of Reclamation
Agronomic Soils and Water Laboratory at Ephrata, WA. Standard procedure
was to weigh the collection bottle, pass the solution through a 0.47 mm
Millipore filter, and then weigh the filtered solids. Results were
expressed in mg/1, with solution density assumed to be 1.00 g/ml. An
evaporating dish technique was used after June 1980 to determine the
concentration of sediments in aliquots from samples which were estimated to
contain greater than 10,000 mg/1 suspended solids. A more detailed outline
of the sediment analysis procedure is given in Appendix D.
Total phosphorus and mineral and total^nitrogen in the nutrient samples
were analyzed with a Technicon AutoAnalyzer II. The total phosphorous
procedure was based on colorimetric analysis involving reaction of the
ortho-phosphate ion with molybdate to form 1,2-molybdophosphoric acid.
Ascorbic acid reduces this compound under strongly acidic pH to the
chromophoric product, phosphomolybdenum blue (Crouch and Malmstadt, 1967;
Burton, 1973). Color development is accelerated and stabilized by antimony
potassium tartrate (Murphy and Riley, 1962; Watanabe and 01 sen, 1965). The
29
-------
intensity of color for this solution is directly proportional to the
concentration of orthophosphate over a suitable range.
Mineral and total nitrogen determinations were based on colorimetric
analysis involving reaction of N03 (+N02) or NH4 with; a) a copper-cadmium
reduction column, reaction with sulfanil amide under acidic conditions to
form a diazo compound, and coupling with N-1-naphthylenediamine
dihydrochloride to form a reddish-purple azo dye (for N03 + N02, or b)
sodium salicylate, sodium m'troprusside and sodium hypochlorite in a
buffered alkaline medium at a pH of 12.8 to 13.0, to form an ammonia-
sal icylate complex (for NH4). The digestion process for total nitrogen
involved use of an H2S04/K2S04/H 0 mixture in a block digestor at 200°C
(preheating for 1 hr) and 370°C ?for 1 hr).
The digestion process for total phosphorous involved addition of
sulfuric acid and ammonium persulfate to a 50 ml aliquot,- followed by
autoclaving for 30 minutes at 121 C (Gales et al., 1966; Technicon
Industrial Systems, 1973; EPA, 1929). Colorimetric analysis for total
phosphorus using the AutoAnalyzer II followed the procedure of Canelli
and Mitchell (1975). There was some concern that heavily sedimented samples
would not completely release their various forms of phosphorus during this
relatively weak digestion procedure (Boyle, 1980). As a result, heavily
sedimented samples were diluted prior to digestion by a.factor of l-,5-,
or 10-fold, depending upon estimated sediment load.
Continuous Monitoring of Runoff
Since daily variations in sediment concentration can be pronounced,
continuous monitoring of changing sediment loads is desirable. This was
not feasible for each individual flume, but periodic continuous monitoring
of selected flumes was carried out using a Manning S-4050 portable
discrete sampler. Samples were collected automatically from the flume
under study on a bi-hourly basis. Collection volume was approximately 900
ml, with 500 ml being used for sediment analysis and the remainder for
phosphorus analysis. A total of thirteen 48-hour studies were carried out
in 1980. Some fields were monitored twice in this fashion, once in mid
season and again during a late-season irrigation of fall-planted crops.
By merging these bi-hourly sediment samples with the continuously
recorded flume discharge, the interaction between runoff volume and
sediment concentration could be deduced. It was tempting to speculate that
sediment concentration should show a well-defined dependence upon runoff
volume for a given site. No consistently significant correlation between
the two parameters was found, however. Such a relation would have greatly
simplified prediction of sediment losses, since runoff data for each flume
were recorded continuously throughout the season.
Use of the continuous sampler during the 48-hour studies still
provided useful data for the prediction of seasonal sediment losses. By
assuming that the two-hour interval between samplings was sufficently brief
to preclude significant fluctuations in sediment concentration, an accurate
30
-------
estimation of daily sediment loss could be made. During continuous
monitoring of a specific flume, manual collection of six additional grab
samples (3 each day during the 48-hour monitoring period) from the flume
provided normal estimates of daily sediment loss. These values could then
be correlated with measurements for that flume as provided by the
continuous 48-hour sampling program.
Sediment Loss Calculations
Two methods were used to estimate sediment losses. Both methods
employed the three daily grab samples and their associated sediment
concentrations. Method I involved computing an average sediment
concentration for the three daily samples, and multiplying this mean value
by the daily runoff volume (obtained from the flume record). Method II
involved computing the average for three daily pairs of sediment
concentration and instantaneous runoff rate values.
On a weekly basis, method I involved the computational sequence:
S Q C = Sediment loss in metric tons/week, (mt/wk) (7)
where S = average sediment concentration
from 3 daily grab samples for
the week, (mg/1)
Q = runoff volume, (I/week)
C = conversion factor, (10~9 mt/mg).
whereas method II involved the sequence: (8)
(S Q) C = sediment loss in mt/week,
where (S Q) = weekly average of paired sediment
concentrations and associated
instantaneous runoff rates,
C = 10'9 mt/mg.
Sediment loss calculations for individual fields within the study area
are presented in Section 8. While it was hoped that the flume records would
provide sufficient data for these estimates, there were instances where
they did not. In some cases a flume had malfunctioned, recording erroneous
levels of runoff or perhaps not even recording runoff at all. In other
cases, a flume was not installed prior to early-season irrigation. From
irrigation records provided by the QCBID, dates of water delivery to a
given farm unit could be determined. If runoff through a flume on a field
of that farm unit was not recorded during the corresponding periods, an
adjustment was necessary. In making this adjustment, a runoff percentage
was first estimated from extrapolation or interpolation of available data.
Next, an average sediment concentration was assigned, based upon average
values preceding and following the period in question. Such adjustments,
31
-------
while lacking in precision, did provide a resultant record which was
considerably more accurate than i.f the adjustments had been omitted.
Since irrigation water diverted from the West Canal already contained
small amounts of sediment and phosphorus, calculation of net sediment and
phosphorus losses was based on the relationship:
N = 0 - I (9)
where N = net amount of sediment or phosphorus lost from
the study area, mt or kg
0 = outflow of sediment or phosphorus from the study
area, mt or kg
I = inflow of sediment or phosphorus to the study area,
mt or kg.
The respective inflow and outflow of sediment and phosphorus were
determined in turn from the relationship:
0 or I = C W K! K2 (10)
where C = weekly average concentration of sediment
or phosphorus, mg/1,
W = water diverted or discharged from the
study area per week, ha-m,
K! = constant = 107 1/ha-m,
K2 = constant = 10"9 mt/mg.
Sediment and phosphorus concentrations in water diverted to and
discharged from the study area were monitored, as described previously, by
the USGS (Water Resources Division). Study-area volume discharges were
provided as well by the USGS, and study-area volume diversions were
calculated from records provided by the QCBID.
SAMPLING OF INDIVIDUAL FURROWS
Individual-furrow studies were also conducted during the 1978-1981
irrigation seasons, to determine the variations in sediment loss
accompanying a range of furrow-inflow stream sizes and managemental
techniques such as cutback irrigation. Infiltration, runoff and advance
rates were all used in determining individual-furrow sediment loss values.
Inflow and Rate of Advance
A section of the field to be monitored for individual-furrow measurements
was first selected and staked at 30.5 m (100 ft) intervals, beginning at the
head ditch. Inflow to each experimental furrow was adjusted to one of 3 or 4
pre-determined rates by careful vertical positioning of a siphon tube, while
timing discharge into a 3.8 liter (1 gallon) container with a stop watch.
32
-------
Initial flow adjustments were often performed on non-measured furrows until
desired inflow was achieved. Each inflow rate was normally tested in
duplicate to assess the degree of within-field variation in rate of advance,
runoff, and sediment loss. Cutback of stream inflow rates was employed for
many of these duplicates in 1981, usually within 60 minutes after the stream
had reached the end of the furrow. Consequently, the number of exact
replicates was reduced, but results from these studies could be better used
to design water management approaches, to help improve irrigation efficiencies
and to lessen field-wide sediment loss.
Advance rates along each furrow were recorded for each of the
stations, until the stream had reached the end of the furrow. Additional
information pertaining to crop type and stage of growth, meteorological
conditions, farmer irrigation practices, furrow spacing and prior tillage
was also recorded. Irrigation information which was commonly recorded
included average application rate, spacing of siphon tubes (whether
watering alternate furrows, every third or fifth or seventh furrow, etc.),
and distance between individual furrows. Some attempts were also made to
characterize antecedent soil moisture status. In addition, surface-soil
texture and average slope for each farm unit had been previously determined
for inclusion in the analyses.
Outflow and Sediment Loss Measurements
At the end of each furrow being monitored, a small pit was dug for the
purpose of collecting sediment samples and measuring runoff rate. Each pit
was large enough to allow free fall of runoff, through a 30 to 60 cm section
of 5 cm O.D. PVC pipe, into a collection container (Figure 4). It was
critical that each pipe be positioned in the furrow upstream from the pit
so as to allow unrestricted flow of water. Any pooling of water behind
the pipe would cause some settling of suspended solids and consequently
incorrect sediment loss values, as well as making runoff rate determinations
somewhat ambiguous due to variations in wetted-furrow perimeter. Outflow
rate was determined by collecting discharge from the pipe in a 3.8 liter
metal container, while timing with a stop watch to provide an accurate
volumetric measurement.
Initial runoff rates were normally measured within 30 minutes after
the irrigation stream had reached the end of the furrow. Immediate runoff
samples were not taken, due to the need to flush from the pipe and soil and
other debris which may have accumulated during its placement. Outflow
rates for each furrow were then recorded at 15-minute intervals for the
first hour, at 30-minute intervals for the following hour, and at 60-minute
intervals thereafter until late afternoon or early evening. By this time,
runoff rates had normally leveled off. The following morning (after inflow
had been checked and readjusted if necessary), two additional samples were
collected at 15-minute intervals. Subsequently, inflow was terminated and
recession time was recorded. Recession time was defined as the time
elapsed between termination of inflow and cessation of runoff.
33
-------
Free
Flowing
Outflow
Irrigation
Furrow
Figure 4. Outflow sampling device.
Sediment loss from individual furrows was estimated from Imhoff cone
readings. Each time runoff rate was determined, a corresponding 1 liter
sample was collected, poured into a clean Imhoff cone and allowed to settle
undisturbed for 15 minutes. An excellent correlation (r2 = 0.85, df = 703)
has been established between total suspended solids (TSS) and Imhoff cone
readings following 15 minute settling periods for 1 liter runoff samples
from the study area. The volume of sediment settling in the cone during
this time interval was converted to a sediment concentration (mg/1) using
the equation:
Y = 200 + 1180 X
where Y = TSS, mg/1
X = 15-minute Imhoff cone reading, ml/I
(11)
Stice (1982) has worked out more detailed correlations for several textural
classfications of surface-irrigated soils throughout Washington state. The
relationship works best for coarse-textured soils, which have better settling
characteristics than do finer-textured soils. Soils with higher clay
34
-------
percentages can therefore be expected to evidence decreased correlations.
Equation 11 appears to work well, however, for the predominant silt loam
soils of the study area (van Nieuwkoop; 1979).
METEROLOGICAL INSTRUMENTATION
A class A weather station, located 3 km west of the study area,
provided daily records of wind velocity, solar radiation, rainfall, free
water surface evaporation, and maximum/minimum and wet-bulb/dry-bulb
temperatures. Wind velocity was measured at a height of 2 meters using a
totalizing 3-cup anemometer (Weather Measure Corporation). Wind speed was
read in kilometers of wind travel per day. Daily maximum and minimum
temperatures were measured using a maximum-minimum self registering
thermometer (Taylor No. 5458). Wet- and dry-bulb temperatures were also
measured once daily, using a sling pyschrometer between 8:00 and 9:00 a.m.
Precipitation was recorded utilizing both a standard volumetric, and a long
term, event-recorder, tipping-bucket, rain gauge (Weather Measure
Corporation). Daily incoming (shortwave) solar radiation was recorded
using a net radiometer (Campbell Scientific Corp.). A summary of weather
data is presented in Table 4.
ECONOMIC STUDY OF BMP ADOPTION
Considerable effort has been expended in identifying cost effective
Best Management Practices (BMPs) for irrigation return flow management
as part of local, state, and federal 208 planning activities. Less
attention has been given to motivating irrigators to adopt these practices.
Yet, identifying cost effective BMPs and motivating their adoption are
inseparable aspects of the overall problem. Nonetheless emphasis typically
is placed only on the identification aspect, frequently to the exclusion of
motivational considerations. It is not too surprising that many
recommendations have fallen on deaf ears.
Because farm level motivations must be understood if cost effective
BMPs are to be successfully implemented, some discussion is devoted to how
farm level decisions are made. Factors such as farm objectives, financial
structure, tax structure, farm size, legal organization, and risk and
uncertainty are discussed. Then a broadened cost effectiveness model
that incorporates the role of selected financial factors in motivating
irrigation return flow management in Washington's Columbia Basin is
presented. Although BMPs for irrigation return flow may address the control
of several pollutants, this discussion focuses on the control of sediment
loss from irrigated fields.
A Traditional View of Cost Effectiveness
Traditional applications of cost effectiveness, whether in irrigation
return flow management or other agricultural settings, are constructed
around single period models of a "typical" or "representative" profit
35
-------
maximizing farm. The applications generally proceed in two stages. First,
attention is devoted to the farm's production choices and its rate of
resource use per unit of production activity. Production activities that
maximize net returns to land, overhead, and management are selected subject
to various combinations of inputs such as labor, fertilizer, energy, and
equipment. A profit maximizing plan is developed without regard to
resulting environmental damage. Third party environmental spillovers,
e.g. sediment and/or nutrient losses from runoff and deep percolation, are
considered only in a bookkeeping sense. Estimates of gross outflows of
such pollutants are made for the current profit maximizing plan. The first
stage analysis yields a baseline or benchmark solution for comparison with
stage two solutions in which alternative pollution abatement practices are
implemented. Cost effectiveness estimates are then derived as differences
in net farm income (relative to the benchmark solution) divided by abatement
levels achieved.
This two stage approach is conceptually appealing since it is
rigorously grounded in economic theory and is analytically tractable given
the generally available tools and data. Moreover, the findings are
readily understandable by farmers and policy makers alike. However, this
•traditional analytical framework forces the problem into a convenient but
somewhat artificial model structure. By confining the producer's adoption
or non-adoption decision into a static (single period) concept of production
efficiency, abatement prescriptions may seriously err. Not only might the
cost effectiveness ranking of control practices differ from the order in
which they would be adopted, but the actual cost of implementing a given
practice might be overestimated. An otherwise "acceptable" farming
practice may fail to receive warranted consideration.
Factors Affecting Adoption
Considerations such as farm size, financial structure, legal organ-
ization, tax treatment, managerial objectives and abilities, and the
ability to bear risk are prominent factors in the ultimate choice and speed
of adopting pollution control practices. A common thread links each of
these factors; adoption or nonadoption decisions are time-dependent
resource allocation problems. Thus, by blending short-run production
theory with the longer-run theory of investment, greater insight into the
adoption process might be gained. Accordingly, the relatively rich body of
farm growth research offers a rigorous base for developing more realistic
models of cost effective environmental quality control.*
Alternative Farm Objectives--
Managerial objectives differ widely throughout agriculture, though
three managerial objectives have received greatest attention by economists:
(1) maximization of net worth, (2) maximization of disposible income, and
*A particularly good summary of farm growth research is provided in
Economic Growth of the Agricultural Firm, Tech. Bull. 86, College of
Agriculture Research Center, Washington State University, 1977.
36
-------
(3) maximization of multidimensional utility function or hierarchical goal
structure. By far the most common objective is maximizing the present
value of net worth.** This objective is essentially a broadened concept
of profit maximization over time, in which the firm's equity is allowed to
expand. Unlike the simple profit maximizing approach, the net worth
objective permits accumulation both of net income and wealth in the form
of durables like land and equipment. The explicit treatment of investment
renders the net worth objective more appropriate than short-run profit
maximization when modeling the adoption of structural pollution abatement
practices requiring substantial investment.
Accumulation of investment and equity over time raises the important
question of an appropriate length (horizon) for the planning process. In
the context of 208 programs, a relatively short horizon seems most appro-
priate. Policy makers rarely are interested in the continuum of production
adjustments and investment decisions throughout a farm's life, particularly
when cost sharing may be involved. Instead, their concern is with
immediate adjustment prospects and impacts. Likewise, farmers confronted
with an adoption decision are more interested in the immediate issues of
opportunity cost and financial solvency. Accordingly, a relatively short,
five to ten year, planning horizon seems most appropriate. This time
framework coincides with most finance periods, enabling linkage of annual
production decisions with longer run investment decisions.
The tradeoff between consumption and asset accumulation has raised a
controversy in the farm growth literature which has some implication with
respect to water quality improvement programs. For example, maximizing the
present value of disposable income or consumption over time may, in some
situations, be a more realistic determinant of actual managerial objectives.
Factors such as current income, age of household head, number of persons in
the household, change in net worth from the previous year, and annuitized
value of total assets all condition the importance of disposable income
vis-a-vis investing in environmental quality controls. Farmers of advanced
age or with large families are more likely to be concerned with disposable
income (financial liquidity) than with net worth. Consequently, they may
view structural control practices unfavorably with respect to the effects
on disposable income. In contrast, young farmers or those with small
family obligations typically are more concerned with accumulation of
productive assets and wealth. Structural practices are usually preferred
by such individuals, other factors being equal.
Tangible and intangible goals unquestionably strike at the heart of
BMP adoption motivations. Accordingly, both the net worth or consumption
maximization objectives can be criticized for compressing into a single
dimension what in reality is more likely to be an extremely complex multi-
dimensional set of goals. The desire for leisure, family needs, or a
better tractor than the neighbor's can each affect farmers' attitudes
**A number of variants on the net worth theme can be found in the farm
growth literature, most of which addresses the flow of funds between
years and the investment in durables.
37
-------
towards adoption. However, modeling goal structure is so complex and farm
specific as to be of little use in motivating widespread adoption. Policy
makers need broad quidelines to analyze the question of control practice
adoption, not farm-specific analyses for individual farmers. The infinite
variation in goals and goal ranking associated with individual management
decisions is completely impractical from a policy standpoint.
Because irrigation return flow management can take many forms, ranging
from subtle, non-structural changes in cultural practices to sizeable
capital commitments in structural controls, the net worth objective appears
to be preferable. The maximum net worth function focusses greater attention
on credit usage and financing. Moreover, consumption can serve as a
constraint on asset accumulation.
Financial Structure—
Acquisition of specific control practices, and particularly structural
ones, depends not only upon their impact on net revenues and net worth, but
also on the magnitude, structure, and liquidity of a farmer's assets and
liabilities. Thus, the financial structure of the farm is a critical
consideration in the ability to adopt a particular technology. Control
practices can be acquired via retained earnings (cash), credit, lease, or
even reduction of debt. The extent to which these sources of finance are
available to a farm effectively limits the number of relevant pollution
abatement practices. Cash flow becomes an overwhelming consideration. The
ability to secure necessary down payments for the purchase of structural
control devices; the ability to meet annual obligations including household
expenses and repayment of loans; the ability to purchase annual operating
inputs; and the ability to maintain asset liquidity all influence a
manager's decisions to adopt pollution abatement practices. These factors
can also be significant considerations in policies designed to induce
adoption of particular control practices. Consider the situaton in which
a down payment is the limiting factor. Either a guaranteed loan or a loan
on the downpayment might be all that is necessary to induce adoption.
Tax Structure, Farm Size, and Legal Organization—
Effective tax management is an increasingly important consideration in
the adoption of agricultural technologies. Decisions to incur large capital
expenditures on pollution abatement, in contrast with nonstructural short
term control methods, cannot be divorced from tax implications. While
both qualify as deductible expenses, federal income tax statutes favor
structural control practices. In particular, structural practices benefit
from tax deferral, investment credits, and conversion of ordinary income
to capital gains.
Income tax deferment from one tax period to another through accelerated
depreciation increases short term cash availability. Thus, tax deferral can
contribute to liquidity in early (and often critical) cash flow periods.
Investment in most depreciable property with a useful life of three or more
years, e.g. most structural control devices, is also eligible for an
investment credit—a dollar for dollar reduction in tax liability. Cur-
rently, the maximum investment credit rate is 10 percent of the qualifying
amount. Those investments that increase property value may also be
38
-------
eligible for preferential tax treatment at the time of asset disposal.
Only 40% of capital gains are taxable. Accordingly, after-tax profits can
differ dramatically from before-tax profits, thereby altering both the
absolute and relative cost effectiveness of alternative control measures.
The extent of benefit to be derived from alternative tax management
strategies (excluding investment credit) is dependent upon the level of
farm income which varies directly with farm size, apart from any
consideration of economies of size. Given the progressive income tax
structure and regressive treatment of expenses, larger farms benefit more
from available tax advantages than do smaller farms with less income.
Consider, for example, the after-tax cost of one dollar of expense on high
and low income brackets. The after-tax cost in a 50 percent tax bracket is
only 50 cents, whereas it is 70 cents in a 30 percent bracket, and one
dollar if no taxes are being paid. The impact is clear; cost-effectiveness
of control practices may differ by farm size. Because of tax structure,
more costly (before-tax) practices may in fact be more cost-effective on an
after-tax basis for larger farms.
Capital-intensive control practices may be more attractive to large
farms for other reasons. Large, capital-intensive farms may view labor-
intensive control practices as unacceptable because of tight labor con-
straints or increased managerial difficulties attending an enlarged labor
force.
Debt/equity position directly influences the percent of expenses that
are deductible in otherwise comparable farm scenarios. The higher the
equity position the greater the deduction benefits.
A final consideration in assessing the role of income taxes on
adoption of abatement practices is the legal organization of the farm.
Though single proprietorship remains the dominant legal organization in
commercial agriculture, incorporation is becoming increasingly prevalent.
Apart from the legal benefits, like sheltering personal assets from extensive
liability, corporations enjoy special tax treatment unavailable to single
proprietorships. This differential tax treatment comes in a variety of
forms, but most pronounced are different statutory tax rate limits and the
ability to shelter income.* Regardless of the tax advantages, the potential
after-tax income remains a fundamental concern—not minimization of taxes.
Risk and Uncertainity--
Regardless of the type of control practices under consideration, an
important factor affecting adoption is the risk or uncertainty which
producers perceive as being associated with the practice. Four main
sources of risk and uncertainty influence managerial decision making in
"Corporate tax rates are relatively flat, in contrast to the progressive
individual rates. Corporations face a maximum tax rate of 46 percent,
whereas single proprietorships are taxed at a current maximum rate of 50
percent. Corporations also can shelter income from taxation with
retaining earnings.
39
-------
agriculture: (1) changes in technology, (2) changes in the legal and insti-
tutional setting, (3) yield variability, and (4) price variability. Price
variability poses the greatest source of risk to farmers in irrigated agri-
culture. Cash flow problems, for example, can become severely aggravated
by price fluctuations. Inability to service debt may threaten the economic
viability of the farm, a somewhat more likely situation for the low equity
farmer. However, insofar as structural practices cut down on production
risks, a preference for structural practices might exist.
Despite the potentially profound role of risk on adoption motivations,
incorporating risk into adoption models does not appear to be practical at
this time. Perceptions of risk, as well as abilities and desires to bear .
perceived risks, vary widely among individual producers.
40
-------
SECTION 7
STRUCTURAL CHANGES
As was explained in the Introduction of this report, a cost sharing
program was established for the Block 86 study area to provide a means of
constructing facilities on the participating farmer's land. Full ownership
of these facilities rests with the land owner. The cost share contributed
by the research grant was viewed as an incentive for the farmer's
participation and an aid to providing facilities necessary for the research
work. All facilities were designed by the faculty members of Washington
State University who were associated with the research project.
The on-farm facilities included buried pipelines, concrete-lined head
ditches, gated pipe systems, sprinkler systems (both center-pivot and solid-
set), and sediment basins. Table 5 lists the details of most of these
changes and Table 6 completes the details by describing the sediment basins
including the basins constructed by the farmers without cost-share funding.
A total of $69,825 of research grant funds were spent on cost-sharing for
construction of on-farm facilities. Since certain areas received benefit
from more than one of the facilities listed in Tables 5 and 6, a cost per
hectare directly benefited is difficult to obtain.
In addition to giving details of sediment basin construction within
the study area, Table 6 gives results of records kept on the degree to
which sediment basins filled and dates on which farmers cleaned the basins.
The two sediment basins which have two dates of construction listed in
Table 6 each had experienced difficulties during the 1979 irrigation season
attributed to design and/or construction deficiencies. Hence, these two
basins were re-constructed before the 1980 irrigation season.
41
-------
TABLE 5. DETAILS OF FACILITIES CONSTRUCTED ON PARTICIPATING FARMER'S LAND
ro
Type of
Change
Buried Pipeline
Buried Pipeline
Buried Pipeline
Buried Pipeline
Concrete-Lined
Head Ditch
Gated Pipe
Center-Pivot
Sprinkler Systems
Solid Set
Farm
Unit No.
40,41,103
42,43,44,
45,48
43,107
64
54,64,65
43,46,47
48, 53
44,45,81
107
Benefit To
Approximate
Area (hectare)
72
110
20
10
80
88
70
9
Comments
Conveys center pivot overflow
and tailwater to improved drain
Conveys center pivot overflow
and tailwater to improved drain
Replaced irrigation lateral W69A
Replaced ditch through field to
supply new concrete lined head
ditch
Replaced earthen head ditches
Replaced earthen head ditches
Replaced furrow irrigation on
steep slopes
Replaced furrow irrigation in
Sprinkler System
clean-cultivated orchard
-------
TABLE 6. DETAILS OF SEDIMENT BASINS IN OPERATION DURING PROJECT PERIOD
OO
Basin
No.
1
2
3
4
5
6
7
8
g**
Farm
Unit No.
49
50
48
53
53
53
65
64
57
Date of
Construction*
3/19/79
3/23/79
3/20/79
5/11/79
4/24/80
5/11/79
4/26/80
4/28/80
4/10/79
4/16/79
3/30/79
30.5
35 x
33.5
35 x
36.6
36.6
42.7
15.2
30.5
21.3
48.8
Size
(m)
x 7.6
7.6 x
x 6.1
8.2 x
x 9.1
x 9.1
x 9.1
x 7.6
x 7.6
x 4.6
x 12.
x 1.2
1.2
x 1.2
1.2
x 1.2
x 1.2
x 1.2
x 1.2
x 1.2
x 1.2
2 x 1.2
Delivery
Area
(hectare)
84
50
5.3
16
19
3.6
37
22
91
Basin
% Full
95
100
100
20
25
40
5
10
100
60
100
80
5
10
60
50
80
50
Capacity
Date
7/23/79
3/7/80
6/6/80
10/25/80
3/7/80
10/10/80
3/7/80
10/10/80
3/7/80
8/18/80
3/7/80
8/18/80
8/18/80
3/7/80
6/15/80
3/7/80
10/14/80
3/7/80
Cleanout
Date
8/17/79
4/7/80
8/12/80
4/7/80
8/5/80 (Partly)
2/10/81
4/24/80
3/2/81
4/26/80
3/2/81.
-------
TABLE 6 (Continued)
Basin
No.
10
11
12
13
14
15
16
17
18
19
Farm
Unit No.
54
41
41
107
47
57
57
57
57
57
Del ivery
Date of Size Area
Construction* (m) (hectare)
4/1/79
4/4/79
4/4/79
3/17/79
v
4/7/80
Before
Project
Before
Project
4/2/80
4/2/80
4/2/80
30.5 x 6.1 x 1.2
18.3 x 4.6 x 1.2
18.3 x 4.6 x 1.2
30.5 x 7.6 x 1.2
15.2 x 4.6 x 1.2
19.8 x 15.2 x 1.2
18.3 x 12.2 x 1.2
36.6 x 6.1 x 1.2
18.3 x 6.1 x 1.2
110 x 6.1 x 1.2
20
55
55
—
38
91
91
91
91
91
Basin
% Full
30
85
90
5
100
0
0
0
0
100
10
75
10
70
100
90
50
Capacity riPinn.it
LI eanou L
Date Date
3/7/80 2/12/81 (half)
6/15/80
10/14/80
3/7/80
10/17/80
3/7/80
10/17/80
3/7/80
10/15/80
8/1/80 3/3/81
3/7/80
10/13/80
3/7/80
10/13/80
10/13/80
10/13/80
10/13/80
-------
TABLE 6 (Continued)
Basin
No.
20
21
Farm
Unit No.
55
56
Date of
Construction*
4/2/80
4/1/80
Size
(m)
24.4 x 4.6 x 1.2
12.2 x 3 x 1.2
Del ivery
Area
(hectare)
40
25
Basin
% Full
100
100
100
Capacity
Date
5/5/80
10/14/80
5/5/80
Cleanout
Date
9/25/80
*Basin No. 1 through 13 were either originally constructed or cleaned on the date indicated using
grant funds for cost-sharing. Basin No. 14 through 21 were constructed by the farmers without
assistance of project funds or personnel.
**This basin was filled in mechanically early in the 1980 irrigation season because of a dispute
between adjacent land owners over property boundary location.
-------
SECTION 8
RESULTS AND DISCUSSION
The previously described 800-hectare irrigated tract was monitored
from 1977 through 1981 to evaluate the magnitude of existing pollution from
irrigation return flows, and to determine the effects of on-farm implemen-
tation of best management practices (BMP) designed to reduce sediment and
nutrient losses. During 1977 and 1978 data were obtained mainly for
comparative purposes and to design subsequent structural improvements.
During the winter of 1978-1979 structural changes were made, including:
construction of additional sediment basins and redesign of existing basins,
construction of check dams in eroding gulleys of some farm units, conversion
to center-pivot and solid-set sprinkler systems, installation of buried
pipes to convey center-pivot overflows directly to drains, installation of
a gated pipe on-farm delivery system, and installation of additional
concrete-1ined head ditches. In addition, non-structural management
practices were evaluated, including: use of non-erosive stream sizes,
irrigation scheduling, use of cutback irrigation techniques, and more
carefully controlled water application to improve on-farm irrigation
efficiency. Results of the monitoring programs are presented from (1)
area-wide (2) individual-field and (3) individual-furrow viewpoints.
AREA-WIDE SEDIMENT AND NUTRIENT LOSSES
Sediment and nutrient losses from the study area's main drains were
monitored continuously throughout the study period. Sediment and phosphorus
losses from 1977 through 1981, as calculated from a mass balance approach,
are presented in Tables 7-11. These tables indicate significant reductions
in sediment losses throughout the project period, and particularly after
construction of improvements. Respective reductions in net sediment loss
from 1977 to 1978, 1979, 1980 and 1981 were 42%, 81%, 89% and 77%. Hence,
the decrease of sediment discharge between 1978 and 1979, when most
structural improvements were implemented, was particularly dramatic. The
slight increase of sediment discharge for the 1981 irrigation season can be
attributed to the planting of somewhat more erosive row crops. Average
trends for 1977-78 as compared with 1979-81 are shown in Figure 5. This
figure also shows that irrigation return flows decreased an average of
only 2.2% after implementing the structural changes, which demonstrates the
effectiveness of the constructed improvements even more.
Relative reduction in sediment and phosphorus loss before and after
construction of structural improvements are presented in Table 12. While
46
-------
TABLE 7. MONTHLY MASS BALANCE FOR SEDIMENT AND TOTAL PHOSPHORUS, 1977 IRRIGATION SEASON
Month
April*
May
June
July
August
Sept.
Oct.*
Total
Water
Diverted
153.4
148.3
203.8
229.5
226.0
83.1
36.4
1080.5
Water
Discharged
(ha-m)
66.0
81.0
82.1
98.1
106.1
78.9
38.7
550.9
Sediment
Delivered
89.3
36.2
140.2
116.0
95.4
2.4
1.1
480.6
Sediment
Discharged
(mt)
1313.6
939.8
953.8
839.1
612.3
145.2
63.0
4866.8
Net
Loss
1224.3
903.6
813.6
723.1
516.9
142.8
61.9
4386.2
Total P
Del i vered
97.6
56.7
143.8
104.3
73.0
15.4
7.3
498.1
Total P
Discharged
(kg)
752.1
606.4
680.8
559.0
503.3
169.5
93.8
Net
Loss
654.5
549.7
537.0
454.7
430.3
154.1
86.5
3364.9 2866.8
Irrigation season was from April 1 to October 20.
-------
TABLE 8. MONTHLY MASS BALANCE FOR SEDIMENT AND TOTAL PHOSPHORUS, 1978 IRRIGATION SEASON
oo
Month
March*
Apri 1
May
June
July
August
Sept.
Oct.*
TOTALS
Water
Diverted Di
(ha-m)
4.2
81.4
144.7
257.4
223.9
176.6
84.7
49.8
1022.7
Water
scharged
5.2t
51.2
67.0
100.8
98.7
86.1
68.4
52. 9t
530.3
Sediment
Del ivered
0.4
12.2
89.8
169.7
73.4
80.5
1.7
1.6
429.3
Sediment
Discharged
(mt)
0.5
118.5
228.2
684.5
1226.9
536.8
95.9
93.7
2985.0
Net
Loss
0.1
106.3
138.4
514.8
1153.5
456.3
94.2
92.1
2555.7
Total P
Del ivered
1.3
29.8
74.7
151. 8
93.4
65.8
14.1
9.6
440.5
Total P
Discharged
(kg)
10.4
166.8
280.7
646.4
1092.7
452.5
102.1
111.2
2862.8
Net
Loss
9.1
137.0
206.0
494.6
999.3
386.7
88.0
101.6
2422.3
^Irrigation season was from March 27 to October 26.
tin some instances water discharged is greater than water diverted, due to the influence of
groundwater.
-------
TABLE 9. MONTHLY MASS BALANCE FOR SEDIMENT AND TOTAL PHOSPHORUS, 1979 IRRIGATION SEASON
Month
April*
May
June
July
August
Sept.
Oct.*
TOTALS
Water
Diverted Di
(ha-m)
50.4
209.7
248.1
215.5
173.0
71.0
51.5
1019.2
Water
scharged
33. 5t
85.1
105.7
102.8
106.7
56.lt
51. 7t
541.6
Sediment
Del ivered
7,1
185.6
172.3
50.9
37.3
2.4
2.4
458.0
Sediment
Discharged
(nit)
26.2
236.7
363.0
379.9
228.5
28.9
37.5
1300.7
Net
Loss
19.1
51.1
190.7
329.0
191.2
26.5
35.1
842.7
Total P
Delivered
11.0
140.0
99.5
57.8
62.8
18.4
7.0
396.5
Total P
Discharged
(kg)
31.5
240.4
543.4
555.2
254.6
58.5
85.8
1769.4
Net
Loss
20.5
100.4
443.9
497.4
191.8
40.1
78.8
1372.9
^Irrigation season was from April 5 to October 27.
tGroundwater contributes significantly to discharge.
-------
TABLE 10. MONTHLY MASS BALANCE FOR SEDIMENT AND TOTAL PHOSPHORUS, 1980 IRRIGATION SEASON
en
o
Month
April*
May
June
July
August
Sept.
Oct.*
TOTALS
Water
Diverted Di
(ha-m)
64.6
149.0
158.8
201.6
170.8
77.9
40.2
862.9
Water
scharged
44.9
78.2
96.0
98.1
97.1
72. 2t
39.lt
525.6
Sediment
Del i vered
16.9
274.1
94.5
160.4
58.5
5.3
1.7
611.4
Sediment
Discharged
(mt)
55.5
373.2
263.5
256.5
127.8
26.5
9.7
I
1112.7
Net
Loss
38.6
99.1
169.0
96.1
69.3
21.2
8.0
501.3
Total P
Del i vered
34.4
145.1
58.4
94.6
56.1
27.5
17.3
433.4
Total P
Di scharged
(kg)
117.2
400.5
418.7
398.3
225.9
64.6
59.8
1685.0
Net
Loss
82.8
255.4
360.3
303.7
169.8
37.1
42.5
1251.6
^Irrigation season was from April 1 to October 25.
tGroundwater contributed significantly to discharge.
-------
TABLE 11. MONTHLY MASS BALANCE FOR SEDIMENT AND TOTAL PHOSPHORUS, 1981 IRRIGATION SEASON
Month
March*
April
May
June
July
August
Sept.
Oct.*
TOTALS
Water
Diverted Di
(ha-m)
7.7
97.3
130.5
184.2
204.5
193.5
91.3
35.2
944.2
Water
scharged
' 7.5|
56.4
66.8
80.4
98.1
94.7
65.9
36.lt
505.9
Sediment
Del i vered
2.5
64.2
76.0
115.4
94.3
52.9
8.6
2.4
416.3
Sediment
Discharged
(mt)
5.0
44.8
349.8
227.6
415.9
205.8
143.5
28.2
1420.6
Net
Loss
2.5
-19.4
273.8
112.2
321.6
152.9
134.9
25.8
1004.3
Total P
Delivered
3.7
63.7
105.3
111.8
82.4
73.0
19.3
4.0
463.2
Total P
Discharged
(kg)
13.8
144.7
318.3
333.2
590.6
272.2
165.7
28.6
1867.1
Net
Loss
10.1
81.0
213.0
221.4
508.2
199.2
146.4
24.6
1403.9
^Irrigation season was from March 25 to October 22.
tGroundwater contributed significantly to discharge.
-------
l/l
I/)
O
C
O)
CL>
400
300
200
100
0
— Avg. 1977-1978
.__ Avg. 1979-1981
25
20
15
10
Apr May Jun
fO
O
c
Jul Aug Sep Oct
Month
Figure 5. Average weekly runoff and net sediment loss for the study area, 1977-1981.
-------
TABLE 12. RELATIVE CHANGES IN NET SEDIMENT AND PHOSPHORUS LOSSES FROM
1977-1978 TO 1979-1981
% Reduction % Reduction
Year compared to 1977 over previous year
1977
1978 42 42
Sediment 1979 81 67
1980 89 41
1981 77 -100*
Phosphorus
1977
1978
1979
1980
1981
—
16
52
56
51
— —
16
43
9
-12*
^Negative sign indicates a net gain 'over the previous year.
reductions in phosphorus loss were also generally significant, it appears
that measures which controlled sediment losses were not equally effective
in controlling phosphorus losses as well. This is probably because
phosphorus is associated with clay-sized particles which remain suspended
in runoff waters during flow through typical sediment basins. Thus, while
substantial amounts of sediment were prevented from leaving the study area,
only minor regulation of phosphorus losses was accomplished. This conclusion
is supported by the observation that end-of-field sediment/phosphorus ratios
were often around 1500, whereas similar ratios measured at the main drain
for the total study area were only about half this value during 1979-1981.
In accounting for the dramatic decline in net sediment loss between
1978 and 1979, additional factors should be mentioned. As mentioned above,
many structural improvements were added during the 1978-79 overwinter
period. Not to be overlooked, in addition, is the psychological effect of
a well-publicized monitoring program. This may have oriented some growers
toward more careful irrigation patterns irrespective of any improvements in
physical facilities. Following the 1978 season, a further change occurred
when sugar beets were taken out of production in the Columbia Basin. Sugar
53
-------
beets are an erosive crop because of their need for frequent cultivation,
thinning and heavy prolonged watering.
Between 1979 and 1981 no new structural improvements were added to the
study area. Sediment losses for 1980 were still somewhat less than for
1979 (Table 12), despite a slight increase in the area cropped to row
crops, and the effects of the Mt. St. Helen's ashfall (which deposited 2 to
3 cm of unconsolidated ash throughout the study area). Both of these
factors were apparently offset by more careful water management. Changes
in net phosphorus loss from the study area from 1979 to 1980 to 1981 were
minor (Table 12).
There is an important point to remember when studying the data of
Tables 7 through 11. Net sediment and phosphorus losses reported are
primarily from the surface-irrigated fields of the study area. Since 34.4%
of the study area was irrigated using overhead sprinklers (mostly center-
pivots) by 1979, reported sediment losses represent only 505 hectares of
the 769 hectare study area. Figure 6 shows that the average overall
reduction (1977-78 to 1979-81) was by a factor of 4.6 for sediment and 2.2
for phosphorus, whereas cultivated area during the same period remained
essentially constant. Judging from such data, a successful program for
reducing sediment discharge was accomplished through a combination of the
above changes.
Although decreases in sediment and phosphorus losses from the study
area were impressive, total water diversion also decreased over the study
5000
2500
-o
d)
00
(4472)(2903)
4386 . 2867
(2463)
2422
Sediment
r} Phosphorus
-.3000
(2578)
2556
(1284)
1373
(835)
843
A
(1152)
1252
(1261)
1404
(506)
501
(976)
1004
2000
1000
CD
in
3
S-
o
O.
O
a.
1977
1978
1979
1980
1981
Figure 6.
Net sediment and phosphorus losses from study area, 1977-1981.
Note that the values in parentheses were corrected by the USGS
to account for errors in compositing samples.
54
-------
period. Table 13 shows the trend in irrigation diversions from 1977 through
1981. Since the area under irrigation remained virtually constant, more
careful water management and annual weather patterns (e.g., the cool, wet
summer of 1980) also contributed to these reductions for the study area.
Figures 7 through 10 illustrate the distribution of sediment loss and
drainageway discharge throughout the 1978-1981 irrigation seasons. Again,
it is. evident that there was a dramatic decline in sediment loss between
1978 and 1979. Total (area-wide) losses were heaviest during the early
summer months, though individual fields were generally most erosive during
heavy early-season irrigations following cultivation. Drainageway
sediment-loss patterns were strongly influenced by degree of sediment basin
filling as the season progressed. The pattern for 1980 evidences an
interesting variation. Figure 9 shows a prominant spike in sediment
discharge during late May. This spike represents erosion of recently
deposited Mt. St. Helen's ash. Figure 11 shows that approximately 180
metric tons of additional sediment and about 200 kg of additional
phosphorus were transported out of the study area during the first two
weeks following the eruption. This immediate loss represented <0.2% of the
more than 90,000 mt of ash which fell on the 800-ha study area (USDA-SCS,
1980). Most of the ashfall was incorporated during subsequent cultivations,
or was lost as intermingled soil- and ash-derived sediments. The early
season timing of the ashfall led to much of its retention by unfilled
sediment basins of the study area.
During the week of May 25, 1980, many growers attempted to alleviate
problems associated with the ash through cultivation or even replanting.
Many growers also felt that irrigation would somehow rectify the problem.
As a result, considerable sediment loss occurred during the week of May 25.
Effects of the ash on phosphorus loss appeared more complicated. Table 14
TABLE 13
STUDY AREA IRRIGATION DIVERSION
SUMMARY, 1977-1981
Year Total Irrigation Diversions
(ha-m)
1977 1081
1978 1023
1979 1019
1980 863
1981 944
55
-------
U1
en
2
^
i
.c
o
c
Qi
Apr
-300
-200
-100
C
CD
E
•r—
•a
01
Figure 7. Seasonal distribution of main drain discharge (runoff) and area-wide net sediment
losses, 1978.
-------
en
to
M-
O
c
13
o:
24 -i
18 -
12 -
6 -
Apr
-150
C
CD
Ol
I/O
Oct
Figure 8. Seasonal distribution of main drain discharge (runoff) and area-wide net sediment
losses, 1979.
-------
24 H
en
oo
to
.c
o
c
18 -
12 -
6 -
Runoff
Apr May
Jun
Jul
Month
T I
Aug Sep Oct
-150
-100
-50
c.
OJ
T3
Ol
1/1
Figure 9. Seasonal distribution of main drain discharge (runoff) and area-wide net sediment
losses, 1980.
-------
Runoff
en
vo
I
to
o
c
May
•o
Ol
CO
-50
Jun
Jul Aug
Month
Sep Oct
Figure 10. Seasonal distribution of main drain discharge (runoff) and area-wide net
sediment losses, 1981.
-------
120 r-
c
O)
-5
0)
I/O
60
0
115
173
82
28
8
zz.
Sediment
LJ Phosphorus
80
42
31
200
100
CD
s-
o
.c
a.
o
jz
a.
0
May 18 May 25
Week of
June 1
May 11
Figure 11. Net sediment and phosphorus main-drain discharges, spring 1980.
presents net weekly sediment and phosphorus losses (and corresponding
sediment/phosphorus ratios) from the study area during the spring, 1980.
Because of the sharply lower phosphorus content of the ash than that of
typical sediments eroded from the study area, the sediment/phosphorus ratio
was perceived as a useful index to ash content of subsequently eroded
materials. Phosphorus losses increased sharply the week of May 18, so the
average sediment/phosphorus ratio for that week remained similar to values
observed during the previous four weeks. During the week of May 25 the
average sediment/phosphorus ratio climbed 303% above the average for the
previous week, however, while net sediment losses increased only 40%. The
following week sediment losses decreased 64%, while sediment/phosphorus
ratios remained high. By the week of June 8 sediment/phosphorus ratios
seemed largely back to normal, with the relatively large phosphorus losses
during the weeks of June 8 and 15 accompanied by more typically mid-season
levels of sediment loss.
Since apparent effects of the ashfall on sediment and phosphorus
losses were most pronounced during the first three weeks following the
eruption, a probable re-construction of events could be as follows: Sediment
and phosphorus increases during the week of May 18 were primarily ash-related.
60
-------
TABLE 14. STUDY AREA NET SEDIMENT AND PHOSPHORUS LOSSES, SPRING 1980
Week
of
Apr
May
June
20
27
4
11
18
25
1
8
15
Sediment
(mt)
13.
3.
-48.
8.
82.
114.
41.
53.
38.
27
21
44*
04
07
81
90
43
87
Phosphorus
(kg)
26.
10.
20.
28.
173.
79.
30.
102.
131.
76
44
67
32
24
97
72
10
70
Sediment/Phosphorus
Ratio
478
307
658**
284 "
474
1435
1364
523
295
*
During the week of May 4, inflow of sediment (via West Canal irrigation
diversions) totaled 62.05 mt. Since only 13.60 mt of sediment were
measured leaving the study area during this week, the net loss was
computed as -48.44 mt.
**
Since 20.67 kg of phosphorus were transported out of the study area
during the week of May 4, this sediment/phosphorus ratio is based only on
the accompanying 13.60 mt of sediment transported out of the study area.
Net loss of sediment would undoubtably have been much greater were it not
for considerable retention of eroded soil and ash in the area's sediment
basins. Upon first consideration, it might appear that much of the
phosphorus of the ash (Fruchter et al., 1980 reported as much as 3200 ppm
PpO,- in ash from the Moses Lake region) was transported out of the study
area during the first week following the eruption. Weather records indicate
that 2.2 cm of precipitation fell during the period May 18-June 28, 1980.
This rainfall, along with post-ashfall irrigation, could have transported
considerable amounts of ash-derived phosphorus to the area's drainageways.
Early reports on soluble nutrients contained in the ash, however, did not
list phosphorus as a water-soluble component (Fruchter et al., 1980). Another
possibility is that much of this "phosphorus" was not really phosphorus at
all. Soluble silica is also determined by the molybdenum blue method.
Relatively high contents of water-soluble silica in the freshly formed
61
-------
and/or fractured ash particles could also have leached during subsequent
rainfall and found its way to the area's drainageways.
During the weeks of May 25 and June 1 most growers attempted to
incorporate the ash, which had formed a somewhat impervious layer in
irrigation furrows. Though water diversions were cut back sharply during
this period, those irrigations which did occur took place on extremely
erodable (recently cultivated) fields. The increased sediment/phosphorus
ratios during the weeks of May 25 and June 1 suggest a large contribution
of relatively phosphorus-free ash to the eroding sediments. By the weeks
of June 8 and 15 water diversions were nearly back to normal. The large
phosphorus losses which were maintained during these two weeks (Table 14)
imply considerable erosion as well. Apparently the area's sediment basins
remained effective in retaining considerable amounts of eroded material,
whereas phosphorus retention was much less successful.
Leached ash contained only about 1% to 2% as much total phosphorus as
did surface soils of the study area. Table 15 shows that average sediment/
phosphorus ratios for runoff from seven fields increased 2.7-fold during
the 10 days immediately following the May 18 eruption. This would
correspond to approximately 60% ash among sediments eroded from these
fields during this period. As the ash was incorporated, this ratio
eventually decreased once more. By monitoring subsequent changes in
sediment/phosphorus ratios throughout the irrigation season, the relative
contribution of ash to sediment losses could be assessed (Peterson, 1982).
Figure 9 also evidenced a slight increase in sediment loss during late
July and early August of 1980. Though the amount involved may not have
seemed significant, this peak was associated with a dredging operation on
the study-area's drainageways. Approximately 55 mt of additional sediment
were transported out of the study area during this operation.
TABLE 15
SEDIMENT/PHOSPHORUS RATIOS IN THE STUDY AREA
RUNOFF AS AFFECTED BY THE MT. ST. HELENS ASHFALL
Average Ratio 10 Average Ratio for the First
Days Prior to 10 Days Following
Flume May 18, 1980 May 18, 1980
10 1680 2806
11 1332 3492
12 1471 2764
62
-------
Cumulative sediment losses from individual fields within the study
area were considerably larger than area-wide sediment losses. Sediment
entrapment basins, when properly managed, may reduce the mass of sediments
transported, from a field by as much as 90%. Of the 769 ha under
cultivation, 505 were surface-irrigated. In 1979, flume-monitored fields
totaled 167 ha, and these.fields lost an estimated 1208 mt of sediment. In
1980, monitored fields totaled 146 ha and lost an estimated 897 mt of
sediment. These values could be used to estimate area-wide sediment losses
via a simple ratio and proportion approach, and could also be inserted into
appropriate regression equations in order to predict phosphorus loss.
Sediment and phosphorus losses for 3 years as calculated by the above
method were compared with U.S.G.S. measurements at the sampling station on
the main drainageway. Results are shown in Table 16. Values under the
heading "WSU" represent estimates of area-wide sediment and phosphorus
losses as extrapolated from flume records for those surface-irrigated
fields which were monitored. Values shown under the heading "USGS"
represent actual measurements of net sediment and phosphorus losses from
the study area via the main drainageway. Differences between corresponding
values for a given year are a measure of the amounts of sediment or phos-
phorus retained in study-area sediment basins and drainageways. From these
data, it appears that only about 20% of the sediment discharged from
surface-irrigated fields subsequently left the study area during either
1979 or 1980.
A second notable feature of Table 16 is the close agreement each year
in phosphorus-loss estimates from the two sources. Earlier, it was
proposed that phosphorus in irrigation runoff from fields of the study area
was largely unaffected by passage through the area's sediment basins, since
the phosphorus is generally associated with highly dispersed clay-sized
minerals. Agreement between the two estimates of phosphorus loss
TABLE 16
COMPARISON OF SEDIMENT AND
PHOSPHORUS LOSS ESTIMATES
1979 1980 1981
WSU USGS WSU USGS WSU USGS
Sediment 3653 843 3101 501 4048 1004
(mt)
Phosphorus 1326 1373 1398 1252 1289 1404
(kg)
63
-------
substantiates the claim that relatively little phosphorus is retained by
sediment basins of the study area.
INDIVIDUAL-FIELD SEDIMENT AND PHOSPHORUS LOSSES
Sediment Losses
During 1978 and 1979, from 12 to 15 monitoring sites (flumes) were
maintained on various fields of the Block 86 study area. Grab samples,
coupled with continuous records of flow through each flume, provided the
necessary data base for sediment loss estimates. In 1980, four additional
flumes were installed and two were removed, and two additional flumes were
removed during 1981. Tables 17 through 20 provide sediment-loss estimates
for the various flumes from 1978 through 1981. The data show a wide range
of sediment-loss values. Wheat fields generally exhibited the least, and
bean and corn fields the most, loss during this period.
Tables 21 and 22 show that variations in sediment loss for a given
crop were roughly correlated with field slope. The two wheat fields in
Table 21 represented by flumes 4 and 19 have unequal slopes and yet
comparable sediment losses. This implies that still another factor was
involved in this case. Since surface-soil texture for these two fields was
relatively uniform (both > 90% silt loam), a probable additional factor is
the water application rate, reflecting how well the irrigation water has
been managed. Table 23 illustrates that the field represented by flume 19
had a considerably greater proportion of irrigation runoff than did the
field represented by flume 4. Furthermore, the field having the higher
percent runoff had its greatest slope in the first 100 meters from the head
ditch, where the cumulative effect of 25 years of surface irrigation was
dramatically evident. At the top of this field, erosion has lowered the
soil surface up to 1 meter below the level of the concrete head ditch.
Much of the soil transported from this region has subsequently been
TABLE 17
SEDIMENT LOSS ESTIMATES FOR
INDIVIDUAL FIELDS, 1978
Flume
1
3
11
13
14
15
Crop
Wheat/mi nt
Corn
Sugarbeets
Beans
Corn
Corn
Area
(ha)
44.3
19.6
5.5
10.5
3.9
8.1
Sediment Loss
(mt/ha)
1
1
30
14
7
2
P Loss
(kg/ha)
1
0.6
13.0
11.0
7.0
2.0
64
-------
TABLE 18. SEDIMENT AND PHOSPHORUS LOSS ESTIMATES FOR INDIVIDUAL FIELDS,
1979
Flume
2
3
4
6
10
11
13
14
15
16
17
18
19
20
Crop
Corn
Corn
Wheat
Corn
Peas
Beans
Wheat
Beans
Beans
Wheat
Corn
Orchard
Wheat
Mint
Area
(ha)
16.1
19.1
14.6
9.5
4.1
5.5
10.5
3.9
8.1
8.8
19.3
10.7
25.0
14.9
Sediment Loss
(mt/ha)
18.0
21.4
3.5
2.9*
12.0
13.8
0.8
50.5
5.4
0.2
6.3
0.4
3.1*
2.1*
P Loss
(kg/ha)
7.1
11.2
1.0
0.9
7.3
5.5
0.7
20.2
2.2
0.3
2.6
0.2
1.7
1.3
*F1ume data were insufficient for detailed analysis. Figures in this
case are merely estimates from available data.
redeposited downfield, where the slope is less. Consequently, the 3.1
mt/ha estimated sediment loss for this field in 1979 would have been
substantially larger were it not for mid-field and end-of-field
redeposition of sediments.
Table 21 also shows that the field monitored by flume 14 was
particularly susceptible to erosion, in spite of its moderate (3.3%)'slope.
This may be due, in part, to over-irrigation, or to irrigation with
excessively large streams. There are actually pronounced multiple slopes
on this field, however, ranging from 2.2% to 4.1%. Hence, the 3.3% figure
represents only a weighted average for the entire field. Unfortunately,
65
-------
TABLE 19. SEDIMENT AND PHOSPHORUS LOSS ESTIMATES FOR INDIVIDUAL FIELDS,
1980
Flume
2
3
4
7
9
10
11
12
13
14
15
16
17
18
21
22
Crop
Corn
Corn
Beans
Wheat
Beans
Carrots
Wheat
Carrots
Peas
Wheat
Peas
Wheat
Wheat
Wheat
Wheat
Orchard
Asparagus
Carrots
Area
(ha)
16.1
19.1
14.6
5.9
4.4
4.1
5.5
76.4
10.5
3.9
8.1
8.8
19.3
10.7
9.9
5.7
Sediment Loss
(mt/ha)
15.1
5.8
5.8
3.6
19.5
11.9
3.8
6.3
4.1
11.6
0.8.
0.8
6.1
0.6
3.9
5.4
P Loss
(kg/ha)
8.8
5.6
3.5
0.9
14.2
4.5
1.1
3.8
2.1
5.3
0.5
0.4
3.5
0.3
3.5
2.7
the flume data could not be resolved further into sub-portions for areas
of different slopes.
The other bean and wheat fields of Table 21 showed more moderate
sediment loss. Corn, which is also an erosive crop, exhibited a
substantial range of sediment-loss values. The fields represented by
flumes 2 and 3 had a known infiltration problems, due to sodic-soil
conditions. Over-irrigation has been common for this particular farm unit.
66
-------
TABLE 20. SEDIMENT AND PHOSPHORUS LOSS ESTIMATES FOR INDIVIDUAL FIELDS,
1981
Flume
2
3
4
9
10
11
13
14
15
16
17
18
22
Crop
Corn
Corn
Wheat
Wheat
Wheat
Peas
Beans
Peas
Peas
Wheat/Peas
Wheat
Orchard
Wheat
Area
(ha)
16.1
19.1
14.7
4.4
4.1
5.5
10.5
3.9
8.1
8.8
19.4
10.7
5.7
Sediment Loss
(mt/ha)
13.6
22.3
0.81
5.10
4.5
5.8
9.86
8.20
2.56
0.50
1.63
3.03
0.63
P Loss
(kg/ha)
10.0
19.0
0.6
3.6
2.4
5.1
11.3
1.5
2.0
0.3
1.2
2.6
0.6
The amount of runoff from these fields was such that even moderate
accompanying sediment concentrations would produce appreciable amounts of
sediment loss.
More difficult to explain is the amount of erosion through flume 17 in
1979 compared to 1980. An irrigated wheat crop normally erodes considerably
less than an irrigated corn crop. The unanticipated high erosion rate for
this wheat crop may have resulted, in part, from exceptional amounts of
runoff. Table 24 shows that this field exhibited greater than 50% runoff
from 04 May through 07 June 1980. Earlier in the irrigation season an
unlined, earthen ditch had been constructed on the east side of this farm
unit to carry excess irrigation water (spillage) directly through the flume
and into the field's sediment basin. The unprotected soil in this waste
ditch appears to have contributed significantly to increased sediment
concentrations in the resultant runoff. In addition, volcanic ash from
67
-------
TABLE 21. FIELD SLOPE AND SEDIMENT LOSS, 1979.
Flume
Crop
Slope
Sediment Loss
(mt/ha)
17
6
2
3
15
11
14
16
13
4
19
Corn
Corn
Corn
Corn
Beans
Beans
Beans
Winter Wheat
Winter Wheat
Winter Wheat
Winter Wheat
1.4
1.5
2.5
3.0
1.5
3.1
3.3
1.0
2.1
3.4
4.3
6.3
2.9
18.0
21.4
5.4
13.8
50.5
0.2
0.8
3.5
3.1
Mt. St. Helens fell during this period, with a potential for increasing
runoff sediment concentrations even further. Other 1980 sediment-loss
estimates for spring wheat did not appear excessive, however. In 1981 this
field was again cropped to spring wheat, with resultant sediment loss
(approximately 1.6 mt/ha) much more in line with that for other spring wheat
crops of the study area.
Crop-specific relationships can be highlighted by comparing selected
fields on successive years. Table 25 shows such comparisons for 1978 through
1981. Some fields remained in the same crop during successive seasons,
whereas others were rotated to alternate crops. The corn crop represented
by flume 2 showed only a slight drop in erosion between 1979 and 1980-81,
whereas a marked drop in erosion was evidenced between 1979 and 1980 for
flume 3, representing a second field of the same farm unit. The main
difference in the latter case appeared to be the construction of a series
of check dams in 1980 at the end of this particular field. An additional
sediment basin (7.6 x 15.3 x 1.2 m) was also constructed on F.U. 53 in 1980.
Flow from this basin also discharged through flume 3, thus further reducing
measured sediment concentrations. These structures were missing from the
68
-------
TABLE 22. FIELD SLOPE AND SEDIMENT LOSS, 1980
Flume
2
3
4
9
16
7
17
15
11
14
Crop
Corn
Corn
Beans
Beans
Spring Wheat
Winter Wheat
Winter Wheat
Winter Wheat
Spring Wheat
Spring Wheat
Slope
(%)
2.5
3.0
3.4
4.7
1.0
1.3
1.4
1.5
3.2
3.3
Sediment Loss
(mt/ha)
15.1
5.8
5.8
19.5
.8
3.6
6.1
0.8
3.8
11.6
TABLE 23. IRRIGATION SUMMARY, FLUMES 4 AND 19, 1979
Flume
4
19
Di
1
0
version
(ha-m/ha)
.13
.47
Runoff
0.19
0.21
Percent
17
45
Runoff
field represented by flume 2. Runoff values for the fields monitored via
flumes 2 and 3 were 48% and 45% of diversions, respectively. Hence, the
difference in erosion between the fields cannot be attributed to any
obvious change in irrigation practice. Sediment losses through flume 3
increased once more in 1981, possibly because of sediment-pond filling or
failure to adequately maintain check-dam performance.
69
-------
TABLE 24. IRRIGATION SUMMARY, FLUME 17, 1980
Week of Diversion Runoff Percent Runoff
(ha-m) (ha-m)
Apr 27
May 4
11
18
25
June 1
8
15
22
29
July 6
13
20
2.03
1.99
0.55
1.17
1.18
0.99
1.38
1.38
1.08
0.89
1.38
1.38
0.69
0.51
1.00
0.35
0.70
0.78
0.49
0.56
0.54
0.44
0.34
0.47
0.53
0.34
25
50
63
60
66
50
41
39
41
39
34
39
48
A second important point to notice from Table 25 is that cropping to
wheat following beans dramatically reduced erosion (by 3- to 6-fold in most
cases). However, two of the most erosive beans fields in 1979 remained
among the most erosive wheat fields in 1980. The fields represented by
flume 4, though, showed only a 1.7-fold increase in erosion when this
cropping sequence was reversed, i.e. when wheat in 1979 was followed by
beans in 1980. On this farm unit, a gated pipe was used to irrigate the
lower portion of the field. This device allowed the irrigator to reduce
erosion somewhat by more careful regulation of stream size. In addition,
runoff from the upper part of the field passed under the gated pipe and
contributed directly to the irrigation of the lower field. The gated pipe
effectively shortened the length of run, and allowed more mid-field
sediment deposition, on this 670 meter field.
70
-------
TABLE 25. FIELD COMPARISONS, 1978-1981
1978 1979
Flume
2
3
4
9
10
11
13
14
15
16
17
18
22
Crop Sediment Crop Sediment
Loss Loss
(mt/ha) (mt/ha)
Corn
Corn 1 Corn
Wheat
__
Peas
Beets 30 Beans
Beans 14 Wheat
Corn 7 Beans
Corn 2 Beans
Wheat
Corn
Orchard
18.0
21.4
3.5
--
12.0
13.8
0.8
50.5
5.4
0.2
6.3
0.4
—
1980
1981
Crop Sediment
Loss
(mt/ha)
Corn
Corn
Beans
Beans
Carrots
Wheat
Peas
Wheat
Wheat
Wheat/Peas
Wheat
Orchard
Carrots
15.1
5.8
5.8
19.5
11.9
3.8
4.1
11.6
0.8
0.8
6.1
0.6
5.4
Crop Sediment
Loss
(mt/ha)
Corn
Corn
Wheat
Wheat
Wheat
Peas
Beans
Peas
Peas
Wheat
Wheat
Orchard
Wheat
13.6
22.3
0.81
5.1
4.5
5.8
9.86
8.2
2.56
0.50
1.63
3.03
0.63
-------
Diurnal and Seasonal Sediment Changes
Grower practices for the study area generally included 24-hour
(occasionally 12-hour) irrigation set-times, so field runoff tended to
follow a diurnal pattern. Shortly after siphon tubes were transferred
from the previous day's furrows to dry furrows, runoff-dropped sharply
(Figure 12). As the wetting front subsequently advanced down the new
furrow, transport of the more erodable soil particles occured. High initial
infiltration rates into dry soil tended to retard furrow advance rate, but
runoff increased subsequently as infiltration rate decreased and the furrow
stabilized. Visual "stabilization" was evident when all soil in and around
the furrow had become saturated and most of the more readily transportable
soil fragments had been eroded.
Runoff rate characteristically increases throughout the irrigation day
(Figure 12). This increase is generally most pronounced during the first
few hours of set time. Thereafter, the increase becomes more gradual as
infiltration slows appreciably and no longer changes rapidly with time.
Even though all siphon tubes are rapidly reset into dry furrows each
morning, the field-wide runoff rate never falls completely to zero before
discharge from newly irrigated furrows begins. Furrows generally begin to
discharge runoff within two to three hours of irrigation initiation, which
suggests a high advance rate. Actual time lag depends, of course, upon
stream size, field slope, antecedent soil moisture, soil infiltration rate,
crop rooting pattern, evapotranspiration rate, etc.
Relatively high sediment concentrations generally accompany the
initial discharge period (Figure 12). Subsequently, furrow stabilization
produces substantial decreases in runoff sediment concentration despite
steadily increasing runoff rate. This period of increasing runoff rate and
decreasing sediment concentrations is maintained until the next irrigation
cycle begins.
Because the time-dependance of runoff sediment concentration is highly
variable, it is desirable to investigate patterns of sediment loss with
time. Whenever definite patterns can be established, estimates of sediment
loss can be refined. One useful test involves multiplying discrete sediment
concentrations by their corresponding flow-rates throughout the irrigation
day. Such data from the second 24 hours of the period illustrated by
Figure 12 are shown in Figure 13. Accurate calculation of the amount of
sediment discharged from this flume over a 24-hour period would obviously
be possible from the resultant smooth curve. The limitation to calculation
of sediment losses in this manner is the impracticality of collecting
bi-hourly runoff samples throughout an irrigation season.
In addition to daily sediment-loss trends, regular seasonal trends may
also be evident. Figure 14 presents 1980 seasonal sediment losses through
flume 3, beginning when cultivation for this corn field ceased. The
irrigator for this farm unit managed his water applications with unusual
consistency, so daily hydrographs throughout the season showed remarkably
uniform variations. As a result, if sediment samples were taken at
72
-------
Runoff
o
CD
(/I
O
c
o:
11
15
23
7 Xi 15
Clock Time (hrs)
4000
- 3000 r
2000
-1000
c
O)
Figure 12. Hydrographic record of changing runoff rate (curve) and
sediment concentrations (shaded bar graph).
approximately the same time each day their change would reflect primarily
seasonal variations alone. An exponential decay curve can be fit to the 29
data points of Figure 14 with rather strong agreement. The field demon-
strated steadily declining erosivity throughout the irrigation season, once
cultivation had ceased.
Since sediment discharge rate for this field evidenced a uniform trend
with daily and seasonal time, Figures 13 and 14 could be combined into a
single, three-dimensional diagram. Figure 15 is a hypothetical illustration
showing probable sediment response for any point in time for the fields
monitored by flume 3 in 1980. Shaded regions correspond to the trends of
Figures 13 and 14. The remaining surfaces simply follow the trends given
by these two experimental curves.
Sediment Basins
Sediment basins are designed to retain sediments in irrigation runoff
from one or more fields. Sediment basins were found in this research
project to be the most successful means of controlling sediment in
irrigation return flows. By the time they were used erosion had already
occurred in the field, however. For this reason it is strongly recommended
73
-------
en
-a
a>
en
to
c
a>
-a
CD
l/l
3.0
2.0
1.0
100 300 500 700 900
Time (min)
1100
1300 1500
Figure 13. Sediment discharge hydrograph, flume 3, July 23, 1980.
that they be used in conjunction with controlled stream sizes and/or
techniques such as the cutback irrigation method.
Twelve sediment basins were constructed or redesigned in 1979 by
cost-sharing with participating growers. Five of the eight existing basins
(from 1978) were judged to be undersized. The twelve basins, ranging in
size from 5 X 18 X 1 to 12 X 49 X 1 meter, were designed by project personnel
from field data collected in 1978. In 1980, five additional basins were
constructed by growers, although some were very small, bringing the total
to twenty basins for 1980 and 1981. Figure 16 shows the location of these
basins, which collected runoff from 46% of the total area, or about 70% of
the surface-irrigated fields.
Inflow and outflow from each sediment basin were monitored daily, to
evaluate basin performance. Overall seasonal sediment basin performance is
presented in Table 26. On the average, 65% of the eroded sediment was
retained in the improved sediment basins from 1979 through 1981. By
comparison, only 14% and 28% of the sediments were retained respectively in
1977 and 1978. This demonstrates the effectiveness of sediment basins in
controlling pollutants from irrigation return flows, though it must be kept
in mind that only part of the study area was covered by adequate sediment
basins.
74
-------
en
c
i
CD
QJ
CD
fO
_c
o
O
o>
01
CO
6.0-
5.0-
4.0-
3.0 '
2.0-
1.0-
0-
O
Y = 4.13 e-°'586x
R2 = 0.90
oo
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64
Day
July
August
September
Figure 14. Seasonal trend for sediment loss, farm unit 53, fields C and D, 1980. Data points represent
sediment discharge rates between 1500 and 1600 hours each day.
-------
Q
LU
oo
DATE
Figure 15. Three-dimensional representation of seasonal and daily
sediment loss.
Efficiency of a sediment basin varies with its size and with the
quality and quantity of irrigation runoff. For a sediment particle to be
retained in the basin, particle fall velocity must combine with forward
velocity of the water to produce a direction of particle motion that will
allow the particle to reach the bottom of the basin before it reaches the
outflow end. Hence, the rate of runoff and surface area of the basin are
important factors determining basin effectiveness. Figures 17 and 18 show
seasonal operation of two sediment basins constructed on the same farm
unit. The entire farm unit was planted to corn in 1981. Basin no. 2
(Figure 17) served 16 hectares (40 ac.). It remained effective throughout
the irrigation season, with an average efficiency of 85%. The basin was
approximately half-full at the end of the 1981 season. Basin no. 3, on the
other hand, served 23 hectares (56 ac.) of more erosive field. It was
filled completely after only the third irrigation, and remained in operation
for only one month. Tables 27 and 28 present daily sampling data and
efficiencies for two basins. Basin No. 3 achieved an efficiency of 87% for
the first twenty days of irrigation (Table 28). Then, for the next two
weeks, the average efficiency dropped to only 14%. By the end of this
period the basin had filled up. This is a typical pattern for an under-
sized basin. It is interesting to note the negative efficiencies which
76
-------
Sediment Basin
Figure 16. Sediment basin locations in the study area, 1981.
-------
TABLE 26. OVERALL SEASONAL SEDIMENT BASIN RETENTION PERFORMANCE,
1977-1981
Date
1977
1978
1979
1980
1981
Number of
sediment
basins
8
8
15
20
20
Total
sediment
produced
(rot)
5594
4193
3452
3462
4257
Main-drain
sediment
discharge
(rat)
4809
3015
1287
1101
1550
Sediment
basin
retention
(mt)
785
1178
2165
2361
2707
Sediment
retention
(%)
14
28
63
68
64
were reported after filling of the basin. This represents subsequent
erosion of settled particles from the basin itself.
Sediment basins monitored in 1981 removed from 53% to 85% of the
incoming sediments, with an average of 66% removal efficiency for all
basins in the study area. Most of these sediment basins required annual
cleanout. Failure to do so resulted in increased sediment discharge at the
main drain. Where a single under-designed basin (8 X 31 X 1 m) was used to
retain sediment from several farm units of approximately 80 hectares (200 ac.)
total area, the basin had filled completely after only two irrigations,
and up to three cleanouts were required in a given season. This problem
could have been eliminated by either increasing the size of the existing
basin or by constructing more sediment basins upstream from this location.
When properly designed, sediment basins are the most effective and
most widely adaptable measure for erosion control though they fail to
prevent losses of soil from the more erosive within-field locations. The
visual effect of the basins was most impressive on growers and participants
in field tours of the study area. Hermanson and Ziari (1982) from data
collected in this study, prepared a step-by-step procedure for sediment
basin design under central Washington conditions in a publication of the
Washington State University Cooperative Extension Service.
Phosphorus Losses
Since the erosion of soil also removes associated mineral nutrients,
loss of productivity is an eventual long-term result of sediment loss.
Phosphorus, whether of native or of man-made origin, is highly immobile in
soils. This is due (among other things) to its strong affinity for the
78
-------
UD
CTl
to
o
c
o
O
o
10000
8000
6000
4000
2000
cu
o
QJ
O.
>>
O
100
80
60
ou 40
o
20
80
\
Efficiency
— — Inflow Concentration
Discharge
60
I . •"
; •/ "V-,r^^
1 A-' *"***,^
V,N
\
^/v"\.
160 180 200 220
Julian Days 1981
June July August
240
260
40
20
o
CO
Ol
C7I
IB
_C
O
September
Figure 17. Seasonal operation of sediment basin No. 2, Block 86, 1981.
-------
60000
00
o
c.
o
(0
o
c
o
o
40000-
2000C -
100
80
c
O)
o
OJ
Q.
O
c
O)
u
60
40
20
160
June
Efficiency
Inflow Concentration
Discharge
80
60
40
o
OJ
l/l
Ol
CD
o
•r-
Q
20
170 180
Julian Days 1981
July
190
200
Figure 18. Seasonal operation of sediment basin No. 3, Block 86, 1981.
-------
TABLE 27. SEDIMENT BASIN PERFORMANCE, BASIN NO. 2 (9 X 43 X 1 m), 1981
00
Date
6-11
6-12
6-15
6-17
6-18
6-19
6-20
7-2
7-3
7-6
7-7
7-8
7-9
7-10
7-11
7-13
7-15
7-16
7-17
7-18
7-20
7-21
7-22
7-23
7-24
7-25
Flow
(cfs)
.13
.01
.10
.01
.15
.08
.04
.25
.24
.19
.24
.63
.39
.69
.40
.50
.53
.62
.71
.89
.81
.65
.76
.65
.72
.73
Flow
(ha-m)
.008
.001
.006
.001
.009
.005
.002
.016
.015
.012
.015
.039
.024
.043
.025
.031
.033
.039
.044
.055
.050
.040
.047
.040
.045
.045
Sediment Cone.
(mg/D
In Out
2206
1659
10733
1605
9013
8829
2294
3510
1676
1365
1553
2400
2896
2832
2784
1196
1576
2696
2118
1309
1677
2309
2398
2832
3024
2370
533
733
440
507
345
503
253
489
467
338
305
735
519
553
568
352
418
475
565
406
413
308
384
284
344
395
Sediment in
Cum.
(mt/day) (mt)
.70
.04
2.7
.04
3.3
1.7
.22
2.2
1.0
.64
.91
3.7
2.8
4.8
2.7
2.5
2.1
4.1
3.7
2.9
3.3
3.7
4.5
4.5
5.3
4.2
.70
.82
6.22
6.22
9.5
11.2
11.4
27.8
28.8
29.4
30.4
34.1
37
42
45
50
52
56
60
66
69
73
78
83
88
93
Sediment Out Sediment Retained Retention
Cum. Eff.
(mt/day) (mt) (mt/day) (mt) (%)
.2
.02
.11
.01
.13
.10
.02
.30
.27
.16
.18
1.1
.50
.93
.56
.43
.54
.72
.98
.89
.82
.49
.72
.45
.61
.71
.2
.2
.4
.4
.5
.6
.8
1.1
1.9
2.1
2.3
3.4
3.9
4.8
5.9
6.7
7.2
7.9
8.8
10.6
11.4
11.9
12.6
13.0
13.6
15.0
.5
.02
2.6
.03
3.2
1.6
.20
1.9
.7
.5
.7
2.6
2.3
3.9
2.1
2.1
1.6
3.4
2.7
2.0
2.5
3.2
3.8
4.0
4.7
3.5
.5
.6
6
6
9
11
11
27
27
27
28
31
33
37
39
43
45
40
51
55
58
61
65
70
74
78
76
56
96
68
96
94
89
85
73
71
79
71
83
82
81
76
75
83
74
68
76
86
84
90
88
83
-------
TABLE 27 (Continued)
CO
ro
Date
7-27
7-28
7-29
7-30
7-31
8-1
8-4
8-5
8-6
8-7
8-8
8-10
8-11
8-12
8-13
8-14
8-15
8-17
8-19
8-20
9-11
9-14
9-17
Flow
(cfs)
.60
.61
.82
.94
.79
.53
.73
.79
.79
.81
.76
.73
.72
.66
.66
.76
.69
.66
.63
.74
.48
.59
.52
Flow
(ha-m)
.037
.030
.05
.06
.05
.03
.05
.05
.05
.05
.05
.05
.04
.04
.04
.05
.04
.04
.04
.05
.03
.04
.03
Sediment Cone.
(mg/1)
In Out
2179
2245
1676
1489
2189
1709
1159
1082
773
870
828
513
611
799
805
597
709
958
921
480
523
469
460
254
165
181
188
157
107
119
138
81
94
87
71
74
76
80
90
94
109
81
77
80
73
110
Sediment in
Cum.
(mt/day) (mt)
3.2
3.4
3.4
3.4
4.2
2.2
2.1
2.1
1.5
1.7
1.5
.9
1.1
1.3
1.3
1.1
1.2
1.5
1.4
.9
.6
.7
.6
101
103
106
110
114
116
122
124
126
128
129
130
131
132
134
135
136
139
140
153
155
157
158
Sediment Out Sediment Retained
Cum.
(mt/day) (mt) (mt/day) (mt)
.37
.25
.4
.4
.3
.1
.2
.3
.2
.2
.2
.1
.1
.1
.1
.2
.2
.2
.1
.1
.1
.1
.1
15.4
15.7
16.1
16.5
16.8
17.1
17.3
17.6
17.8
18.0
18.4
18.5
18.6
18.7
18.8
19.0
19.4
20.2
20.3
22.4
22.7
23.0
23.3
2.8
3.1
3.0
3.0
3.9
2.1
1.9
1.8
1.3
1.5
1.3
.8
1.0
1.2
1.2
.9
1.0
1.3
1.3
.8
.5
.6
.5
86
87
90
94
97
99
105
106
108
110
111
112
112
113
115
116
117
119
120
131
132
134
135
Retention
Eff.
(%)
89
93
89
87
93
94
91
87
90
90
89
87
88
91
90
85
87
89
91
84
85
84
76
TOTAL
158
23.3
135
85
-------
TABLE 28. SEDIMENT BASIN PERFORMANCE, BASIN NO. 3 (9 X 37 X 1 m), 1981
c»
CO
Date
6-10
6-11
6-12
6-15
6-17
6-18
6-19
6-20
6-21
7-1
7-2
7-3
7-6
7-7
7-8
7-10
7-11
7-12
7-13
7-14
Flow
(cfs)
.41
.37
.27
.37
.30
.25
.23
.38
.01
.57
.90
.74
.58
.51
.49
.55
.50
.61
.55
.55
Flow
(ha-m)
.03
.02
.02
.02
.02
.02
.02
.02
.001
.04
.06
.05
.04
.03
.03
.03
.03
.04
.03
.03
Sediment Cone.
(mg/1)
In Out
13434
37233
11727
4229
8747
1402
25557
11851
979
30719
11463
9980
5004
3580
4017
4678
3733
2262
2778
3032
1976
2367
933
394
505
158
2031
1349
364
5589
4615
1998
1878
2777
2402
3765
2489
6735
4856
5623
Sediment in
Cum.
(mt/day) (mt)
14
33.7
7.8
3.8
6.4
.9
21.3
11.0
.01
42.9
25.3
18.1
7.1
4.5
4.8
6.3
4.6
3.4
3.7
4.1
14
48
71
79
85
85
107
129
129
178
197
251
258
263
273
274
283
286
290
294
Sediment
(mt/day)
2.0
2.1
.06
.04
.04
.01
1.7
1.3
.01
7.8
10.2
3.6
2.7
3.5
3.0
5.1
3.0
10.0
6.5
7.6
Out
Cum.
(mt)
2
4
5
6
7
7
8
11
11
19
29
40
43
46
52
57
60
70
76
84
Sediment Retained
(mt/day) (mt)
12
31.6
7.2
3.4
6.0
.8
19.6
9.7
.01
35.1
15.1
14.5
4.4
1.0
1.8
1.2
1.6
-6.6
-2.8
-3.5
12
44
66
73
78
78
99
118
118
153
168
211
215
217
220
221
222
215
214
210
Retention
Eff.
(%)
85
94
90
91
94
89
89
89
63
82
63
78
62
27
39
20
33
-198
- 75
- 85
Total
294
84
209
71
-------
reactive surfaces of clays and its tendency to form some very insoluble
solid-phase compounds. The research project also examined the magnitude
of phosphorus losses from surface-irrigated fields, primarily by testing
the existence of correlations between sediment and phosphorus loss.
A statistically significant area-wide linear correlation was found
between sediment and phosphorus concentrations for each irrigation season.
Table 29 shows the results of these tests for 1979 and 1980, with phosphorus
as the dependent variable. Analysis of runoff from individual fields also
showed a significant linear dependence of phosphorus upon sediment, as
reported previously by Boyle (1980). Linear correlation coefficients were
usually greater than 0.85 for individual fields.
Table 30 shows estimated phosphorus losses from individual fields in
the study area for 1979 and 1980. Although any loss of plant nutrients is
undesirable, the magnitude of the losses was generally insignificant when
compared to the amounts of phosphorus regularly applied by growers and/or
immobilized annually in the soil by "fixation" reactions. The soil itself
represented a greater sink for the annual removal of phosphorus than did
erosion per se. Of course, eroded phosphorus can contribute to increased
phosphorus levels of area waters, including Lower Crab Creek and the
several downstream pools of the Columbia River system. It can be of
considerable environmental concern.
One of the major steps taken to reduce erosional losses from the
study area has been the construction of sediment basins. The basins are
effective so long as accumulated sediments do not prevent adequate retention
and settling times. Most basins were cleaned at least once each season
during the study period, to provide maximum retention efficiency. Many
TABLE 29. LINEAR REGRESSION PARAMETERS OF PHOSPHORUS ON AREA-WIDE SEDIMENT
LOSSES
y = a + bx 1979 1980
Intercept:
Slope:
ii ,n
a
"b"
0.26
3.63 x 10"4
0.36
4.5 x 10"4
Correlation
Coefficient: "r" 0.82 0.84
Standard Error of the
Estimate: "S . " 0.048 0.059
y *x
Number of
Observations: "n" 921 1466
84
-------
TABLE 30. PHOSPHORUS LOSSES FOR MONITORED FIELDS.
1979
Flume
2
3
4
6
7
9
10
11
13
14
15
16
17
18
19
20
21
22
Crop
Corn
Corn
Wheat
Corn
--
—
Peas
Beans
Wheat
'Beans
Beans
Wheat
Corn
Orchard
Wheat
Mint
—
—
Total
Phos.
(kg/ha)
11.0
7.1
1.0
0.9
—
--
7.3
5.5
0.7
20.2
2.2
0.4
0.2
1.7
0.9
1.3
—
—
1980
Total
Crop Phos.
(kg/ha)
Corn
Corn
Beans
--
Wheat
Beans
Carrots
Wheat
Peas
Wheat
Wheat
Beans/Wheat
Wheat
Orchard
—
—
Asparagus
Carrots
8.8
5.6
3.5
--
0.9
4.6
1.1
2.0
0.5
14.2
0.3
3.8
0.4
3.4
--
--
3.5
2.9
1981
Total
Crop Phos.
(kg/ha)
Corn
Corn
Wheat
--
--
Wheat
Wheat
Peas
Beans
Peas
Peas
Peas/Wheat
Wheat
Orchard
—
—
—
Wheat
10.0
19.0
0.6
—
—
3.6
2.4
5.1
11.3
1.5
2.0
0.3
1.2
2.6
—
—
—
0.6
filled rapidly during the early part of each season, due to heavy irrigation
of newly planted and/or recently cultivated fields, or improper design.
85
-------
Whereas sediment basins were effective in reducing transport of
sediments from fields to drainageways, transport of phosphorus was not
equally reduced. Apparently, phosphorus associated with clay-sized (<2 (jm)
particles remained suspended in runoff waters during flow through most
sediment basins. Thus, only minor regulation of phosphorus losses was
accomplished. This is supported by the observation that sediment/phosphorus
ratios measured at flume sites were generally in the range of 1500 ± 500.
Average sediment/phosphorus ratios measured at the main drain, on the other
hand, were 755 ± 385 in 1979 and 513 ± 143 in 1980 (excluding the weeks of
May 25 and June 1, 1980).
When making sediment-pond evaluations it should be remembered that,
while retention basins can be effective in reducing the amount of sediment
entering drainageways, soil has still been lost from eroded fields. Such
soil is rarely redeposited at or near its point of origin as the sediment
basin is cleaned.
INDIVIDUAL FURROWS STUDIES
Individual tests were carried out on 175 sets of furrows during the
1979-81 irrigation seasons, to evaluate the effects of stream size and
other irrigation management techniques, including cut-back irrigation, on
sediment losses. Infiltration rate, advance rate, deep percolation and
irrigation efficiencies were also calculated from these data.
Parameters which were evaluated in order to determine their effects on
sediment loss included: infiltration, run off, stream size, irrigation
scheduling, cutback irrigation, number of irrigations, crop type,
cultivation practices, and field slope.
Infiltration Rate
The capacity of a furrow to absorb water (its intake rate) is a
function of several parameters. The rate at which water infiltrates
through the wetted furrow perimeter, though defined so as to be independent
of furrow length, is dependent on stream size, slope, soil surface
conditions, furrow geometry, and time during the season. Overall trends
for each season substantiated that larger stream sizes produced somewhat
higher infiltration rates. Lower inflow rates, on the other hand, are
associated with a smaller depth of water in the furrow. The resultant
decrease in wetted perimeter produces decreased infiltration area and
thereby a reduced furrow intake rate. Infiltration rates computed from
the Kostiakov-Lewis function show that streams with higher inflow rates
generally have higher "k" values (representing the intercept of the
function i=kt , when the logarithm of the furrow intake rate is plotted
against the logarithm of total intake time). Changes in "n" values,
representing the slopes of these log-log plots, were not statistically
different for different stream sizes, however. This implies that intake
rate decreased at approximately the same rate with time, independent of
86
-------
stream size. Larger stream sizes maintained higher intake rates than did
smaller stream sizes throughout the duration of an irrigation set.
Table 31 presents typical data from four fields cropped to peas or
wheat during a one-month period in 1981, illustrating the above
relationships. Increased inflow to the head of the furrow produced a
consistent increase in "k" value, whereas "n" values showed no consistent
trend with stream size. The pattern was not completely uniform, however,
due to variations, in "n" (e.g., the sharp increase in "k" between 0.25
and 0.38 I/sec). Analysis of variance showed greater variation in slope
values than in intercept values.
Differences in field slope had less effect on infiltration rate than
did differences in stream size. With all other factors which affect water
intake being held constant, steeper slopes tended to be associated with
decreased infiltration rates.
The greatest changes in infiltration rate for a particular field
appeared to be affected by cultivation and, to a lesser extent, by the
type of crop grown. Close growing crops, such as peas and wheat, are not
cultivated during the growing season. Row crops such as corn and beans, on
the other hand, are cultivated two or more times in the early stages of each
growing season for weed control. Infiltration rates then tend to decrease
steadily following cultivation, as the furrow becomes stabilized through
removal of easily detached soil particles from the furrow surface and as a
compacted layer forms as a result of repeated soil structure breakdown
during wetting.
As an example of the decreasing infiltration rate over time for row
crops such as corn, beans, and new mint, Table 32 presents data from three
TABLE 31. INFILTRATION AT 24 HOURS AS A FUNCTION OF STREAM SIZE
No. of Inflow
Samples (I/sec)
9
7
10
7
0.19
0.25
0.38
0.50
k
2.44
2.94
5.33
6.49
n
0.22
0.21
0.16
0.21
i
(cm/hr)
0.49
0.62
1.65
1.42
Standard
Deviation
sk sn
0.007
0.007
0.028
0.026
0.120
0.066
0.072
0.112
Data from 4 fields (54A, 81A, 49A, and 49B); May 5 - June 17, 1981;
peas and wheat.
87
-------
TABLE 32. INFILTRATION AND RUNOFF FOR FRESHLY CULTIVATED VS. STABILIZED
FURROWS
Infiltration Percol.
No. of
Samples
Inflow
(I/sec)
i
(cm/hr)
Depth
(cm)
Runoff Loss
%
-------
Runoff and Sediment Loss
/
Variables which affect infiltration rate, such as recent cultivation,
crop cover and density, slope, and antecedent soil moisture, also affect
runoff and sediment loss. Runoff can be assumed to be a function of the
following independent variables:
(1) Inflow rate to the furrow
(2) Field length
(3) Infiltration constants "k" and "n"
(4) Duration of irrigation
(5) Field slope
(6) Furrow geometry
In the preceding section, it was demonstrated that the infiltration
constants "k" and "n" reflect the average depth of infiltration during an
irrigation. In a similar manner, "k" and "n" explicitly define the rate of
runoff. As "n" becomes more negative, intake rate decreases and runoff
must therefore increase. Since the volume of runoff from individual
furrows is computed from inflow-outflow data, the rate at which runoff
increases with time is a mirror image of the furrow intake rate (Figure 19).
The relationship of Figure 19 shows how prolonged irrigation leads to
decreased efficiency as runoff rate increases. As irrigation proceeds,
furrows with higher inflow rates become increasingly less efficient than
furrows with smaller inflow rates, and increased runoff volumes are
generated. Furrow intake rates also vary directly with field length.
Longer fields have a greater capacity to absorb water, based on their
increased area for infiltration. Therefore, other factors being equal,
runoff will be less from longer fields at a given stream size.
Antecedent soil moisture and cultivation also affect runoff rate.
These effects are normally reflected by changes in the infiltration
constants. Frequent irrigation of relatively moist soil results in a high
percentage of runoff. In contrast, preirrigation of drier soils for
fall- or spring-planted crops yields relatively little runoff and a cor-
respondingly higher irrigation efficiency. Table 33 demonstrates these
relationships with data obtained in 1979 from a single field of new carrots.
Preirrigation for the carrots was begun in August. Frequent irrigations
were necessary in this case to prevent moisture stress on the germinating
seed or on susceptible young seedlings. The third irrigation was charac-
terized by 76.1% runoff of the total application, compared to only 28.2%
runoff during the preplant irrigation.
Although the preplant irrigation was characterized by less runoff, a
corresponding increase in application efficiency (e ) did not materialize
(Table 33). Instead, a somewhat higher percentage of the total application
was lost as deep percolation. Since drier fields with higher intake rates
also have longer advance times, they can lose a higher percentage of the
applied water below the root zone than would be typical of wetter fields.
Sediment loss was low by comparison during the preplant irrigation, but
89
-------
o
a;
to
QJ
to
+J
c
a
s_
3
U_
0.20
0.15
0.10
0.05
O Intake rate .
Runoff rate
_L
240 480 720 960
Time (min)
1200 1440
0.20
0.15
0.10
0.05
o
c
Figure 19. Furrow intake and runoff rate relationships, late season, 1979.
Data selected from field 54A, cropped with corn.
increased during subsequent irrigations. The increased volume of runoff
from stabilized and previously moistened furrows turned out to be the
dominant influence in this case.
As indicated earlier, cultivated row crops generally experience
greater rates of sediment loss than do non-cultivated field crops under
comparable conditions. Wheat, peas, and alfalfa, in fact, are sometimes
effective in actually trapping suspended sediment from irrigation waters.
The percentage runoff from each of these crops (whether cultivated or
non-cultivated) generally increases as the season progresses. Pre-
irrigation or early-season irrigation on freshly cultivated furrows
generally produces lower runoff rates than observed for stabilized or non-
cultivated furrows. The rate of sediment discharge per unit volume of
runoff, however, tends to be highest for early-season irrigations. Graphs
on some of the following pages will illustrate sediment loss as a function
of the percentage runoff, which increases with increasing stream size.
Furrow Stream Size
Sediment loss and achievement of a reasonable irrigation efficiency
are directly dependent on furrow stream size. Too small or too large a
stream size will each produce adverse effects on irrigation performance.
Too small a stream size will apply too much water to the top of the field
and too little to the bottom. This causes deep percolation and nonuniformity
90
-------
TABLE 33. RUNOFF AND SEDIMENT LOSS RELATIONSHIPS OF SOME DRY AND PREVIOUSLY MOIST SOILS
Stage
Preplant (dry)
At planting
2 weeks later
Infiltration
Depth
(cm)
28.17
8.20
4.19
ea*
%
56.6
42.8
22.2
Percolation
Loss
%
15.2
5.1
1.7
Runoff
%
28.2
52.1
76.1
Sed. Loss
(kg/furrow)
59
120
255
Sed. Cone.
(mg/1)
6,442
7,039
10,140
Data from Field 49E, 1979, 24 hr. set at 0.38 I/sec furrow inflow rate
^Application efficiency.
-------
of application. On the other hand, too large a stream size will result in
increased runoff and serious erosion hazards. In this study, individual
furrows were irrigated throughout the growing season with stream sizes
ranging from 0.12 to 0.63 I/sec (2 to 10 gpm). Sediment discharge and rate
of runoff were measured as advancing water reached the end of each furrow.
Higher stream sizes (higher than the empirical non-erosive stream, Equation
1, page 11) generally produce excessive runoff and high sediment loss.
Sediment loss from the time runoff begins normally reaches a peak within
the first few hours (depending on field management and geometry) and then
gradually declines to a steady-state level, as shown in Figure 20. The
data for the sediment discharge hydrograph in Figure 20 were collected from
a bean field (49D) with an inflow rate of 0.38 I/sec (6 gpm). In this
example the peak was reached within the first hour of runoff, with almost
50% of the sediment being lost during this period for this recently
cultivated field. After 150 minutes relatively little sediment was
generated, although runoff continued to increase with respect to time.
Figure 21 demonstrates irrigation of a corn field after a recent
cultivation on a very erosive field of 3% slope. Irrigation with a 0.25
I/sec (4 gpm) stream size resulted in 56% runoff and 11.9 mt/ha of sediment
loss (assuming sediment is produced at the same rate from every furrow in
the field, with a 0.6 meter furrow spacing; normally irrigators use an
alternate-furrow method of irrigation for corn, however). When the stream
size was doubled to 0.5 I/sec (8 gpm), 87% runoff and 113 mt/ha of sediment
loss occurred. This represents an 850% increase in sediment loss. Hence,
sediment loss increases exponentially with furrow inflow rate. In another
case, where stream size was reduced by 30% from 0.38 I/sec (6 gpm) to 0.25
I/sec (4 gpm), sediment losses were reduced by 80%. It is clear from this
example that a continuous stream size of 0.25 I/sec or less should be set
in the furrows of this field, with higher stream sizes being totally avoided.
A word of caution is appropriate, however. These kinds of high sediment
loss-stream size relationships are applicable only to early-season
irrigations, or after a recent cultivation. Later in the season, after
several irrigations, most of the easily erodible soil particles have been
transported along the furrows. The furrows become stabilized and will no
longer produce as much sediment (e.g., Table 32).
Irrigation Scheduling
During the 1980 irrigation season, irrigation scheduling was conducted
on three fields of the study area, to evaluate the effectiveness of improved
scheduling in reducing sediment losses. Too-frequent irrigation increases
costs of labor and water, causes excessive sediment and nutrient losses,
and also reduces the efficiency of surface irrigation systems.
Three fields planted to beans were selected for these studies, with
various field slopes and lengths of run. Furrows were selected for study
from the irrigation-scheduled area, and were compared with the farmer's
(control) furrows. For each irrigation, runoff and soil loss were measured
at the end of selected furrows.
92
-------
800 i-
UD
CO
O>
tO
O
1/1
O
c
O)
Ol
600 -
400 -
200
50
O Cultivated furrow
D Stabilized furrow
100 150 200 250
Time since runoff began (min)
300
350
Figure 20. Sediment discharge hydrographs for field 49D, 1981, 0.38 I/sec inflow, cropped with beans.
-------
vo
(O
Ul
O
ID
dl
50
40
30
20
10
200
400
1000
600 800
Minutes
Figure 21. Sediment loss from farm unit 64 after cultivation, 1981.
1200
0
1400
O)
o
to
to
to
o
O)
-------
An irrigation scheduling model developed by Jensen et al. (1970) was
utilized. It was judged to be best suited to this Pacific Northwest area.
This program is generally known as the "USDA-ARS Computerized Irrigation
Scheduling Program." The model predicts crop consumptive use (ET ) by
inputting soil moisture content at field capacity, along with weather data
such as minimum and maximum temperature, wet and dry bulb temperature,
solar radiation, precipitation, pan evaporation and wind speed. Using this
scheduling model, the number of irrigations on all three fields was reduced
from eight (for the farmer's fields) to five.
Saving three irrigations is considerable in terms of water, labor, and
soil conservation (Tables 34 and 35). Soil losses from all three fields
were reduced by an average of 35%, and 37% less water was applied than via
the farmer's methods. There were no statistical differences at the 95%
level of confidence between the yields obtained by five versus eight
irrigations, with the exception of one field (46a).
Cutback Irrigation
During the 1981 irrigation season numerous furrows were tested to
evaluate the effect of cutback irrigation on sediment loss. Inflow to each
furrow was reduced by half, usually within the first hour from the start of
runoff. Cutting streams back immediately after they reach the end of the
furrow will usually cause a temporary recession of water in the furrow,
which in the case of smaller stream sizes may last for several hours.
Figure 22 shows sediment losses from cutback and non-cutback irrigation on
a bean field.
As shown in the graph, when stream sizes were cut in half from 0.25,
0.38, and 0.50 I/sec (4,6, and 8 gpm), sediment losses were reduced
respectively by 81%, 71%, and 80%. Table 36 summarizes cutback and non-
cutback effects for this field. Table 36 shows that water applications for
TABLE 34. SUMMARY OF SOIL LOSS RESULTS, SEDIMENT LOSS IS IN KG. PER
IRRIGATION PER FURROW
Field
Inflow Outflow Sediment
Rate Rate Maximum Loss
(1/s) (1/s) Efficiency (%) (ml/1)
Sediment
Loss
(kg/irrigation)
40b
46a
46a
81b
.25
.38
.38
.13
.04
.06
.22
.05
84.0
82.0
42.5
62.0
0.4
4.7
36.5
5.0
2.17
30.64
728.92
24.86
95
-------
TABLE 35. SUMMARY OF IRRIGATION SCHEDULING MODEL PERFORMANCE FOR THE
IRRIGATION SEASON (SEDIMENT LOSS AND WATER USE CALCULATIONS ARE
BASED UPON THREE AND TWO IRRIGATIONS OF CULTIVATED FURROWS FOR
THE FARMER AND MODEL MODES, RESPECTIVELY. F = FARMER TREATMENT,
M = MODEL TREATMENT)
Soil Loss Water Used Crop Yield
T-*,-.. n..**!... (raT/ha) (ha-m) bu/a
Field
40b
40b-c
46a
46a-c
81 b
81b-c
(1/s) (1/s) F M F M F
.25
.38
.38
.13
.19
.04 1.08 0.65 0.44 0.27 49.11
There was no significant sediment loss up to .38 1/s.
.06 57.27 38.32 0.44 0.28 55.89
.22
.05 50.30 33.74 0.13 0.08 29.08
.11
M
45.70
47.55
25.36
cutback furrows were on the average 46% less than those for non-cutback
furrows, while at the same time runoff was reduced by an average of 53%. A
striking feature of the cutback systems was the fact that, although stream
size (water application) was cut by a factor of 50%, total water intake was
reduced by an average of only 24% (the lowest reduction, for 0.38 I/sec
furrow intake rate, was only 8%). This demonstrates that higher
efficiencies can be achieved by applying 50% less water for most of the
irrigation set, without appreciable loss in infiltrated depth of water.
Deep percolation losses as shown in Table 36 were not excessive for any of
the stream sizes, although they were generally reduced somewhat by cutback
irrigation. Application efficiencies were improved by an average of 47%
when the cutback was used.
One of the problems of using the cutback method is determining when
and how much the stream size should be cut. Van Nieuwkoop (1979) approached
this problem mathematically, and presented the optimal time of cutback
(assuming that inflow rate should not be less than furrow intake rate at
any time during the irrigation; i.e., no recession) as follows:
1.604 F (.65 T. + T ) nT - .65 T,
T — — —
(12)
KL (n+1)
n+1
96
-------
in
in
O
c
OJ
•I—
"O
o»
00
60
50
40
30
20
10
—Non-Cutback
—Cutback
25
20
15
10
0
0 100 200 300 400 500 600 700 800 900
Minutes
QJ
O
to
c
o
to
o
CD
OJ
to
Figure 22. Sediment losses from cutback and non-cutback irrigation systems.
-------
TABLE 36. SEDIMENT LOSS AND IRRIGATION EFFICIENCY FROM CUTBACK AND NON-CUTBACK FURROWS OF FIELD
47A (BEANS), DURING THE FOURTH IRRIGATION
UD
00
Stream Size
Initial Final
(1/s) (1/s)
.25 .25
.25 .13
(Cutback)
.38 .38
.38 .19
(Cutback)
.50 .50
.50 .25
(Cutback)
Set
Time
(hr)
12
24
36
12
24
36
12
24
36
12
24
36
12
24
36
12
24
36
Volume
Appl ied
(liter)
10902
21804
32706
6231
11682
17912
16353
32706
49059
9343
17522
25699
21804
43608
65412
12401
23303
34205
Furrow
Intake
(liter)
6568
12776
18859
4255
7601
10676
4864
8240
11212
4137
7601
10959
5496
10398
15100
4467
7571
10304
Runoff
Volume
(liter)
4334
9028
13847
1976
4081
7234
11489
24465
37846
5209
9922
14740
16307
33209
50311
7934
15732
23901
Runoff Deep
Percolation
(%) (%)
40
41
42
32
34
39
70
75
77
56
57
57
75
76
77
64
67
69
5
3
2
2
1
1
3
1
1
3
1
1
2
1
1
2
1
1
Application Sediment
Eff. Loss
(%) (mt/ha)
55
56
56
66
65
60
27
24
22
42
42
42
23
23
22
34
32
30
5.1
1.0
32.0
9.4
61.1
12.3
-------
where T = cutback time after water has reached the end of the field (min)
Vrf
F = initial inflow rate
T, = total advance time
T = net application time
d
L = length of field
K, n = constants from the Kostiakov-Lewis intake function
Solving this equation using a secant method of iteration yields an
optimal cutback time. The cutback inflow rate (F ) can be found as:
FC = 0.623 KL (,65TL + Tc)n (13)
The volume of runoff (ROV) can also be expressed mathematically as:
ROV = F(TL + Tc) + (.623KL (.65TL + TC» 7.48 V^, (14)
where V. is the infiltrated volume. Volume of runoff can then be computed
for various values of T . Table 37 presents an example for calculating the
optimal cutback time, wnich produces the least runoff. In this case the
time is found to be 300 minutes. Figure 23 shows the same results graphically.
Although cutback irrigation is very effective in reducing erosion and
pollutants in irrigation return flow, continued use of smaller nonerosive
stream sizes in many cases is even more effective. This is primarily
because of the large amount of sediment loss which occurs during the first
hour of runoff, before cutback is implemented. As shown in Table 36, when
stream size was cut from 0.38 to 0.19 I/sec, 9.4 mt/ha of sediment were
produced. Only 5.1 mt/ha of sediment were produced from a non-cutback
furrow inflow rate of 0.25 I/sec, a reduction of 42%. Larger streams are
often required to advance water down the furrows, however, especially
early in the irrigation season or after a cultivation. In these cases a
combination of somewhat smaller stream size and use of the cutback method
would probably prove to be most effective.
Number of Irrigations
Sediment discharge from a given furrow decreases as the growing season
advances and as the channel in the furrow becomes stabilized. Irrigation
water advances down the furrow through the channel of least resistance,
hydraulically leveling the channel via deposition of transported soil
particles. This process requires several irrigations, however. Because of
loosened soil particles due to planting or cultivation, sediment losses are
usually highest per unit volume of runoff during pre-irrigation and/or
99
-------
TABLE 37. TOTAL RUNOFF IN RELATION TO TIME OF CUTBACK
Cutback time
(T nrin)
C*
50
100
200
300
400
500
600 .
700
800
900
1003
Cutback inflow rate
d/s)
.25
.23
.20
.18
.17
.15
.14
.14
.13
.12
.12
(gpm)
3.94
3.64
3.20
2.89
2.65
2.47
2.32
2.19
2.08
1.99
1.91
Runoff
(liter)
4368
3547
2725
2582
2831
3377
4107
4974
5950
7022
8195
(gal)
1154
937
720
682
748
892
1085
1314
1572
1855
2165
i
k
n
L
q
TL
Dn
Ta
P
W
General
nformation
= 0.1992
= -0.492
= 154 m
= 0.32 1/s
= 357 min
= 9.52 cm
= 1003 min
= 0.356
= 1.07 m
during the irrigations immediatly following planting or cultivation (Table
33). During preirrigation, furrows are normally drier and intake rates are
very high. This results in relatively low runoff. As the season progresses,
intake rates reduce and runoff increases. This is compensated by the fact
that there are less easily erodible particles available by this time.
Figure 24 shows the effect of the number of irrigations on sediment
loss from a typical carrot field. At 40% runoff, the amounts of sediment
produced were 3.2, 1.2, and 0.2 mt/ha for pre-, first, and fourth irri-
gations, respectively. This is a reduction in sediment loss of 63% and 93%
for the latter two cases. Great care must be taken during preirrigation
and the first few irrigations, because of the higher erosive power of the
stream and the presence of more readily erodible soil particles. Use of
cutback techniques during early-season irrigations is more desirable and
may even be more feasible (because of less demands on farm labor resources
before the season gets completely under way) than later in the season.
100
-------
en
O
c
3
J_
CD
tO
$-
t/o
2000 -
1500
1000
500
100 200 300 400 500
Cutback time, tr (min)
600
0
700 800
8000
6000
4000
2000
-------
o
ro
ro
1/1
in
o
QJ
c:
o
to
O
c
QJ
O)
CO
Figure 24. The effect of percent runoff and number of irrigations on sediment loss from a new
carrot field.
-------
Crop Type
The type of crop being grown in a given field can have a pronounced
effect on the amount of sediment which that field produces. Data in Tables
17 through 20 demonstrate the effect of crop type on field-wide sediment
loss. Close-grown crops such as wheat and peas, which are normally planted
in large amounts in the study area, tend to produce the lowest sediment
losses. This is because they generally produce greater densities of
vegetative cover, due to their prostrate growth or.tillering. This reduces
the flow velocity in the average furrow. Such crops extend their growth
into the furrows, serving as small check dams and causing deposition of
sediments. This can reduce sediment loss from the furrows substantially.
On the other hand, row crops with clean furrows (commonly requiring several
cultivations) tend to produce much greater amounts of sediment. In this
study it was found that rotation from wheat to beans in the same field
generally produced a 4- to 6-fold increase in sediment loss. Figure 25
shows the amount of sediment produced from four of the crops common to the
study area, as a function of runoff percentage. All fields had 3% slopes
and were at a comparable irrigation period (the fourth irrigation). Corn
and beans resulted in the highest sediment losses, while peas produced the
lowest. Corn, in this case, was being irrigated immediately after
cultivation and thus was particularly erosive. Irrigating these fields
with a stream size which produces 50% runoff would result in forty times
more sediment loss from the corn than from the pea field. It is also
interesting to note in Figure 25 that even high amounts runoff from the
field planted to peas produced only small amounts of sediment discharge.
This is typical of close-grown crops, after full cover has been reached.
Cultivation
Most row crops (i.e. beans, corn) require from one to three culti-
vations each season, depending on field conditions, weediness, and the
management ability and goals of the grower. Irrigation immediately
following cultivation, or for that matter any practice that disturbs the
furrow surface, will increase sediment discharge significantly.
After cultivation, irrigators tend to irrigate the dry soil with very
large stream sizes (usually 2 to 3 times the nonerosive stream size) in
order to increase the advance rate to the end of the field (push the water
through). They then continue to irrigate with the same stream size for the
remainder of the diurnal cycle. This can cause excessive sediment loss
because of the predominance of readily erodible soil particles at this time
of the season. Reduction in the number of cultivations or more careful
management of the irrigation water (i.e. use of a cutback system) during
this period can dramatically reduce total seasonal sediment losses. Figure
26 demonstrates the effect of cultivation on a bean field of 3% slope. The
sediment loss at 50% runoff is 48% and 76% lower for the third and fifth
irrigations, respectively, than for the first irrigation after
cultivation. Cultivation actually increases soil intake rate and thus
reduces runoff, but sediment discharge is greatest for a given volume of
runoff immediately following a cultivation (Figure 26).
103
-------
o
-P"
to
to
O
O)
E
0)
oo
60
55
50
45
40
35
30
25
20
15
10
5
0
0
Beans
Carrots
20
30
26
24
22
20
18
16
14
12
10
8
6
4
2
0
40 50 60
Runoff (%)
70
80
90 100
-------
o
en
E
i/)
i/j
O
_J
•»->
1
•o
0)
C/5
55
50
45
40
35
30
25
20
15
10
5
0
10
20
30
40 50 60
Runoff (55)
80
90 100
O)
o
(O
l/l
c
o
(/J
o
c
(U
Ol
Figure 26. Sediment loss from a field of beans as a function of runoff percentage with time
since cultivation.
-------
Infiltration rates tend to decrease steadily for several irrigations
following a cultivation, as the furrows become stabilized through removal
of easily erodible soil particles and as a compacted layer forms in each
furrow. Table 32 presented infiltration and sediment losses for recently
cultivated as well as stabilized furrows in bean and corn fields.
Reduction was 17% to 65% in infiltration, and up to 64% in sediment loss,
for the stabilized as compared to the cultivated furrows.
Field Slopes
Field slopes in the study area ranged from 1% to 5% for a majority of
the farm units. This range is typical of many of the irrigated farmlands
in Washington state. Sediment loss and the mechanisms of erosion can
change substantially for fields of differing slopes. Generally, the
steeper slopes will produce much higher sediment loss. The force of the
flowing stream generally increases with slope, as infiltration rate
decreases due to the smaller depth of flow and the reduced wetted furrow
perimeter. Figure 27 shows the effect of field slope on sediment losses
from two fields planted to beans. Both the first irrigation and the
irrigation immediately following a cultivation were done during approximately
the same period in each case. Sediment losses at 50% runoff for the field of
3% slope were 83% and 144% higher, respectively, for the first irrigation and
for an irrigation immediately following a cultivation, than for the field of
1.9% slope. (50% is a common rate of runoff for this study area.) Another
interesting point from this figure is the fact that recently cultivated
furrows on 3% and 1.9% slopes will produce, respectively, 100% and 50% more
sediment than first irrigations on the same slopes. Severity of erosion
reduces by a factor of two as slope is reduced by only 36%.
SEDIMENT LOSS MODEL FOR AN IRRIGATION FURROW
Background
Analysis of water and sediment movement in a pervious channel, usually
that of an irrigation furrow, and the associated problems of describing
this movement have been of interest to engineers for many years. The
equations governing water and sediment movement in an irrigation furrow
constitute a complex mathematical problem because of unsteady nature of the
flow, nonlinearity of the equations, and large number of variables involved
in the process.
The complexity of such a problem exists because many variables enter
into the functional relationship, and because the governing differential
equations cannot be integrated analytically except under very restrictive
and simplifying conditions, in which case the resulting solutions have
limited practical application. The solution of the complete equations of
unsteady flow problems by numerical techniques is a recent activity made
feasible by high speed computers.
106
-------
1/1
o
c:
CD
•r—
T3
-------
During the past several years, various efforts have been made to
develop computational techniques (Abbott and lonescu, 1967; Amein and
Fang, 1975; Fread, 1974; and Liggett and Cunge, 1975) applicable to natural
channel flow. Some studies (Bassett, 1972; Katbpodes and Strelkoff, 1977;
Strelkoff and Katopodes, 1977) were done to simulate flow in a border
strip. More recently, Souza (1980) and Kabir and King (1981b) reported
numerical models for flow simulation in an irrigation furrow.
In order to improve the water quality of irrigation return flow,
various investigations were conducted (Allen et al. 1977; Fitzsimmons et
al., 1977; and Robinson and Brockway, 1980). Some of these studies
emphasized on-farm structural changes while others were intended to
develop best management practices. While most of these studies were
directed towards farmwide improvements, only a few efforts have been made
to study the sediment loss from individual furrows. Kabir and King (1980)
developed an emperical technique capable of predicting the soil loss from
individual furrows for steady state conditions. More recently, Kabir and
King (1981a) reported a method of predicting soil loss from an irrigation
furrow for both steady and unsteady conditions.
A numerical model capable of predicting soil loss from irrigation
furrows has been developed. The model numerically solves the zero-inertia
equations of motion by using an implicit finite difference scheme to
compute hydraulic and sediment transport parameters in an irrigation
furrow. Details of the model developed for describing the water flow are
given in Appendix B. The way in which the model treats sediment movement
is also detailed in the following section.
Sediment Transport
Sediment movement in an irrigation furrow involves the process of
detachment and transport of the soil particles by the flow. The sediment
transport rate for a flow is determined by the available detached soil
particles and the transport capacity of the flow. Bagnold (1960) reported
that the transport capacity of a flow is proportional to the stream power
which is defined as
Ps = Y Q0 Se - (15)
in which y is the unit weight of water, Q is the outflow rate, and S is
the slope of the energy line. Due to the°high infiltration rate of tne
soil at the beginning of irrigation, the outflow rate from a furrow is low
and thus the stream power is low. But with the decline of infiltration
rate of the soil, the outflow rate increases and thus the stream power
increases with time.
Due to the transport of finer sized particles from the bed and
bank of the furrow, the availability of the sediment particles to the
stream decreases with time. This availability function is defined as
108
-------
. -I
A = A (t) (16)
in which X and 4 are constants and their values depend on initial soil
moisture and cultivation condition of the furrow. The sediment transport
rate is given by the product of stream power and availability function.
Qs = Ps A (17)
The model solves eqs. 823 and 824 (Appendix B) for all times to compute
outflow rate Q which is used to compute sediment discharge from an
irrigation furrow.
Model Verification
The procedure presented in Appendix B was programmed in FORTRAN for
the AMDAHL-470 computer. In order to verify the model performance the
results of the numerical experiments were compared with the field
experimental results. Two comparison tests are presented in this report.
The first is referred to as ARIZONA, the particulars of which are given in
Table 38 and also Figure 28, where the computed and measured advance curves
are shown. The solid curve shows the measured and the dotted curve shows
the computed advance rate of the furrow stream. The field observations for
the ARIZONA test were made at the University of Arizona, Tucson, Arizona.
Due to the unavailability of the field sediment data for the ARIZONA
test, the computed sediment discharge rate was compared with a second test,
which is labelled WASHINGTON, the particulars of which are shown in Table
38 and also in Figure 29 where the advance curves are compared. The field
observations for this test were performed by Washington State University in
the Block 86 study area. Figure 30 shows a comparison of the sediment
discharge rate for the WASHINGTON test.
Both Figures 28 and 29 show close agreement between the computed and
measured advance curves. Figure 30 shows satisfactory agreement between
the measured and computed sediment discharge rates.
The method developed can be used in designing furrow irrigation
systems to minimize soil loss from the fields by selecting proper
streamsize and length of furrow. It can also be used for on-farm water
management practices to obtain appropriate streamsize for furrow irrigated
fields. The model can also be used in designing sediment basins to
determine the sediment load delivered to the basin.
IMHOFF CONE
A method of measuring suspended sediment concentration in water is of
concern to farmers and others in Washington involved in the implememtation
109
-------
TABLE 38. PARTICULARS OF COMPARISON TESTS
TEST ARIZONA WASHINGTON
Q0, i/s
t , min
CO
S
0
n
k, cm/hra
a
L, m
Wt, min
d5Q, urn
1.33
200
0.00103
0.022
1.316
0.497
101
4
70
0.504
200
0.0329
0.02
1.36
0.7405
339
4
4.2
of plans for reduction of soil loss from irrigated farmland under section
208 of Public Law 92-500 and its amendments. Ideally, the method chosen
should be relatively inexpensive, be easily understandable by the general
public, and accurately reflect conditions in the field.
Filtration techniques for measuring sediment concentration were ruled
out because of their expense and excessive time requirements. Turbidity
was considered but investigations (King et al., 1978) showed this method
also to be unacceptable because of inherent inaccuracies. Many other
variables, which could be measured easily and accurately, were considered.
The parameter of settleable solids using an Imhoff cone was determined to
be the best compromise of the selection criteria listed above. The
decision was made realizing that settleable solids measurements would not
always represent the total suspended sediment for a particular return flow.
The Imhoff cone method meets the criteria of being inexpensive and
easily understandable. Its reliability and accuracy for irrigation
tailwater were unknown. Due to the nature of the water quality problems in
irrigated areas, and the type of program desired, it was decided that a
procedure less sophisticated than standard techniques for sediment
measurement would be adequate. Preliminary measurements indicated that a
15-minute settling time in the Imhoff cone would produce acceptable
accuracy for Washington soils.
Carter and Berg (personal communication, D. L. Carter, USDA-ARS,
Kimberly, ID, 1981) conducted studies in southern Idaho during the 1981
110
-------
30
20
Computed
Ol
10
Measured
20
40 60
Distance (m)
80
100
Figure 28. Advance curves for ARIZONA test.
-------
80
,_•
-------
8xl03
c
•r—
E
CD
01
CO
_c
o
in
O)
T3
cu
oo
6xl02
Time (min)
Figure 30. Sediment discharge rates for WASHINGTON test.
-------
irrigation season in which they used settling times in the Imhoff cone of
15, 20, 25 and 30 minutes. Their results demonstrate that all of the
settling times tried gave good estimates of sediment concentration in
runoff waters from silt loam soils. For the settling times of 15, 20, 25.
and 30 minutes they obtained r2 values of 0.86, 0.90, 0.90 and 0.92,
respectively, for linear regression of the Imhoff cone reading versus the
sediment concentration obtained by standard filtration techniques (American
Public Health Association, 1971). Results improved with the longer
settling times, although the improvement was not great. They recommend
using the 30 minute settling time, although they state that results for the
other settling times studies were acceptable.
A survey of irrigated farmland in the Yakima and Columbia Basins was
conducted using irrigation district and Soil Conservation Service office
records (Stice, 1982; King and Stice, 1982). The location of surface
irrigated units were obtained from irrigation districts. These locations
were compared with Soil Conservation Service soil maps to determine the
surface texture of each farm unit. The detailed results of this survey are
listed in Table 39. There are approximately 385.000 hectares of irrigated
farmland in the Yakima and Columbia Basins. Roughly 42 percent (163,000
hectares) of this is irrigated by either furrows or corrugations.
Approximately 93 percent of this surface irrigated farmland is comprised of
soils with a surface texture of loamy sand, sandy loam, loam, or silt loam.
The sampling criteria required that there not be any significant settling
of suspended material before the tailwater reached the sampling point. It
was essential that the stream bed have a significant and sudden drop in
elevation at the sampling point. This allowed the flow to separate from
the stream bed and insured that the bed flow was included i.n the sample and
that the sediment carried in the tailwater was well-mixed. The flumes and
pipes of the Block 86 study sites satisfied these criteria.
Samples were taken in pairs at each site. One sample of each pair was
used for determination of total suspended sediment concentration, and the
other was used to determine settleable solids concentration. The settleable
solids concentration was determined with an Imhoff cone according to the
procedure described by the American Public Health Association (1971),
section 224-F, with a settling time of 15 minutes. The total suspended
sediment concentration was determined in accordance with the procedure
described in the American Public Health Association (1971), section 224-C.
Initially this procedure was altered by using a Reeve-Angel filter disk
instead of the Whatman Disk suggested by section 224-C. Erratic results
were encountered and resulted in the use of the Whatman filter disk for
the remainder of the study. The two measurements obtained thus represented
one data pair used in developing statistical models.
Stice (1982) and King and Stice (1982) give detailed results for soils
from both the Yakima and Columbia River Basins showing the correlation
between total suspended and settleable solids in irrigation tailwater from
furrow-irrigated fields. Figure 31 gives results for the soils of the
Block 86 study sites (See also Equation 11).
114
-------
TABLE 39. DISTRIBUTION OF IRRIGATED LAND IN WASHINGTON'S MAJOR IRRIGATION DISTRICTS
Area of Surface Irrigated Land (hectare)
Irrigation
District*
QCBID
ECBID
SCBID
KRD
WIP
SVID
RID
TOTAL
Percent
of Total
Irrigated
Land
Percent of
Surface-
Irrigated
Land
Total
30,456
15,718
10,132
23,845
46,084
21,992
14,713
162,940
42.29
100
Loamy
Sand
3,323
5,635
2,121
1,559
942
149
13,729
3.56
8.43
Sandy
Loam
11,539
4,640
4,747
4,581
9,896
1,938
37,341
9.96
22.92
Loam
429
2,215
16,720
10,858
7,688
6,906
44,816
11.63
27.50
Silt
Loam
14,807
7,661
96
557
23,972
3,466
5,441
56,000
14.54
34.37
'Other
Texture
357
208
3,275
1,821
5,113
- -
280
11,054
2.87
6.78
Irrigated
Area (hectare)
94,962
58,681
79,741
24,114
60,583
38,041
29,138
385,260
*Quincy Columbia Basin Irrigation District (QCBID)
East Columbia Basin Irrigation District (ECBID)
South Columbia Basin Irrigation District (SCBID)
Sunnyside Valley Irrigation District (SVID)
Roza Irrigation District (RID)
Kittitas Reclamation District (KRD)
Wapato Irrigation Project (WIP)
-------
CTl
cn
c:
01
-a
01
GO
-o
O)
-a
c
0)
Q.
I/)
73
CO
60,000
50,000
40,000
30,000
20,000
10,000
o
sus. sed. = 200 + 1180 (set. sol.)
r2 = 0.95
10 20 . 30 40
Settleable Solids (ml/I)
50
Figure 31. Correlation of suspended sediment (mg/1) with settleable solids (ml/1) as measured with
the Imhoff cone for tailwater from furrow-irrigated fields in the study area, 1978.
-------
From the Imhoff cone studies (Stice, 1982; King and Stice, 1982) it
was concluded that the Imhoff cone cannot be expected to precisely measure
the total suspended sediment concentration in a sample of tailwater from a
furrow irrigated field. Hence its widespread use as a regulatory standard
is not recommended.
For certain soil textures, the statistical results from field samples
indicated very good correlation of the Imhoff cone reading with suspended
sediment concentration. This is especially true for sandy loam and loamy
sand. These two textures account for approximately 31 percent of the land
in Washington which is surface (furrow, corrugation) irrigated. Another 62
percent of the surface irrigated land is classified as silt loam and loam.
Results for these latter two textures were not as good, but perhaps were
still acceptable for certain uses of the Imhoff cone. Perhaps the best use
of the Imhoff cone for irrigated agriculture would be as a tool for
comparing the relative effectiveness of various practices in reducing
sediment loss from a given field. Such use should be beneficial to a
farmer, conservation district personnel, or irrigation district personnel
as an evaluation aid. Such use is consistent with implementation of a
voluntary plan for involving farmers in a program to reduce sediment
discharge from irrigated farmland.
If accurate estimates of total suspended sediment concentrations are
needed, it is recommended that the procedure described in the American
Public Health Association (1971), section 224-C be used.
It is further recommended that various settling times be used in any
further studies of this type. The work of Carter and Berg (personal
communication, D. L. Carter, USDA-ARS, Kimberly, ID, 1981) in southern
Idaho suggests that for the silt loam soils used in their studies,
correlation of Imhoff cone reading with suspended sediment concentration
increased with settling times in the range from 15 to 30 minutes.
MODELING NITROGEN LOSSES FROM FURROW-IRRIGATED SOIL
Attempts have been made to predict nitrogen movement below the root
zone based on the amount of nitrogen fertilizer added, crop uptake, and the
proportion of water passing through the root zone. Predictions made using
this approach often overestimate nitrogen leaching by 30 percent of more
(Pratt and Adriano, 1973). This simple approach fails to account for the
many possible combinations of water and fertilizer application. Sprinkler
irrigation with nitrogen dissolved in the water, for example, may result in
considerably different amounts of leaching than furrow irrigation following
banded fertilizer application.
An approach recently suggested by Batu and Gardner (1978) and Ross and
Koplik (1979) as a simple and efficient method of handling two-dimensional
solute transport problems is to approximate the two-dimensional flow field
by a group of streamtubes (Raats, 1970; Batu, 1979) and to treat solute
transport within each streamtube as one-dimensional. A major limitation of
this approach is that the concept of streamtubes loses significance for
117
-------
transient-flow situations. Water and solute transport in unsaturated soil
are not steady-state processes. To describe solute transport under these
conditions, it is necessary to solve the transient convective-dispersive .
transport equation subject to appropriate boundary conditions. In two
dimensions, this equation is:
-dJ dJ
RUPN + Q = -g-S - g-S • (18)
_o
where: p = density of water , kg m
3 -3
6 = volumetric water content , m m
C = solution concentration , kg kg
t = time , s
-3 -1
RUPN = root uptake of solute , kg m s
-3 -1
Q = other sources or sinks , kg m s
_2 -i
J = solute mass flux density , kg m s
Guymon et al. (1970) and Nalluswami et al. (1972) have solved this equation
using the finite element method, and Bresler (1975) has solved the equation
with the finite difference approach.
Although the two-dimensional convective-dispersive transport equation
has previously been solved, a physically based description of root uptake
has not been included in these solutions nor has the problem of solute
transport from a line source been addressed.
In our model, a network analysis approach is taken to describe water
transport through soil. The resulting equation is solved using an
alternating-direction implicit (ADI) method. Solute transport is simulated
by considering convection to be the dominant process and letting diffusion
be reflected by numerical dispersion. Since plant uptake plays an important
role in solute transport, a physically based p-lant uptake subroutine is
included in this analysis.
The model is applied to the problem of simulating nitrogen transport
out of a fertilizer band and through the soil for various arrangements of
band placement relative to the water level in nearby furrows. Such simula-
tions are made for surface irrigation in every furrow and in alternate
furrows.
Effects of microbial and chemical transformations of nitrogen, cation
exchange, and anion exclusion are ignored in the present analysis. Such
processes do not materially affect the path taken by a solution as it
118
-------
travels through the soil, and it is this path which is our principal
concern at present. Although these processes are not included here, the
model is structured so that they may be included at a later date with
little additional effort.
The model described in Appendix C was programmed in BASIC. It was run
on a TRS-80, model I microcomputer with 48 K memory using the Microsoft
BASIC compiler. On the average, it took 1 day to simulate a 1 week
irrigation cycle for the 12 by 13 node grid in Figure C8.
Furrow irrigation with band placement of nitrogen was simulated with
the model in order to investigate the effect of management practices on
convective nitrogen transport. Situations simulated include furrow
irrigation of a sand, silt loam and clay; furrow irrigation and varied
nitrogen placement; every-furrow versus every-other-furrow irrigation; and
alternating furrow irrigation. For these simulations irrigation was
continuous and plant uptake was not considered. The role of plant uptake
was then investigated by simulating a 30-day irrigation period for a mature
potato crop.
A homogenous, isotropic soil at an initial water content of 25 percent
was assumed irrigated for all simulations. In all cases, the furrows were
assumed to be 90 cm apart and 20 cm deep. Width of furrow bottoms and
depth of water in furrows were assumed to be 5 and 2 cm, respectively.
Nitrogen was assumed placed at a lateral distance of 40 cm from the furrow
and simulations were terminated when 50 percent of the placed nitrogen
leached below 68 cm, the assumed depth of the potato root system.
The result of simulations of every-furrow irrigation using hydraulic
properties of sand, silt loam, and clay are presented in Figure 32. For
these simulations the band of nitrogen was placed at a depth of 18 cm,
level with the height of water in the furrow. Hydraulic properties of the
soils are listed in Table 40. The results in Figure 32 show that, for a
given amount of drainage, less nitrogen is leached from a sand than from a
silt loam and less is leached from a silt loam than a clay. As soil
TABLE 40. HYDRAULIC PROPERTIES OF THE SOILS USED FOR THE SIMULATIONS
Ye Ks 6s
Type (J/kg) b (Kg s m"3) (m3 m"3)
Sand -0.66 3.4 3E-3 0.5
Silt loam -2.1 4.2 1E-3 0.5
Clay -17 6.1 1E-4 0.5
119
-------
60
silt loam
ro
o
"O
QJ
-C
O
i-
(U
NI
OJ
QJ
cn
sand
40
30
20 -
10 -
100
150 200 250
Mass of Water Drained (kg)
300
350
400
Figure 32. Nitrogen leaching as a function of drainage for continuously furrow-irrigated sand,
silt loam, and clay. Fertilizer placed at the 18 cm depth with every furrow being
irrigated.
-------
texture becomes finer, the matric component of water potential becomes more
important relative to the gravitational potential in determining the
wetting pattern from a furrow. Therefore, in fine-textured soils there is
a greater lateral component to flow from a furrow and more of the water
which infiltrates the soil moves toward the fertilizer band and may carry
nitrogen as it travels through the soil. In a sandy soil, on the other
hand, more of the water moves vertically and emerges from the root zone
without encountering nitrogen. This can also be seen in Figure 33, where
drainage water and nitrogen as a percentage of the totals are plotted
against lateral distance from the furrow. The lines designated sand, silt
loam, and clay are plots of cumulative water or nitrogen drained at the
time when 40 percent of the fertilizer nitrogen had appeared below 68 cm.
Both nitrogen and water leaching reflect a greater lateral component to
flow in the clay. Of the water drained from the clay, 15 percent carried
nitrogen, while only 11 percent of the water draining from the sand carried
nitrogen. Consequently more drainage is required to remove a given
quantity of nitrogen from a sandy soil than from a clay soil. In fact it
took twice the drainage to remove 40 percent of the applied nitrogen from
the sand as from the clay.
As a wetting front advances from a line source it becomes elongated
downwards due to the influence of gravity. Therefore, placing one
fertilizer band higher in the profile should reduce the fraction of infil-
trating water which passes through the band and more of the fertilizer
should be retained in the root zone. Results of simulations to demonstrate
this are shown in Figure 34. A silt loam soil was assumed irrigated in
every furrow for these simulations. Fertilizer band placements were (1) 20
cm deep, level with the furrow bottom, (2) 18 cm deep, level with the depth
of water in the furrow and (3) 13 cm deep. By raising the band 7 cm in
this silt loam soil, approximately one more pore volume of drainage was
required to remove 50 percent of the fertilizer nitrogen from the root
zone. From this we see that altering the placement of the fertilizer band
relative to the level of water in the furrow can be an effective management
option in retaining fertilizer against over-irrigation. Placing the band
at shallower depths should be most effective in coarse-textured soils, where
the gravitational potential is relatively more important in determining the
shape of the wetting front.
Irrigating every other furrow rather than every furrow is an
irrigation practice in parts of Washington state. Simulated alternate-
furrow irrigation was compared to conventional irrigation for a silt loam
soil with the fertilizer banded 18 cm deep. The results of these simula-
tions are given in Figure 35. The practice of irrigating every other
furrow is more effective at keeping nitrogen in the root zone than is the
practice of irrigating every furrow. In fact, nearly 20 cm more drainage
was required to move 50 percent of the applied nitrogen below the depth of
68 cm when every other furrow was irrigated as opposed to irrigating every
furrow. When irrigating every furrow, water which passes through the
fertilizer band soon reaches symmetry (no flow) line DE (Figure C5,
Appendix C) and travels downward along this line. If every other furrow
is irrigated, the IJ is the symmetry line and water which passes through
the band may continue to move laterally towards IJ. Because of the longer
121
-------
ro
IN3
0)
-t->
to
to
o
c
o
s-
o
Q.
O
(X
110
100
90
80
70
60
50
40
30
20
10
0
WATER
sand
silt
NITROGEN
sand
10
20 30 40
Lateral Distance from Furrow (cm)
50
Figure 33. Spatial distribution of water and nitrogen leaching below the 68 cm depth during
continuous every-furrow irrigation of sand, silt loam, and clay soils.
-------
ro
co
-o
O)
-C
o
res
O)
s_
CD
N
5-
O)
O
60
50
40
30
20
10
20 cm 18 9m
50
100 150 200
Mass of Water Drained (kg)
250
300
Figure 34. Nitrogen leaching as a function of drainage for continuous every-furrow irrigation
of a silt loam with various positions of fertilizer placement.
-------
-a
0)
o
o
60
50
40
30
20
10
every other furrow
every furrow
50 100 150 200 250 300 350 400
Mass of Water Drained (kg)
450
500 550
Figure 35. Nitrogen leaching as a function of drainage for continuous every-furrow and
every-other-furrow irrigation of a silt loam with the fertilizer placed 18 cm deep.
-------
trajectory of water passing through the fertilizer band in the every other
furrow case, more water crosses the bottom boundary FJ (Figure C5, Appendix
C) before nitrogen crosses FJ than if every furrow is being irrigated. The
trajectory of water passing through the band will be longer for fine-
textured soils so the model predicts that the every-other-furrow irrigation
practice will be relatively more effective in fine-textured than in coarse-
textured soils.
Since there is greater lateral movement of nitrogen if every other
furrow is irrigated, the practice of irrigating every other furrow but
alternating furrows for successive irrigations has been advocated as a
nitrogen-conserving measure in the potato-growing region of Washington
State. The rationale for this recommendation is that the practice should
produce cyclic movement of nitrogen laterally between the furrows rather
than down and out of the root zone. Nitrogen distributions for four days
of continuous every-other furrow irrigation and four days of alternating-
furrow irrigation are given in Figures 36a-d and 37a-d, respectively. A
silt loam soil with nitrogen at the 18 cm depth was assumed for the simu-
lations. Nitrogen concentrations are relative to the initial concentration
in the band. In Fig. 36a-d we see that, for the first two days of con-
tinuous every-other-furrow irrigation, nitrogen movement is primarily
lateral, away from the source, with little vertical spreading. By the end
of the third day, however, water and nitrogen were approaching the symmetry
line and nitrogen was moving appreciably downward. After four days of
continuous irrigation significant downward movement of nitrogen had
occurred, but nitrogen did not appear to move downward toward the bottom of
the root zone until the symmetry line had been reached. In Figure 37a-d,
on the other hand, we see that alternating-furrow irrigation moves the
nitrogen back and forth between furrows only to a limited extent; with the
band initially at the 18 cm depth the nitrogen was moved downward at a more
rapid rate with alternating-furrow irrigation than if the same furrow was
always irrigated (the every-other-furrow case). This downward trend is
especially noticeable by the third and fourth days. Such results suggest
that alternate-furrow irrigation is not always a nitrogen-conserving
practice relative to every-other-furrow irrigation. It is possible that,
had the fertilizer been placed higher in the profile, the alternating-furrow
technique would have been more effective in keeping nitrogen in the crop
root zone.
All simulations reported to this point have assumed continuous
irrigation without plant uptake for as long as necessary to move 50 percent
of the banded fertilizer out of the crop root zone. While continuous
irrigation is not a realistic practice, such simulations allow more rapid
evaluation of the utility of the model and of the relative merits of
various management practices. A more realistic irrigation regime,
consisting of 1 day of irrigation every 7 days, was also simulated both
with and without root uptake. The results of these simulations, for which
a silt loam soil was assumed irrigated in every furrow and nitrogen assumed
placed at the 18 cm cepth, are given in Figure 38. Root uptake was
included with the assumed root distribution of a mature potato crop (Table
Cl, Appendix C). Intermittent irrigation resulted in more nitrogen
leaching per unit drainage than continuous irrigation. This was true even
125
-------
(A) Day 1
Dav 2
ro
.3
.2
.05
.01
(C) Day 3
*furrow being
irrigated
• initial
nitrogen
placement
.1
.05
.02
.01
.05
.02
.01
.005
(D) Day 4
.02
.01
.005
.001
Figure 36. Nitrogen distribution for 4 days of every-other-furrow irrigation for a
silt loam soil. Nitrogen placed at the 18 cm depth. Lines represent
relative concentrations in the band at indicated levels from inside to
outside.
-------
r>o
(A) Day 1
.3
.2
.05
.01
(C) Day 3
(B) Day 2
*furrow being
irrigated
•initial
nitrogen
placement
.1
.05
.02
.01
.05 .
.02
.01
.005
(D) Day 4
.02
.01
.005
.001
Figure 37. Nitrogen distribution for 4 days of alternating-furrow irrigation for
a silt loam soil. Nitrogen placed at the 18 cm depth. Lines represent
relative concentrations in the band at indicated levels from inside to
outside.
-------
60
ro
CO
-o
o>
_G
o
-------
if plant uptake was included in the simulation. To move 40 percent of the
applied nitrogen below the 68 cm depth took only 70 percent as much
drainage when irrigation was intermittent. The appearance of nitrogen in
the leachate with less drainage when irrigation was intermittent resulted
from greater vertical movement of water relative to lateral movement after
irrigation was terminated during each weekly cycle. When infiltration is
stopped, the gravitational potential becomes relatively more important in
determining water flow and nitrogen moves deeper into the soil with the
redistributed water. Continuous irrigation simulations give results
similar to those expected with a streamtube approach. The significant
differences between this (continuous) simulation and those of more realistic
irrigation regimes indicate that assuming steady flow dividing the flow
field into streamtubes and treating transport as a one-dimensional process
within each streamtube is not an adequate approach for simulating solute
transport during furrow irrigation.
The results of Figure 38 make it important to demonstrate that results
from continuous irrigation comparisons would not be reversed for
intermittent irrigation settings. Toward this end, simulations were made
of weekly irrigations for a mature potato crop growing in both a sand and a
silt loam soil. Results of these simulations are shown in Figure 39. As
in Figure 32, we see that more water was required to move a given quantity
of nitrogen out of the root zone of a sand than of a silt loam whether
irrigation was intermittent or continuous. In fact, the greater relative
retention of nitrogen in the root zone of a sand became more pronounced
when irrigation simulations may be used to compare management practices on
a qualitative basis, but for quanitative comparisons it is important that
practices be simulated much as they are actually employed by the operator.
Nitrogen balances after three weeks of intermittent irrigation
including plant uptake are presented in Table 41. More nitrogen is
retained against over-irrigation in the every-furrow mode for a sand than
for a silt loam. Even more nitrogen is retained if every other furrow is
irrigated instead.
Kemper et al. (1975) suggest the need for separting the routes by
which water and nitrogen enter the plant as a method for keeping nitrogen
at positions in the profile available to the plant. Placing the fertilizer
in bands near the surface and irrigating every other furrow appear to be
management-practices which aid in keeping nitrogen in the root zone during
over-irrigation. Placing the band higher in the profile reduces the
fraction of irrigation water passing through the band, lengthens its
trajectory, and lessens nitrogen leaching. Placing the band too near the
surface, however, could result in considerable movement of nitrogen to the
surface, where few roots exist. This would make the nitrogen essentially
unavailable to plants. Irrigating every other furrow further lengthens the
trajectory of water carrying nitrogen and reduces nitrogen leaching. In
addition, it seems reasonable that the soild could then be irrigated so that
most plant water would be extracted from region ABCDEF (Figure C5, Appendix
C) while most of the nitrogen retained in region DGHIJE. Sufficient water
would have to be extracted from region DGHIJE to bring the needed nitrogen
into the plant, however.
129
-------
60 r
CO
CD
-a
01
_c
o
to
O)
O)
N
Ol
CD
o
20 h
10 U
50
150 200 250
Mass of Water Drained (kg)
300
350
Figure 39. Nitrogen leaching as a function of drainage for furrow-irrigated sand and silt loam
soils with plant uptake. Nitrogen placed at 18 cm deep with every furrow irrigated.
-------
TABLE 41. NITROGEN BALANCE AFTER THREE WEEKS OF IRRIGATING A MATURE
POTATO CROP
Every-furrow Every-furrow Every-other-furrow
Irrigation practice silt loam sand . silt loam
root zone N 11.97
plant N 39.21
drainage N 48.82
100.00
drainage water (kg) 225.0 515.8 149.3
leaching fraction 0.70 0.79 0.50
nitrogen concentration
(ppm) 62.0 6.0 0.6
ECONOMIC MODEL
Discussion in Section 5 considered all impacts of adoption decisions
to greater or lesser degrees. But incorporating each of them into a cost
effectiveness model is infeasible at this time. Practicality suggests
immediate attention be given first to existing institutions and programs
that provide some incentive to adopt pollution abatement technologies.
This analysis illustrates the importance of selected financial factors in
the adoption of erosion control practices in irrigated agriculture. Speci-
fically, four interrelated financial factors are examined: farm size,
debt/equity position, cash flow and 1980 federal income tax statutes. The
empirical analysis is presented in the context of irrigation return flow
management for erosion control in Washington's Columbia Basin.
A multiperiod linear programming model was developed to capture the
essence of farm-level decision making in the adoption of erosion control
practices. Present value of net worth was maximized over a ten year
planning horizon; horizon length was chosen to be consistent with
depreciation schedules associated with structural control practices.
Within each year, four general activity classifications comprised the
model: crop production activities, purchasing activities, separable income
tax activities and transfer activities. The crop producton activites
involved five crops (wheat, field corn, late potatoes, alfalfa and dry
beans) grown under eight irrigation systems. Furrow irrigation provided
the benchmark system to which all other systems were compared (furrow with
131
-------
sediment pond, cutback, pumpback, gated pipe, side roll, center pivot,
center pivot with corner catcher*). Purchasing activities included payment
for land, labor, irrigation system, operating loans and a fixed annual
living expense. Tax activities included 1979 federal tax rates and invest-
ment credits. Transfer activities provided for the carry-forward of
surplus capital and unused investment credit. Rotational requirements and
irrigation method were fixed for each model run. The resultant solutions
reflect intertemporal cash flow variations originating from differential
impacts of selected financial factors on erosion control practices.
Killingsworth (1980) presented a detailed discussion of model specification
and assumptions.
Both before- and after-tax income estimates were developed for two
farm sizes (65 ha and 260 ha), each under two debt/equity positions (25 and
75 percent land equity). All before- and after-tax income estimates were
developed as present values of ten year income streams. Included in these
estimates are residual values from structural practices that extend beyond
the terminal period. A fixed family living expense was also deducted from
all' income estimates.
Findings
The importance of an after-tax cash flow analysis in estimating the
cost per ton of erosion control cannot be overemphasized. The conventional
before-tax analytical framework not only systematically overestimated cost
effectiveness, but lead to an inappropriate ranking of control practices
(Tables 42 and 43). Before-tax cost of sediment control was found to be up
to five times greater than corresponding after-tax estimates. The largest
differential was found with the high equity/large farm scenario, the
smallest differential was found with the low equity/small farm scenario.
Several interrelated factors contributed to the before- and after-tax
differentials. Three such factors include the tax deductible nature of
erosion control practices, the progressive income tax structure, and
resultant cash flow changes. Erosion control practices shelter income from
taxes, thus reducing the actual after-tax cost of control. The amount of
this reduction is dependent largely upon the marginal tax bracket which the
farm faces. But the progressive nature of federal income taxes contributes
to a more subtle outcome. Cost effectiveness estimates are based upon
income comparisons to the benchmark (furrow) irrigation system. Within any
farm size and debt/equity setting, the benchmark system attains a greater
taxable income. After-tax income differentials are thereby compressed
simply because the benchmark system incurs the. greatest tax liability, it
also incurs the greatest increase in debt, which in turn reduces after-tax
net income most.
^Evidence supports a 10 percent yield grade boost for potatoes grown under
center pivots. Accordingly, the analysis was conducted with and without
this yield differential.
132
-------
TABLE 42. BEFORE- AMD AFTER-TAX COST EFFECTIVENESS OF ALTERNATIVE IRRIGATION SYSTEMS AND CROP ROTATIONS FOR 64.75 ha FARMS BY EQUITY
LEVEL, $/t
I-"
U)
CO
Before-Tax Cost
System Effectiveness
Sediment Pond $1.04
Cutback 3.64
Pumpback 2.28
Gated Pipe 4.02
Side-Roll
Furrow
Relation 2
Furrow .
Rotation 3
25% Land
Equity
Alter- Tax Cost Ranking
Effectiveness Before-Tax
$ .63
2.26
1.34
2.19
-
46.66
7.82
1
3
2
4
5
-
-
Ranking
After- 1 ax
1
4
2
3
5
7
6
Before-Tax Cost
Effectiveness
$1.03
3.58
2.23
3.95
8.74
-
-
75% Land Equity
After-Tax Cost
Effectiveness
$ .51
1.83
1.08
1.76
4.60
37.50
6.78
Ranking
Before-Tax
1
3
2
4
5
-
-
Ranking
After-Tax
1
4
2
3
5
7
6
NOTE: Cost effectiveness is the decrease in net income divided by the amount (t) of soil saved. Comparisons are made with the
benchmark furrow irrigation system and crop rotation of one-third potatoes and two-thirds wheat.
Rotation 2 is one-fifth potatoes, one-fifth beans, one-fifth corn, and two-fifths wheat.
Rotation 3 is five-sixths alfalfa and one-sixth wheat.
-------
TABLE 43. BEFORE- AND AFHR-1AX COST EFFECTIVENESS OF AITfRHAIlVE IRRIGATION SYSTEMS AND CROP ROTATIONS FOR 259 ha FARMS BY EQUITY
LEVEL. $/t
OJ
25% Land Equity
Before-Tax Cost
System Effectiveness
Sediment Pond
Gated Pipe
Cutback
Pumpback
Center-Pivot
Corner Catcher
* ior
Center-Pivot
+ un
Side-Roll
Center-Pivot with
Corner Catcher
Center-Pivot
Furrow .
Rotation 2
Furrow
Rotation 3
$1.03
2.18
3.60
2.24
4.13
5.27
8.21
8.73
9.15
-
-
After-Tax Cost
Effectiveness
$ .33
.46
1.19
.68
1.09
1.54
2.93
3.25
3.44
22.45
3.96
Ranking
Before-Tax
1
2
4
3
5
6
7
8
9
-
-
Ranking
After- Tax
1
2
5
3
4
6
7
C
a
11
10
Before-Tax Cost
Effectiveness
$1.01
2.17
3.56
2.22
4.03
5.21
7.91
8.36
8.68
-
-
75% Land Equity
After-Tax Cost Ranking
Effectiveness Before-Tax
$ .28
.35
.98
.56
.72
1.16
2.01
2.15
2.38 ~~
18.56
3.15
1
2
4
3
5
6
7
8
9
-
-
Rank i no
After-Tax
1
2
5
3
4
6
7
8
9
11
10
NOTE: Cost effectiveness is the decrease in net income divided by the amount (t) of soil saved. Comparisons are made with the
benchmark furrow irrigation system and crop rotation of one-third potatoes and two-thirds wheat.
Center-pivot system includes a 10% yield advantage for potatoes.
Rotation 2 is one-fifth potatoes, ,one-fifth beans, one-fifth corn, and two-fifths wheat.
cRotation 3 is five-sixths alfalfa and one-sixth wheat.
-------
Changes in cost effectiveness ranking were also found for both farm
sizes. Differential tax treatment between structural and nonstructural
practices was the source of these changes. Additional tax savings attending
only structural control practices moved gated pipe ahead of cutback for the
65 ha farm, while center-pivot with corner catchers (10 percent additional
potato yield) moved ahead of cutback for the 260 ha farm. Relatively close
before-tax cost effectiveness estimates appear to be necessary ingredients
for a ranking change. In no situation was debt/equity position found to
affect system ranking.
After-tax cost effectiveness estimates for the high equity position
ranged from $.46 to $4.18/t for the 65 ha farm size (Table 42). Sediment
ponds were most cost effective ($.51) followed by pumpback ($1.08), gated
pipe ($1.76), cutback ($1.83) and side roll ($4.60). These figures con-
trast with before-tax cost effectiveness estimates: sediment ponds ($1.03),
pumpback ($2.23), gated pipe ($3.95), cutback ($3.58) and side-roll ($8.74).
A modest increase in cost effectiveness estimates was found to correspond
to the low equity situation. In contrast, the high equity 260 ha farm
after-tax cost effectiveness estimates averaged 50 percent less than those
for the 65 ha farm. Sediment ponds ($.28) again were most cost effective
followed by gated pipe ($.35), pumpback ($.56), center-pivots with corner
catcher plus yield advantage ($.72), cutback ($.98), center-pivots plus
yield advantage ($1.16), side-roll ($2.01) and average yield center-pivots
with and without corner catchers ($2.15 and $2.38, respectively). Before-
tax cost effectiveness estimates ranged from $1.01 to $8.88. Similar to
the smaller farm size, low equity after-tax cost effectiveness estimates
were slightly higher than the high equity estimates. Changing crop mix
was found to be a relatively inefficient method of controlling sediment
loss.
Implications
• Voluntary adoption of pollution abatement practices requires, in
general, some form of tax or subsidy to the farmer. Cost sharing is cur-
rently the dominant incentive used to motivate the adoption of soil con-
servation practices. It is apparent from the analysis presented here that
relevant agencies involved in program implementation and cost sharing
(e.g., ASCS and state departments of ecology) should embrace existing
institutions such as federal income tax considerations in the determination
of both BMPs and "efficient" subsidy rates. When the goal is to leave the
farmer no worse off financially than before pollution abatement, the more
efficient rate of compensation per metric ton of soil loss coincides with
an after-tax estimate of cost effectiveness.
The findings of this research support a program of variable incentives
depending upon farm size and debt/equity position. No single rate of com-
pensation will be "most efficient" for all farms. Such variable rate pro-
grams undoubtedly would be politically unacceptable, though. A regional
perspective on pollution abatement appears to offer a more politically
palatable framework to capture available program efficiencies resulting
from current income tax statutes. A fixed rate subsidy high enough to
encourage all farms to adopt control practices, regardless of farm size,
135
-------
will promote oversubsidization of large farms. By basing subsidy rates on
larger farm cost effectiveness estimates, region-wide abatement goals could
be achieved at a lower social cost. Larger farms sizes may adopt erosion
control practices at a lower cost per unit of abatement. This implication
would be especially true of large farms in a high equity position. Smaller
farms would find a relatively low subsidy inadequate to motivate adoption,
while larger farms should be willing to adopt.
Underlying the above analysis is a managerial objective of net worth
maximization. Factors such as leisure, risk aversity, and asset
accumulation may weigh heavily on the adoption process. To illustrate the
potential importance of such factors, consider the impact of leisure upon
the adoption of capital intensive irrigation systems. Freedom from managing
irrigaton labor would contribute to realizing more leisure time, thereby
placing a premium on capital intensive systems. This phenomenon (combined
with others, such as less risk of improper water timing) may in fact explain
the dramatic increase in center-pivot irrigation systems evident in the
Columbia Basin, despite the fact that traditional economic models consis-
tently find center-pivots less efficient that furrow irrigation.*
Parametric analysis on the price of hired labor reveals than an
increase from the current $4.50 hourly wage rate to just under a $7.75
wage rate renders center-pivots with corner catchers and 10 percent potato
yield advantage more cost effective than even sediment ponds. Though a 70
percent increase in the price of labor, the $7.75 wage rate appears to be
well above the implicit wage rate perceived by many farmers when considering
the "nuisance" cost associated with managing irrigation labor. A small
sample of approximately 30 Columbia Basin farmers suggests the implicit
wage associated with not having to manage irrigation labor is considerably
greater, averging $10-$15 per hour. However, each of the farmers surveyed
owned center pivots, possibly biasing their response. Further research is
necessary to refine estimates of farmers' perceptions of implicit wages.
Closer examination of an increase in the price of labor to $7.75
provides additional insight into the revealed preference for center pivots.
Rill irrigation, the benchmark system, is labor intensive relative to center
pivots. Thus, the after-tax income differential betv/een rill and center
pivot irrigation narrows as the wage rate rises. After-tax cost effective-
ness of center pivots with corner catchers and 10 percent potato yield
advantage drops 62 percent, from $.72/t to $.28/t. In contrast, cost
effectiveness of sediment ponds remains essentially unchanged at $.29/t.
Both sediment ponds and furrow irrigation employ the same quantity of labor
and, thus are affected equally.
Many states have expected (albeit naively) funding levels during the
208 planning phase to somehow carry over into the implementation phase.
In light of the current political/economic mood, the precarious supply of
*Center-pivot irrigated acreage in the Columbia Basin has increased from
zero acres in 1967 to nearly a quarter of a million acres in 1980.
136
-------
previously limited funds seems to threaten potential program viability.
However, more careful modeling of the adoption process, and more creative
use of existing institutions may offer needed insight to revitalize the 208
program.
Several factors influencing adoption motivations were discussed
briefly, one of which was investigated in an empirical model—the role of
selected financial factors. Factors such as farm size and debt/equity
position were found to be particularly important in determining the actual
after-tax costs of abatement strategies. While it is clear that federal
tax statutes alone do not provide sufficient impetus to motivate adoption
of erosion control practices, failure to include such considerations leads
to serious overestimation of cost effectiveness, improper ranking of BMPs,
and ultimately faulty policy prescriptions. As other critical decision
variables are added.to adoption models, a synergistic effect is certain to
exist. Unfortunately, the seemingly infinite number of unique farm
situations might appear to leave policy oriented research in the same hope-
less predicament that has plagued farm growth research. Practicality, how-
ever, should limit the analysis to a few farm scenarios that focus on the
most important characteristics of adoption motivations.
137
-------
LITERATURE CITED
Abbott, M. B. and F. lonescu. 1967. On the Numerical Computation of
Nearly Horizontal Flows, J. Hydr. Res. 5:96-117.
Allen, R. G., J. R. Busch, G. C. Lewis and D. W. Fitzsimmons. 1977.
Management of Irrigation Stream Size for Improved Water Quality. Paper
No. 77-2580. Presented at the 1977 Winter Meeting, Amer. Soc. Ag. Engr.
Amein, M. and C. S. Fang. 1975. Implicit Numerical Modeling of Unsteady
Flows. J. Hydr. Div., Amer. Soc. Civil Engr. 101:717-731.
American Public Health Association. 1971. Standard Methods for
Examination of Water and Wastewater. Thirteenth Edition.
Ames, W. F. 1977. Numerical Methods for Partial Differential Equations.
Academic Press, New York.
Bagnold, R. A. 1960. Sediment Discharge and Stream Power. U.S.
Geological Survey Circular 421. U.S. Department of the Interior,
Washington, D. C.
Bartsch, A. F. 1972. Role of Phosphorus in Eutrophication. Environmental
Protection Agency. Ecological Research Series., EPA-R3-72-001.
Bassett, D. L.. 1972. A Mathematical Model of Water Advance in Border
Irrigation. Trans, of Amer. Soc. Ag. Engr. 15:992-995.
Batu, V. 1979. Flow Net for Unsaturation Infiltration from Strip Source.
J. Irrig. and Drain. Div., Amer. Soc. Civil Engr. 105:233-245.
Batu, V. and W. R. Gardner. 1978. Steady-State Solute Convection in Two
Dimensions with Nonuniform Infiltration. Soil Sci. Soc. Amer. Proc.
42:18-22.
Bondurant, J. A. 1971. Quality of Surface Irrigation Runoff Water.
Trans. Amer. Soc. Ag. Engr. 14:1001-1003.
Boyle, J. S. 1980. Sediment and Phosphorus Losses from an 800-Hectare
Irrigated Tract in Central Washington. M.S. Thesis. Washington State
University, Pullman, WA.
Brandt, A., E. Bresler, N. Diner, I. Ben-Asher, J. Heller and D. Goldberg.
1971. Infiltration from a Trickle Source: I. Mathematical Models.
Soil Sci. Soc. Amer. Proc. 35:675-682.
138
-------
Bresler, E. 1975. Two-Dimensional Transport of Solutes During Nonsteady
Infiltration from a Trickle Source. Soil Sci. Soc. Amer. Proc.
39:604-613.
Bresler, E. and R. J. Hanks. 1969. Numerical Method for Estimating
Simultaneous Flow of Water and Salt in Unsaturated Soils. Soil Sci.
Soc. Amer. Proc. 33:827-832.
Brown, M. J., D. L. Carter and J. A. Bondurant. 1974. Sediment in
Irrigation and Drainage Waters and Sediment Inputs and Outputs for Two
Large Tracts in Southern Idaho. J. Env. Qua!. 3:347-351.
Brutsaert, W. F. 1971. A Functional Iteration Technique for Solving
Richard's Equation Applied to Two-Dimensional Infiltration Problems.
Water Resources Res. 7:1583-1596.
Burton, J. D. 1973. Problems in the Analysis of Phosphorus Compounds.
Water Res. 7:291-307.
Campbell, G. S. 1974. A Simple Method for Determining Unsaturated
Conductivity from Moisture Retention Data. Soil Sci. 117:311-314.
Campbell, G. S. 1977. An Introduction to Environmental Biophysics.
Springer-Verlag, New York.
Campbell, M. D., G. S. Campbell, R. Kunkel and R. I. Papendick. 1976. A
Model Describing Soil-Plant-Water Relations for Potatoes. Amer.
Potato J. 53:431-441.
Canelli, E. and D. G. Mitchell. 1975. A Seimautomated Procedure for the
Determination of Phosphorus in Waste Waters and Particulates. Water
Res. 9:1093-1098.
Carter, D. L. 1976. Guidelines for Sediment Control in Irrigation Return
Flow. J. Env. Qua!. 5:119-124.
Carter, D. L. and J. A. Bondurant. 1976. Control of Sediments,
Nutrients, and Absorbed Biocides in Surface Irrigation Return Flows.
Environmental Protection Technology Series EPA-600/2-76-237. Office
of.Res. Devel. U.S. Envir. Prot. Agen. Ada, OK.
Carter, D. L., J. A. Bondurant and C. W. Robbins. 1971. Water Soluble
N03~Nitrogen, PO.-phosphorus and Total Salt Balances on a Large
Irrigated Tract. Soil Sci. Soc. Amer. Proc. 35:331-335.
Carter, D. L., J. A. Bondurant and C. W. Robbins. 1976.
Sediment-Phosphorus Relations in Surface Runoff from Irrigated Lands.
Proceedings of the Third Inter-Agency Sedimentation Conference, March
22-25, 1976, Denver, CO., p. 3-41 to 3-52.
139
-------
Carter, D. L., M. J. Brown, C. W. Robbins and J. A. Bondurant. 1974.
Phosphorus Associated with Sediments in Irrigation and Drainage Waters
for Two Large Tracts in Southern Idaho. J. Env. Qua!. 3:287-291.
Carter, R. W. and J. Davidian. 1965a.
Streams. United States Geological
Resources Inventory, Book 1, Chap.
General Procedures for Gaging
Survey Techniques of Water
9, 13 p.
Carter, R. W. and J. Davidian. 1965b. Discharge Ratings at Gaging
Stations. United States Geological Survey Techniques of Water
Resourses Inventory, Book I, Chap. 12, 35 p.
Chaudhari, N. M. 1971. An Improved Numerical Technique for Solving
Multi-Dimensional Miscible Displacement Equations. Soc. Petr. Engr.
J. 11:277-284.
Cole, C. V. and S. R. Olsen. 1959a. Phosphorus Solubility in Calcareous
Soils: I. Dicalcium Phosphate Activities in Equilibrium Solutions.
Soil Sci. Soc. Amer. Proc. 23:116-118.
Cole, C. V. and S. R. Olsen. 1959b. Phosphorus Solubility in Calcareous
Soils: II. Effects of Exchangeable Phosphorus and Soil Texture on
Phosphorus Solubility. Soil Sci. Soc. Amer. Proc. 23:119-121.
Cooley, R. L. 1970. A Finite Difference Method for Analyzing Liquid Flow
in Variablity Saturated Porous Media. US Army Corps of Engr., Hydrol.
Engr. Center, Davis, CA. Technical Paper No. 22.
Cooley, R. L. 1971. A Finite Difference Method for Unsteady Flow in
Variablity Saturated Porous Media: Application to a Single Pumping
Well. Water Resources Res. 7:1607-1625.
Cowan, I. R.
System.
1965. Transport of Water in the Soil-Plant-Atmosphere
J. Appl. Ecology 2:211-239.
Criddle, W. D., S. Davis, C. H. Pair and D. G. Shockley. 1956. Methods
for Evaluating Irrigation Systems. Agricultural Handbook No. 82,
USDA-SCS.
Crouch, S. R. and H. V. Malmstadt. 1967. A Mechanistic Investigation of
Molybdenum Blue Method for Determination of Phosphate. Analyt. Chem.
39:1084-1089.
Davidson, J. M. , D. A. Graetz, P. S. Rao and H. M. Selim. 1978.
Simulation of Nitrogen Movement, Transformation and Uptake in Plant
Root Zone EPA-600/3-78029.
De Boodt, M. and D. Gabriels. 1980. Assessment of Erosion. Wiley-
Interscience. John Wiley and Sons, Ltd. Chichester, Great Britian.
563 p.
140
-------
Douglas, J. , Jr. 1956. On the Relation Between Stability and Convergence
in the Numerical Solution of Linear Parabolic and Hyperbolic
Equations. Soc. Ind. Appl. Math. J. 4:20-37.
Douglas, J., Jr., D. W. Peaceman and H. H. Rachford, Jr. 1959. A Method
for Calculating Multi-Dimensional Immiscible Displacement. Amer.
Inst. Mining Engr. 216:197-208.
Ensminger, L. E. 1952. Loss of Phosphorus by Erosion. Soil Sci. Soc.
Amer. Proc. 17:338-342.
Ensminger, L. E. and J. T. Cope. 1947. Effect of Soil Reaction on the
Efficiency of Various Phosphates for Cotton and the Loss of Phosphate
by Erosion. Agron. J. 39:1-11. •
Environmental Protection Agency. 1973. Pollutant Discharge Elimination,
Form and Guidelines Regarding Agricultural and Silvicultural
Activities. Federal Register, Vol. 38, No. 128, Part III, July 5, 1973.
Environmental Protection Agency. 1977. National Pollutant Discharge
Elimination System, General Permit Program. Federal Register, Vol.
42, No. 24, February 4, 1977.
Environmental Protection Agency. 1979. Methods for Chemical Analysis of
Water and Wastes. Environmental Monitoring and Suppport Lab, EPA-600
4-79-020. Cincinnati, OH.
Federal Water Pollution Control Administration, National Technical Advisory
Committee. 1968. Agricultural Uses. p. 111-178. .In Water Quality
Criteria. U.S. Government Printing Office, Washington, D.C.
Fitzsimmons, D. W., C. E. Brockway, J. R. Busch, G. C. Lewis, G. M.
McMaster and C. W. Berg. 1977. On Farm Methods for Controlling
Sediment and Nutrient Losses. Proceedings of the National Conference
on Irrigation Return Flow Quality Management, Fort Collins, CO., pp.
183-191.
Fitzsimmons, D. W. , J. R. Busch, R. B. Long, K. H. Lind_ebord, G.
M. McMaster, C. E. Brockway,. L. R. Conklin, G. C. Lewis, C. W. Berg
and E. L. Michalson. 1978. Evaluation of Measures for Controlling
Sediment and Nutrient Losses From Irrigated Areas. U.S. Environmental
Protection Agency, EPA Technology Series, No. EPA-600/2-78-138.
Fread, D. L.. 1974. Numerical Properties of Implicit Four Point Finite
Difference Equations of Unsteady Flow. NOAA Technical Memorandum NWS
HYDRO-18, National Weather Service, National Oceanic and Atmospheric
Administration, US Department of Commerce, Silver Spring, Maryland.
38 p.
Freeze, R. A. 1971. Three-Dimensional Transient Saturated-Unsaturated
Flow in a Groundwater Basin. Water Resources Res. 7:347-366.
141
-------
Fruchter, J. S. , D. E. Robertson, J. C. Evans, K. B. Olsen, E. A. Lepel,
J. C. Laul, K. H. Abel, R. W. Sanders, P. 0. Jackson, N. S. Wogman, R.
W. Perkins, H. H. Van Tuyl, R. H. Beauchamp, J. W. Shade, J. L.
Daniel, R. L. Erikson, G. A. Sehmel, R. N. Lee, A. V. Robinson, 0. R.
Moss, J. K. Briant and W. C. Cannon. 1980. Mount St. Helens Ash from
the 18 May 1980 Eruption: Chemical, Physical, Mineralogical and
Biological Properties. Science 209:1116-1125.
Gales, M. E. , E. C. Julian and R. C. Kroner. 1966. Method for
Quantitative Determination of Total Phosphorus in Water. J. Amer.
Water Works Assoc. 58:1363-1368.
Gardner, W. R. 1960. Dynamic Aspects of Water Availability to Plants.
Soil Sci. 89:63-73.
Gardner, W. and C. W. Lauritzen. 1946. Erosion as a Function of the Size
of the Irrigating Stream and the Slope of the Eroding Surface. Soil
Sci. 62:233-242.
Glymph, L. M. and H. C. Storey. 1967. Sediment — Its Consequences and
Control, p. 205-220. Iji N. C. Brady (ed.), Agriculture and the
Quality of our Environment. Amer. Assoc. Adv. Sci. Symposium. Publ.
85, Washington, D.C.
Gossett, D. L. 1975. The Economics of Changing the Water Quality of
Irrigation Return Flow from Farms in Central Washington. M.S. Thesis.
Washington State University, Pullman, WA.
Grissinger, E. H. and L. L. McDowell. 1970.
Quality. Water Res. Bull. 6:7-14.
Sediment in Relation to Water
Guymon, G. L., V. H. Scott and L. R. Herrmann. 1970. A General Numerical
Solution of the Two-Dimensional Diffusion-Convection Equation by the
Finite Element Method. Water Resources Res. 6:1611-1617.
Hermanson, R. E. and F. A. Ziari. 1982. Sediment Basin Design. EB 1109,
Cooperative Extension, College of Agriculture, Washington State
University, Pullman, WA.
Israelson, 0. W., G. D. Clyde and C. W. Lauritzen. 1946. Soil Erosion in
Small Irrigation Furrows. Utah Agric. Expt. Station Bull. 320. 39 p.
Jensen, M. E., D. C. N. Robb and C. E. Franzoy. 1970. Scheduling
Irrigation Using Climate-Crop-Soil Data. J. Irrig. and Drain. Div.
Amer. Soc. Civil Engr. 96:25-28.
Kabir, J. and L. G. King. 1980. Nondimensional Statistical Model of
Furrow Irrigation Sediment Transport. Paper No. PNW 80-211, Annual
Meeting, Pacific Northwest Region, Amer. Soc. Ag. Engr., Great Falls,
MT.
142
-------
Kabir, J. and L. G. King. 1981a. Sediment Loss Prediction Model for
Furrow Irrigation. Paper No. 81-2039, Summer Meeting, Amer. Soc. Ag.
Engr., Orlando, FL.
Kabir J. and L. G. King. 1981b. A Numerical Model of Hydraulics of Furrow
Irrigation. Paper No. PNW 81-211, Annual Meeting, Pacific Northwest
Region, Amer. Soc. Ag. Engr., Edmonton, Alberta, CANADA.
Kabir J. and L. G. King. 1981c. A Numerical Model of Furrow Irrigation
Sediment Transport. Paper No. 81-2529, Winter Meeting, Amer. Soc. Ag.
Engr., Chicago, IL.
Kabir J. and L. G. King. 1981d. Simulating Sediment Transport in Furrow
Irrigation. Proc. Third International Conference On Mathematical
Modeling, Los Angeles, CA.
Katopodes, N. D. and T. Strelkoff. 1977. Hydrodynamics of Surface
Irrigation—A Complete Model. J. Irrig. and Drain. Div. , Amer. Soc.
Civil Engr. 103:309-324.
Kemper, W. D., J. Olsen and A. Hodgdon. 1975. Fertilizer or Salt Leaching
as Affected by Surface Shaping and Placement of Fertilizer and
Irrigation Water. Soil Sci. Soc. Amer. Proc. 39:115-119.
Killingsworth, D. G. 1980. The Effect of Selected Financial Factors on
the Adoption of Erosion Control Practices in Irrigated Agriculture.
M. A. Thesis. Washington State University, Pullman, WA.
King, L. G. and A. C. Janke. 1983. Effects of Farming Practices on
Sediment In Irrigation Return Flow. State of Washington, Water
Research Center, Pullman, WA.
King, L. G. and P. D. Stice. 1982. Correlation Between Total Suspended
and Settleable Solids in Irrigation Tailwater. Agricultural
Engineering Department, Washington State University, Pullman, WA.
King, L. G., D. L. Bassett and J. M. Ebeling. 1978. Significance of
Turbidity for Quality Assessment of Agricultural Runoff and Irrigation
Return Flow. Agricultural "Engineering Department, Washington State
University, Pullman, WA.
King, L. G., P. L. Wattenburger and A. C. Janke. 1977. Water Quality
Monitoring for Selected Irrigated Fields in The Sulphur Creek Drainage
Area, 1976. Agricultural Engineering Department, Washington State
University, Pullman, WA.
Kirkby, M. J. and R. P. C. Morgan, eds. 1980. Soil Erosion. John Wiley
and Sons, Ltd., Chichester, Great Britian. 312 p.
Kuentzel, L. E. 1969. Bacteria, Carbon Dioxide and Algal Blooms. J.
Water Poll. Contr. Fed. 41:1737-1747.
143
-------
Lantz, R. B. 1971. Quantitative Evaluation of Numerical Diffusion
(Truncation Error). Soc. Petr. Engr. J. 11:315-320.
Laursen, E. M. 1958. Sediment Transport Mechanics in Stable Channel
Designs. Trans. Amer. Soc. Ag. Engr. 1:754-758, 762.
Law, J. P., J. M. Davidson and L. W. Reed. 1970. Degradation of Water
Quality in Irrigation Return Flows. Okla. Agric. Expt. Sta. Bull.
B-684. 26 p.
Law, J. P. and G. V. Skogerboe. 1972. Potential for Controlling Quality
of Irrigation Return Flows. J. Env. Qua!. 1:140-145.
Lesczynski, F. B. and C. B. Tanner. 1976. Seasonal Variation of Root
Distribution of Irrigated Field-Grown Russett Burbank Potato. Amer.
Potato J. 53:69-78.
Liggett, J. A. and J. A. Cunge. 1975. Numerical Method of Solution of the
Unsteady Flow Equations, Unsteady Flow in Open Channels, Edited by K.
Mahmood and V. Yevjevich, Vol. 1, Chap. 4, Water Resources
Publications, Fort Collins, CO. pp. 89-182.
Massey, H. F. and M. L Jackson. 1952. Selective Erosion of Soil Fertility
Constitutuents. Soil Sci. Soc. Amer. Proc. 16:353-356.
McDowell, L. L and E. H. Grissinger. 1966. Pollutant Sources and Routing
in Watershed Programs. Proc. 21st Annual Meeting Soil Conserv. Soc.
Amer. 147-161.
McNeal, B. L., N. K. Whittlesey and V. F. Obersinner. 1979. Controlling
Sediment and Nutrient Lossess from Pacific Northwest Irrigated Areas.
Final Report, EPA Grant No. R-805037. Office of Research and
Development, U. S. Environ. Protect. Agency, Ada, OK.
Mech, S. J. 1959. Soil Erosion and Its Control under Furrow Irrigations
in the Arid West. Agric. Res. Ser. USDA Inf. Bull. 184. 6 p.
Mech, S. J and D. D. Smith. 1967. Water Erosion under Irrigation. Ij] R.
M. Hagen, H. R. Haise and T. W. Edminister, eds., Irrigation of
Agricultural Lands. Agronomy 11:950-963. Amer. Soc. Agron., Madison,
WI.
Meyer, L. D. and W. H. Wischmier. 1969. Mathematical Simulation of the
Process of Soil Erosion by Water. Trans. Amer. Soc. Agr. Engr.
12:754-758, 762.
Murphy, J. and J. P. Riley. 1962. A Modified Single Solution Method for
the Determination of Phosphate in Natural Waters. Anal. Chim. Acta.
27:31-36.
Murrman, R. P. and M. Peech. 1969. Effect of pH on Labile and Soluble
Phosphate in Soils. Soil Sci. Soc. Amer. Proc. 33:205-210.
144
-------
Nalluswami, M., R. A. Longenbaugh and D. K. Sunada. 1972. Finite Element
Method for the Hydrodynamic Dispersion Equation with Mixed Partial
Derivatives. Water Resources Res. 8:1247-1249.
National Soil Erosion - Soil Productivity Research Planning Committee,
Science and Education Administration -- Agricultural Research. 1981.
Soil Erosion Effects on Soil Productivity: A Research Prospective.
J. Soil Water Conserv. 36:82-90.
National Technical Advisory Committee on Water Quality Criteria. 1968.
Research Needs. U.S. Dept. Int., Fed. Water Poll. Contr. Admin, p.
69-70.
Naylor, D. V., G. C. Lewis, J. R. Busch and D. W. Fitzsimmons. 1976.
Quality of Irrigation and Drainage Waters in the Boise Valley. Univ.
of Idaho, Agric. Expt. Sta. Bull. No. 97.
Nelson, D. W. and M. J. M. Romkins. 1972. Suitability of Freezing as a
Method for the Determination of Phosphate in Natural Waters. J. Env.
Qua!. 1:323-325.
Obersinner, V. F. 1979. The Economics of Reducing Pollination from
Irrigated Return Flows in Selected Areas of the Pacific Northwest. M.
S. Thesis. Washington State University, Pullman, WA.
Parker, F. W. 1953. Soil and Fertilizer Phosphorus. Agronomy Monographs.
Vol. IV. Academic Press, New York, NY.
Peterson, D. A. 1982. Effects of Fine Volcanic Ash on Surface Irrigation
in Central Washington. M. S. Thesis. Washington State University,
Pullman, WA.
Pratt, P. F. and D. C. Adriano. 1973. Nitrate Concentrations in the
Unsaturated Zone Beneath Irrigated Fields in Southern California.
Soil Sci. Soc. Amer. Proc. 37:321-322.
President's Science Advisory Committee on Water Quality Criteria. 1965.
Restoring the Quality of Our Environment. Report of the Environmental
Pollution Panel. The White House.
Raats, P. A. C. 1970. Steady Infiltration from Line Sources and Furrows.
Soil Sci. Soc. Amer. Proc. 34:709-714.
Robbins, C. W. and D. L. Carter. 1975. Conservation of Sediment in
Irrigation Runoff. J. Soil Water Conserv. 30:134-135.
Robinson, A. R. 1971. Sediment. J. Soil Water Conserv. 26:61-62.
Robinson, C. W. and C. E. Brockway. 1980. Sediment Removal with Vegetal
Filters. Proc. Irrig. and Drain. Div. Specialty Conference, Amer.
Soc. Civil Engr., Boise, Idaho, pp. 422-429.
145
-------
Rogers, H. T. 1941. Plant Nutrient Losses from a Corn, Wheat, Clover,
Rotation on Dunmore Silt Loam. Soil Sci. Soc. Amer. Proc., 6:263-271.
Romkins, M. J. M. and D. W. Nelson. 1974. Phosphorus Relationships in
Runoff from Fertilized Soils. J. Env. Qual. 3:10-13.
Ross, B. and C. M. Koplik. 1979. A New Numerical Method for Solving the
Solute Transport Equation. Water Resources Res. 15:949-955.
Rubin, J. 1968. Theoretical Analysis of Two-Dimensional, Transient Flow of
Water in Unsaturated and Partially Saturated Soils. Soil Sci. Soc.
Amer. Proc. 32:607-615.
Sawyer, C. N. 1947. Fertilization of Lakes by Agricultural and Urban
Drainage. J. New England Water Works Assoc. 61:109-112.
Scarseth, G. D. and W. V. Chandler. 1938. Loss of Phosphate From a
Light-Textured Soil in Alabama and Its Relation to Some Aspects of
Soil Conservation. Agron. J. 30:361-374.
Schnabel, R. R. 1981. Modeling Nitrogen Leaching and Denitrification Losses
From Furrow-Irrigated Soil. Ph.D. Dissertation. Washington State
University, Pullman, WA.
Schuman, G. E., R. G. Spomer and R. F. Piest. 1973. Phosphorus Losses
From Four Agricultural Watersheds on Missouri Valley Loess. Soil Sci.
Soc. Amer. Proc. 37:424-427.
Selim, H. M. and D. Kirkham. 1973. Unsteady Two-Dimensional Flow of Water
in Unsaturated Soils above an Impervious Barrier. Soil Sci. Soc.
Amer. Proc. 37:489-495.
Skogerboe, G. V. and J. P. Law, Jr. 1971. Research Needs for Irrigation
Return Flow Quality Control. U.S. Environmental Protection Agency,
13030.
Smith, G. E. 1967. Fertilizer Nutrients as Contaminants in Water
Supplies, p. 173-186. In N. C. Brady, ed., Agriculture and the
Quality of Our Environment. Amer. Assoc. Adv. Sci. Publ. 85.
Soil Conservation Society of America. 1977. Soil Erosion: Prediction and
Control. Soil Conserv. Soc. Amer., Ankeny, Iowa. 392 p.
Souza, F. 1980. Nonlinear Hydrodynamic Model of Furrow Irrigation. Ph.D.
Dissertation, University of California, Davis, CA.
Steel, R. G. D. and J. H. Torrie. 1980. Principles and Procedures of
Statistics - A Biometrical Approach. 2nd ed. McGraw-Hill, Inc., New
York, NY.
146
-------
Stice, P. D. 1982. Correlating Settleable Solids to Total Suspended
Sediment Concentration of Irrigation Tailwater. M. S. Thesis,
Washington State University, Pullman, WA.
Strelkoff, T. and N. D. Katopodes. 1977. Border Irrigation Hydraulics
with Zero-Inertia. J. Irrig. and Drain. Div. , Amer. Soc. Civil Engr.
103:325-342.
Taylor, A. W. 1967. Phosphorus and Water Pollution. J. Soil Water
Conserv. 22:228-231.
Taylor, A. W. and H. M. Kunishi. 1971. Phosphate Equilibria on Stream
Sediment and Soil in a Watershed Draining an Agricultural Region. J.
Agric. Food Chem. 19:827-831.x
Taylor, C. A. 1940. Transportation of Soil in Irrigation Furrows. Ag.
Engr. 21:307-309.
Technicon Industrial Systems. 1973a. Orthophosphate in Water and
Seawater. Industrial Method No. 155-71W/Tentative. Tarrytown, NY.
Technicon Industrial Systems. 1973b. Orthophosphate in Water and
Wastewater. Industrial Method No. 94-70W. Tarrytown, NY.
Tongyai, M. L. C. 1976. An Evaluation of Finite Difference Numerical
Methods for Soil-Plant-Water Relations Studies. M.S. Thesis.
Washington State University, Pullman, WA.
Tongyai, M. L. C. 1977. Predicting Ponderosa Pine (Pinus Ponderosa Laws.)
Seedling Survival From Environmental and Plant Parameters. Ph.D.
Dissertation. Washington State University, Pullman, WA.
Turner, N. C. 1974. Stomatal Behavior and Water Status of Maize, Sorghum
and Tobacco Under Field Conditions: II at Low Soil Water Potential.
Plant Physio!. 53:360-365.
United States District Court for the District of Columbia. 1975a.
Memorandum, Civil Action No. 1629-73, filed March 24, 1975.
United States District Court for the District of Columbia. 1975b. Final
Judgement, Civil Action No. 1629-73, filed June 10, 1975.
USDA - Soil Conservation Service. 1980. Mount St. Helens Ash Fallout
Impact Assessment Report. Spokane, WA.
U.S. Department of the Interior. 1968. Research Needs. National
Technical Advisory Committee on Water Quality Criteria. Fed. Water
Poll. Contr. Adm. pp. 69-71.
Van Nieuwkoop, J. 1979. Optimal Design of Furrow Irrigation Systems.
M.S. Thesis. Washington State University, Pullman, WA.
147
-------
Vauclin, M., D. Khanji and G. Vachuad. 1979. Experimental and Numerical
Study of a Transient, Two-Dimensional, Unsaturated-Saturated Water
Table Problem. Water Resources Res. 15:1089-1101.
Viets, F. G. 1970. Soil Use and Water Quality -- A Look Into the Future.
J. Agric. Food Chem. 18:789-792.
Viets, F. G. 1971. Fertilizers. J. Soil Water Conserv. 26:51-53.
Wadleigh, C. H. 1968. Wastes in Relation to Agriculture and Forestry.
USDA Misc. Publ. No. 1065. Washington, D.C.
Watanabe, F. S. and S. R. 01 sen. 1965. Test of an Ascorbic Acid Method
for Determining Phosphorus in Water and NaHCO, Extracts from Soil.
Soil Sci. Soc. Amer. Proc. 29:677-678.
Weiss, C. M. 1969. Relation of Phosphate to Eutrophication. J. Amer.
Water Works Assoc. 61:387-391.
Wells, D. M. 1974. Agricultural Waste Management, by the Committee on
Agricultural Waste Management of the Environmental Engineering
Division. J. Env. Engr. Div. , Amer. Soc. Civil Engr. 100:1-6.
Zingg, A. W. 1940. Degree and Length of Land Slope as it Affects Soil
Loss in Runoff. Ag. Engr. 21: 59-64.
148
-------
APPENDIX A
INDIVIDUAL-FURROW DATA
149
-------
TABLE Al. INDIVIDUAL-FURROW DATA
(Jl
O
Farm Length Slope Dale
Unit (ra) (%)
S4A 249.3 2.24 Early season
49A 276.5 1.31 May - June
498 154.2 1.49 June
81A 79.3 0.99 June
49E 126.5 2.64 Early August
49E 304.8 2.64 August
49E 304.8 2.64 Late August
54A 249.9 2.24 Late Season
Inflow
(I/sec)
0.13
0.16
0. 19
0.25
0.38
0.19
0.25
0.38
0.25
0.38
0.09
0.13
0.16
0.19
0.25
0.38
0.50
0.19
0.25
0.38
0.50
0.13
0.19
0.25
0.38
0.50
0.13
0.19
0.25
0.38
0.
1.
1.
1.
2.
5.
6.
3.
6.
7.
2.
3.
4.
5.
8.
23.
18.
2.
4.
3.
4.
1.
1.
1.
2
2
1.
2.
2,
3.
k
975
295
798
798
393
6S4
706
109
370
239
957
978
054
029
900
4/0
288
118
648
063
161
326
509
.417
.057
.438
,036
.438
,987
,581
n Infill. Depth
(cm)
}979
-0.11
-0.11
-0.18
-0.18
-0.22
-0.31
-0.34
-0.21
-0.23
-0.26
-0.14
-0.16
-0. 14
-0.16
-0.21
-0.37
-0.32
-0.19
-0.30
-0.24
-0.28
-0.23
-0.28
-0.26
-0.29
-0.32
-0.11
-0.21
-0.24
-0.21
5.2
7.2
6.7
6.9
7.3
8.2
9.1
9.7
15.6
17.0
12.2
16.7
19.6
21.8
14.7
28.2
30.4
6.3
8.1
8.2
8.9
3.7
3.2
3.4
4.2
4.2
5.8
7.7
8.2
11.3
Runoff
(X)
18.
15.
36.
50.
65.
5.
20.
47.
19.
49.
12.
21,
27
33.
5
28
46.
18.
23.
52.
61.
34.
62.
70,
76.
82.
11
26.
42.
45.
0
2
7
6
5
0
1
|
5
0
0
3
6
7
0
2
2
.4
4
1
.5
2
5
.4
1
1
6
7
6
0
Percolation Loss
(*)
19.
12.
5.
3.
2.
43.
22.
10.
30.
7.
30.
9.
6.
4.
50.
15.
6.
37.
19.
5.
3.
10.
2.
2.
1.
1.
16.
7.
4.
3.
4
4
3
9
0
7
9
0
0
1
6
1
4
2
0
2
7
0
8
1
1
0
8
5
7
3
6
7
2
4
Cumulative Sediment Loss
(kg)
1.
0.
21.
129.
601.
0.
6.
142.
0,
130.
0.
0,
0.
19.
0.
59
158.
1,
8.
119,
313,
2.
59,
171
254.
475.
0.
1.
9.
24.
6
3
5
3
5
3
2
,5
6
2
1
5
.6
3
,5
.4
4
3
3
7
,0
4
,3
.2
,5
2
2
1
8
4
-------
TADIE Al. (Continued)
(jn
farm Length
Unit (m)
40B 134.1
46A 326.1
46B 355.7
408 134.1
61B 260.0
55A 321.0
48B 321.0
558 335.3
Slope Date Inflow
(X) (I/sec)
0.50 7-09 0.38
0.50
0.50
3.50 7-17 0.38
0.50
1.68 7-22 0.19
0.32
0.38
0.38
0.50 7-24 0.19
0. 19
4.70 7-29 0.13
0. 19
0.25
0.38
0.50
4.41 8-07 0.25
0.38
0.50
1.70 8-11 0.50
0.50
2.23 8-19 0.38
0.50
0.50
k
12.329
53.492
11.8/2
3.978
3. 185
1.448
2.454
3.155
3.780
1.844
1.463
0.564
1.006
0.914
2.210
1.448
1.478
2.484
2.911
5.105
4.298
1.920
2.987
2.819
n Infill. Depth
(en)
1980
-0.26
-0.47
-0.29
-0.30
-0.26
-0.20
-0.19
-0.21
-0.21
-0.05
-0.05
-0.08
-0. 16
-0. 11
-0. 18
-0.15
-0.09
-0.12
-0.15
-0.28
-0.24
-0. 10
-0.11
-0.15
27.9
36.9
23.7
7.6
7.7
4.9
7.8
9.5
10.8
16. 1
12.7
4.0
4.3
5.5
8.5
7.0
8.6
13.1
13.2
10.8
11.6
11.8
17.1
13.1
Runoff
<%>
24.4
25.4
54.3
53.2
63.0
32.8
27.4
31.4
17.8
19.0
35.9
40.9
57.3
60. 1
58.4
74.4
9.4
15.3
38.1
50.2
46.9
22.8
15.1
37.4
Percolation Loss Cumulative Sediment Loss
(X) (kg)
13.6
17.9
5.4
5.5
2.0
5.5
13.6
15.0
23.2
1.2
1.3
6.8
2.1
1.2
1.5
0.6
17.6
10.1
4.3
18.9
18.6
6.8
4.6
3.0
2.0
2.6
17.6
870.7
452.3
1.9
3.2
6.8
2.3
0.9
0.4
19.5
144.2
307.1
632.7
750.3
0.3
3.8
66.3
17.7
11.1
1.4
1.2
2.1
-------
TABlf Al. (Continued)
Oi
IN3
Farm
Unit
64A
408
55B
46A
54A
81A
55A
49A
498
81A
53C
Irngth Slope Date
(m) (%)
209.) 1.93 4-01
131.7 0.50 4-10
4-14
387.1 2.23 4-16
349.0 3.30 4-23
258.2 1.63 4-30
97.5 0.96 5-05
321.0 4.41 5-12
280.4 1.50 5-18
147.8 3.30 5-21
97.5 3.20 6-02
548.6 2.55 6-08
Inflow k
(I/sec)
0.50
0.38
0.38
0.38
0.50
0.41
0.41
0.63
0.35
0.38
0.38
0.50
0.38
0.50
0.63
0.19
0.25
0.38
0.50
0.19
0.25
0.38
0.50
0. 19
0.19
0.25
0.25
0.38
0.38
0.50
0.50
0.38
0.38
0.38
7.422
5. 182
87. 158
5.761
12.984
2.347
8.321
4.069
8.214
12.7/1
12.802
14.051
5.212
3.292
4.755
1.737
1.722
1.356
8.016
2.438
2.408
2.957
4.694
3.780
3.338
4.633
3.962
4.511
5.227
5.883
7.361
5.959
3.353
7.742
n Infill. Depth
(cm)
1981
-0.30
-0.08
-0.54
-0.09
-0.23
-0. 15
-0.41
-0.15
-0.36
-0.24
-0. 19
-0.23
-0.26
-0. 12
-0. 14
-0.21
-0. 17
-0. 15
-0.37
-0. 12
-0. 10
-0.12
-0.31
-0.19
-0. 19
-0.20
-0.21
-0.10
-0.31
-0.08
-0.09
-0.34
-0.25
-0.34
13.7
33.5
41.2
33.4
36.0
10.2
8.3
32.9
10.5
33.4
42.0
40.2
11.7
12.6
22.7
5.5
6.9
6.3
10.2
13.4
15.3
16.7
8.5
13.9
12.3
16.1
13.0
28.9
9.5
42.6
50.3
4.6
4.4
5.9
Runoff
(X)
58.
11.
--
10.
32.
25.
42.
26.
49.
36.
13.
44.
23.
15.
14.
40.
43.
66.
59.
23.
0
1
3
3
9
8
3
2
0
5
0
7
7
0
0
5
2
4
2
35.8
53.
82.
48.
54.
55.
64.
47.
82.
41.
31.
5
3
8
8
8
3
1
7
5
3
Percolation Loss
(X)
6.
11.
—
17.
6.
13.
8.
10.
11.
8.
20.
3.
14.
13.
8.
8.
6.
3.
2.
4.
2.
1.
0.
1.
1.
1.
1.
0.
0.
0.
0.
0
0
7
0
0
3
0
0
9
2
8
9
0
3
3
7
8
5
9
5
3
6
3
5
0
0
8
5
9
5
Cumulative Sediment Loss
(*g)
147.
5.
1.
3.
3.
2.
12.
81.
14.
47.
3.
67.
14.
1.
5.
2.
3.
51.
9.
1.
1.
27.
140.
2.
8.
6.
-
7.
6.
9.
4.
9
9
7
7
7
5
0
3
9
1
8
8
9
8
9
0
6
8
5
1
7
9
1
5
3
5
-
5
0
4
7
18.8
-
-
-
-
—
—
31.
13.
9
4
-------
TABLE Al. (Continued)
01
CO
Farm Length Slope Date
Unit (m) (%)
49A 200.4 1.50 6-12
498 150.0 2.68 6-16
81A 97.5 0.96 6-16
54A 258.2 1.63 6-17
558 347.5 2.23 6-19
490 228.6 1.88 6-22
4GA 326.1 3.30 6-23
490 228.6 1.88 6-24
47A 274.3 4.90 6-25
Inflow
(I/sec)
0. 19
0. 19
0.25
0.25
0.38
0.50
0. 19
0.19
0.25
0.38
0.50
0.38
0. 19
0.38
0.25
0.25
0.38
.0.38
0.50
0.50
0.38
0.50
0.63
0.38
0.50
0.38
0.50
0.50
0.63
0.63
0.25
0.38
0.38
0.38
0.50
0.50
0.63
k
4. 115
0.914
2. 149
2.438
1.676
2. 134
2. 149
2.027
3.24b
3.094
3.277
6.297
1.417
2.591
4. 115
3.749
2.667
5.273
5.715
4. 100
4.161
4.023
5.852
1.951
2. 195
12.314
33.269
10.058
37.048
40.447
3.978
2.682
1.859
2.941
1.768
3.216
3.246
n Inlilt
(cm
-0.53
-0. 19
-0.28
-0.28
-0.07
-0.13
-0. 17
-0.23
-0.26
-0.18
-0.25
-0.14
-0. 16
-0.11
-0.30
-0.28
-0.21
-0.35
-0.38
-0.31
-0.21
-0. 18
-0.25
-0. 11
-0.08
-0.41
-0.65
-0.29
-0.51
-0.62
-0.29
-0.20
-0.12
-0.20
-0.11
-0.24
-0.26
2.
3.
4.
5.
12.
11.
8.
5.
7.
12.
8.
31.
12.
14.
7.
7.
8.
7.
6.
7.
13.
15.
14.
11.
15.
12.
10.
20.
21.
13.
7.
9.
10.
. Depth
')
3
3
5
3
7
3
8
9
9
1
5
4
3
9
3
3
5
5
9
2
2
5
7
5
6
1
0
1
2
7
7
2
4
10.0
10.
8.
7.
5
7
9
Runoff Percolation Loss
W (%)
76.
66.
56.
59.
30.
b3.
49
55.
66.
65.
82.
42
0
0
0
0
0
.0
.5
7
.8
.8
.1
.7
39.0
24.
28
28
47
54
68
68
41
49
62
28
27.
45
67
34
43
64
37
52
46
47
59
66
75
.4
.6
.6
.9
.9
.5
.5
.5
. 1
.4
.2
3
.9
.7
.1
.2
.1
.4
.0
.0
.3
.1
.1
.6
1.
1.
1.
1.
2.
1.
3.
1.
1.
1.
0.
1.
4.
7.
18.
18.
6.
4.
5.
5.
6.
3.
5.
4.
3.
10.
5.
13.
12.
9.
11.
4.
3.
5.
3.
2.
1.
4
3
7
7
6
0
1
6
7
5
5
2
4
2
1
1
7
6
3
3
6
9
4
5
5
9
5
9
2
5
0
2
4
1
6
7
9
Cumulative Sediment Loss
(kg)
2.
5.
4.
4.
16.
5.
1.
2.
3.
4.
13.
2.
1.
1.
64.
36.
254.
570.
997.
1145.
17.
55.
373.
1.
2.
69
241.
76
158
609.
75.
549.
8
3
4
4
1
0
6
3
3
6
8
4
4
9
2
2
8
7
9
7
5
7
6
.8
4
.2
.0
.1
.0
.6
. 1
.6
398.8
345.
718.
406
1342
.7
.2
.6
.9
-------
TABLE Al. (Continued)
Farm Length Slupe Date
Unit. (m) (X)
47A/B 274.3/304.8 4.90/3.50 6-25
55B 381.0 2.23 6-30
55B 381.0 2.23 7-01
53A 419.) 2.50 7-02
490 279.2 1.88 7-07
64A 243.8 1.93 7-09
S3C 548.6 2.55 7-10
55B 387.1 2.30 7-14
47A 274.3 4.90 7-15
47A/B 274.3/335.3 4.90/3.50 7-16
Inflow
(I/sec)
0.50
0.50
0.63
0.63
0.25
0.38
0.38
0.50
0.63
0.63
0.25
0.25
0.38
0.50
0.25
0.25
0.38
0.38
0.25
0.38
0.50
0.25
0.38
O.bf!
0.25
0.25
0.38
0.38
0.50
0.50
0.25
0.38
0.38
0.50
0.50
3.
8
2.
3.
11.
12.
7
3S.
14.
3.
2.
A.
3.
5.
1.
1.
0.
1.
0.
0.
1.
2
1.
3.
0.
0.
1.
k
018
733
804
155
582
146
315
022
1P8
642
SCO
008
322
182
109
204
991
021
853
991
966
286
692
246
823
442
783
0.930
0.
0.
0.
0.
0.
6.
4.
853
503
.869
396
975
721
862
n Infill. Depth
(cm)
-0.21
-0.42
-0.23
-0.25
-0.49
-0.44
-0.36
-0.64
-O.DO
-0. 18
-0. 19
-0.24
-0. 19
-0.27
-0.21
-0. 13
-0. 10
-0.21
-0. 11
-0. 10
-0. 16
-0.21
-0. 11
-0.26
-0.04
-0.03
-0.24
-0.21
-0.08
-0.08
-0. 12
-0.02
-0. 17
-0.39
-0.37
8.8
8.0
7.9
7.9
7.0
9.8
9.2
10.7
8.7
13.8
8.7
9.5
11.3
H.5
3.9
6.3
6.3
3.4
5.0
6. 1
8.3
7.5
9.7
7.8
7.6
4.3
4.9
3.0
6.2
3.7
4.5
3.7
4.0
7.7
6.2
Runoff Percolation Loss
(%) (%)
19.
31.
46.
47.
17.
24.
19.
40.
61.
32.
25.
13.
35.
53.
73.
55.
71.
84.
21.
34.
32.
18.
26.
57.
41.
66.
74.
84.
76.
85.
16.
52.
53.
34.
46.
4
1
7
0
7
7
0
8
8
4
7
6
8
3
0
8
0
8
2
9
5
4
0
0
4
2
8
3
2
9
0
1
3
0
5
20.
14.
7.
6.
15.
16.
15.
13.
6.
7.
14.
26.
14.
6.
2.
2.
1.
'•
6.
5.
9.
0.
9.
3.
2.
1.
1.
1.
1.
0.
14.
16.
5.
0.
2.
3
9
0
8
9
1
9
5
6
1
4
2
9
7
3
8
7
3
6
4
3
7
4
6
6
7
3
0
1
8
2
0
6
6
8
Cumulative Sediment Loss
(kg)
43.
254.
797.
1195.
10.
85.
390.
334.
18/8.
22.
131.
18.
102.
186.
302.
175.
1187.
889.
19.
174.
186.
0.
51.
711.
90.
171.
534.
440.
1021.
388.
83.
78.
126.
723.
675.
4
0
5
2
7
9
0
0
8
8
4
9
7
2
4
9
8
9
8
6
2
9
0
2
9
6
8
8
5
1
6
6
7
2
4
-------
TABLE Al. (Continued)
cn
en
Farm
Unit
490
53A
64A
4SC
53C
49D
49D
47A
55A
64A
S3A
length Slope Dale Inflow
(m) (%) (I/sec)
298.7 1.88 7-17 0.25
0.25
0.38
0.50
518.2 2.50 7-20 0.25
0.25
0.38
0.38
0.50
243.8 1.37 7-21 0.38
0.38
0.50
0.50
118.9 7.00 7-22 0.25
0.38
548.6 2.55 7-23 0.25
0.25
0.25
0.50
243.8 1.88 7-24 0.25
0.25
279.2 1.88 7-28 0.25
0.38
3!0.9 4.90 7-30 0.25
0.25
297.2 4.41 7-31 0.50
0.63
243.9 1.93 8-03 0.25
0.38
548.6 2.55 8-04 0.38
0.38
0.50
0.50
k
3.063
8.976
8.077
0.960
1.250
0.884
1.63.1
0.960
1.600
19.660
24.384
10.058
9.296
3.810
9.296
1.539
9.342
9.342
2.134
0.625
0.701
1.676
1 . 966
0.823
0.716
2.469
3.231
1.478
1.996
1.387
1.387
1.295
2.286
n
-0.44
-0.52
-0.03
-0.08
-0. 17
-0.16
-0.18
-0, 11
-0.16
-0.47
-0.49
-0.29
-0.37
-0. 11
-0.37
-0.25
-0.69
-0.24
-O.Z5
, -0.10
-0.17
-0.23
-0.28
-0.14
-0. 10
-0.06
-0.13
-0.21
-0. 18
-0.19
-0.18
-0.20
-0.22
Infill. Depth
(cm)
2.6
5. 1
7.9
6.9
5.0
3.8
6.2
5.7
7.0
14.0
15.1
18.5
11.6
21.8
11.9
3.9
2.4
5.5
5.4
4.0
2.9
4.8
4.2
4. 1
4.5
19.4
16.9
4.8
7.8
5.0
5.1
4.5
7.0
Runoff Percolalion loss
(X)
77.7
57.3
55.1
70.6
23.5
42.5
37.7
44.0
48.4
33.7
--
29.4
59.3
22.6
73.3
38.5
63.3
40.8
57.6
82.8
80.1
61.6
77.7
63.9
SS.7
15.2
42.4
66.9
64.0
47.1
45.5
65.3
45.6
(%)
4.2
6.8
2.3
1.9
8.2
5.7
5.1
3.5
2.5
11.0
--
19.7
7.1
8.2
1.3
6.2
6.2
8.7
3.2
1.5
1. 1
2.7
1.3
1.8
2.3
7.0
3.3
2.6
2.5
4.6
5.2
2.4
3.9
Cumulative Sediment loss
(kg)
953.5
408.9
919.4
963.9
9.0
28.1
53.9
54.8
151.6
265.7
10.4
88.6
976.2
6.1
58.7
75.3
65.4
SO. 1
380.3
504 . 1
193.2
111.8
753.4
17.8
18.7
1.2
113.5
139.0
550.1
4.5
5.7
79.0
16.1
-------
1ABIE Al. (Continued)
Ul
CTi
Farm
Unit
55B
490
Length Slopfl Oat.e Inflow
(m) (X) (I/sec)
387.1 2.23 8-05 0.
0.
0.
279.2 2.10 8-06 0.
0.
25
25
38
25
38
0.38
81A
49A
47A
106.7 3.20 8-07 0.
0.
259.1 2.20 8-10 0.
38
50
63
310.9 4.90 8-11 0.19
0.
0.
0.
0.
0.
19
.25
38
.38
.50
k
1.722
2.347
1.600
1.478
0.457
1.372
11.537
25. 146
10.394
0.945
0.853
0.808
1.494
1.920
2.484
n Infill. Depth
(cm)
-0.28
-0.31
-0.20
-0.22
-0.05
-0. 12
-0. 19
-0.32
-0.29
-0.11
-0.06
-0.06
-0. 12
-0.22
-0.29
3.6
4.2
5.5
4.5
4.0
7.7
36.5
40.4
20.3
S.6
6.7
6.5
8.4
5.9
5.1
Runoff
(%)
59.
53.
59.
64.
79.
59.
14.
35.
38.
33.
18.
42.
50.
65.
77.
6
7 .
5
0
1
4
7
4
2
5
6
6
8
4
9
Percolation Loss
(%)
6.
5.
3.
3.
1.
2.
26.
13.
8.
3.
6.
2.
2.
2.
1.
1
4
5
0
7
7
5
5
5
8
7
7
1
1
6
Cumulative Sediment Loss
(kg)
41.
60.
246.
144.
288.
188.
7.
159.
64.
1.
1.
2
26.
35.
174
5
6
3
4
5
.9
4
7
.4
,0
,9
.9
. 1
.5
.1
-------
APPENDIX B
157
-------
INDIVIDUAL-FURROW SEDIMENT LOSS MODEL
GOVERNING EQUATIONS
The governing equations of unsteady open channel flow in an irrigation
furrow are the so called Saint-Venant equations, which consist of the .
equation of continuity
8-f + r? + rr =°
[Bl]
and the equation of motion
gat
a x
Ag.
= ACS.. - sj + —
[B2]
in which Q = flow rate; y = depth of flow; x = distance measured along the
channel; t = time; A = cross sectional area of the flow; A = infiltrated
volume per unit length; S = bottom slope; S = slope of energy line; g~ =
gravitational acceleration; and P = pressure force on the cross section
per unit weight of water. The overbars denote the dimensional quantities.
Using Manning's equation, S can be defined as
S =
[B3]
C2 A2 R4/3
U 2
in which C = 1.0 m 1/3/s in SI units and 1.49 ft 1/3/s in English units;
"n = Manning's roughness coefficient; and R = hydraulic radi-us.
The infiltrated volume per unit length A is defined by
"A = P z
z w
[B4]
in which P = wetted perimeter; and z = infiltrated depth given by
2=
[B5]
158
-------
where k and a are the Kostiakov functions; and T = infiltration opportunity
time.
For the cross sectional area of the flowing stream, the following
relationship is assumed
8=0^ [B6]
in which C and M are constants and B = top width of the stream. The value
of M defines the shape of the furrow. For M = 0 the furrow is rectangular;
for 0 < M < 1 the furrow is parabolic, and for M = 1 the furrow is
triangular.
The pressure force on the cross section per unit weight of water is
defined as
? = / A dy [B7]
DIMENSIONLESS FORM OF GOVERNING EQUATIONS
__ The following dimensionless variables were defined using ^ , T, X, Y ,
Z as the characteristic variables:
_
Yn
z = t = ; t = ; Az = ; v = S = [B8]
Z T T Zy-n V Sc
in which Q = inflow rate; X = y /S ; T = Xy /Q , y = normal depth of
flow using Q as the flow rate; Z = infiltrated volume per unit area in
time T. S is the characteristic slope defined as
_ = q^ B» [89]
C _2 -2 _ 4/3
CAR
u n n
in which A and R = cross sectional area and hydraulic radius, respectively,
with Q as normal flow. The symbols without the overbar denote the
dimensionless variables.
159
-------
Introducing the nondimensional variables of eq. B8 in eqs. Bl and B2,
the following nondimensional form of governing equations results
[810]
3A
if 5 * - *(!-", Sf) = 0 [Bll]
in which, F is the nondimensional Froude number defined as
F = ^o CB12]
n
K and K are defined as
[B13]
yn
K = A2 R 4/3 [B14]
s n n L J
Souza (1980) reasoned that the acceleration terms in the governing
equations have a negligible effect on the flow of the furrow stream and
thus could be neglected. For such a condition F = 0 and hence the
nondimensional momentum equation (Eq. Bll) reduces to
I* = A (1 - Ks Sf ) [B15]
INITIAL AND BOUNDARY CONDITIONS
The governing equations, eqs. BIO and B15, for the zero inertia model
are subject to certain initial and boundary conditions. The initial up-
stream bpundary conditions are
Q = 0 and y = 0 at t < 0;
Q.n = 1 at 0 < t < tco;
Qin = 0 at t > tCQ; [B16]
Where-t = time of cutoff; and Q. = inflow rate.
co nn
160
-------
During the advance phase of the irrigation, the downstream boundary
condition is
Q=Oandy=OatX=X
-------
OJ
14
12
10
4
2
Rectangular
Cell
Triangular
Cell
I— Ax —I
Oblique
Cell
4
Distance, x
Figure Bl. Computational grid system.
J I
7 8
x=L
162
-------
- xi
2At
zi
J J
-xi
2At
-5 + A"? .. + K(A J + A J )
Zi
[B19]
in which AC is the residual continuity and 6 is a weighting factor. For
0.5 < 0 < 1.0 the difference equation, eq. B19, is unconditionally stable,
and for 6 < 0.5 the scheme is conditionally stable as reported by Fread
(1974). For this model a value of 6 = 0.6 was used.
In Eq. B19, AC is a function of four unknowns, such as y. , Q. ,
y.+,, and Q-+1- So using chain rule one can write
d (AC) = *f- dy.
3(AC)
9Q
3(AC)
3y
ay
j+l 9 (AC)
i+l
dQ
[820]
For the Newton-Raphson method of iterative solution
AC + d (AC) = 0
So Eq. B20 can be written as
a (AC)
a (AC;
1+1
+ AC
[B21]
a (AC
8Q
<
[B22]
Eq. B22 can be symbolically written as
C.dyf1 +C2
= C3dQ
1
[B23]
163
-------
Similarly one can write the equation of motion as
dyf l + M2 dy = M dQ + M dQ + M [B24]
Writing the governing equations for one cell results in two nonlinear
equations with four unknowns of which two unknowns are common to any two
neighboring cells. For a time line of N nodes there are 2N-2 equations with
2N unknowns. The system is then closed by one downstream and one upstream
boundary condition which produces a total of 2N nonlinear equations with 2N
unknowns. This system of equations is then solved by the Newton-Raphson
iterative procedure proposed by Amein and Fang (1975).
For the iterative solution, the unknown values of the previous time
steps were used as trial values. Substitution of the trial values in the
system of equations resulted in residual continuity and momentum equations.
This system of linear residual equations was solved by a double sweep
technique proposed by Liggett and Cunge (1975).
The computational cells in the advance phase were chosen to assume the
oblique form to allow accurate numerical solution as was proposed by
Strelkoff and Katopodes (1977). For the triangular cells located between
the rectangular and oblique cells, the two boundaries were collapsed
together at the beginning of the time interval and the right boundary
moved by Ax at the end of the time step.
SEDIMENT TRANSPORT
Sediment movement in an irrigation furrow involves the. process of
detachment and transport of the soil particles by the flow. The sediment
transport rate for a flow is determined by the available detached soil
particles and the transport capacity of the flow. Bagnold (1960) reported
that the transport capacity of a flow is proportional to the stream power
which is defined as
Ps = Y Q0 Se [B25]
in which Y = unit weight of water and Q = outflow rate. Due to the high
infiltration rate of the soil at the beginning of irrigation, the outflow
rate from a furrow is low and thus the stream power is low. But with the
decline of infiltration rate of the soil, the outflow rate increases and
thus the stream power increases with time.
But due to the transport of finer sized particles from the bed and
bank of the furrow, the availability of the sediment particles to the
stream decreases with time. This availability function is defined as
. -I
A = \ (t) [B26]
164
-------
in which \ and £ are constants and their values depend on initial soil
moisture and cultivation conditions of the furrow. The sediment transport
rate is given by the product of stream power and availability function.
' Qs = Ps A [B27]
The model solves Eqs. B23 and 824 for all times to compute outflow rate Q
which is used to compute sediment discharge from an irrigation furrow.
165
-------
APPENDIX C
166
-------
NITROGEN LEACHING MODEL
WATER BUDGET
The continuity equation for two-dimensional water flow is
-1 -3
where: RUPT = root uptake of water , kg s m
p = density of water , kg m
w
9 = volumetric water content , m m
t = time , s
-2 -1
Jw = water flux density , kg m s
x refers to the horizontal dimension, m
z refers to the vertical dimension , m
Water flow in the vapor phase is usually negligible in wet soils, and since
it is our purpose to simulate irrigated conditions, transport in the vapor
phase will be ignored in this analysis. Flow of liquid water in soil may
be described by a form of Darcy's equation. For flow in the horizontal
direction an appropriate equation is
Jw = -kWcWm/dx. [C2]
Similarly for the vertical direction,
Jw = -k(H<) (dWi/dz + g) [C3]
~3 -1
where: k(40 = unsaturated hydraulic conductivity, kg m s
¥m = matric potential , J kg
-p
g = gravitational acceleration , m s
167
-------
Assuming a unique relationship between 0 and Vm, applying the chair rule to
the storage term (-rr = -^ ^r) , and substituting Eq. [C2] into Eq. [Cl]
yields an equation in a single dependent variable,
•
CC4]
dQ
where CPT(H') is the specific water capacity (^5). The functional rela-
tionships between 0 and ¥ and between V and k(y) are assumed (Campbell,
1974) to be
0 = 0 (-£-) i/1J [C5]
C V0
and
\t(w} = k fj.i'x rrfii
"^v ' / "^ Viij ) L*^^'J
3 -3
where: 0 = saturation water content , m m
k = saturated hydraulic conductivity , kg m s .
The constants H'e and b are obtained by regressing In y against In 0, with
n = 2 + 3/b. Employing Eq. [C5], the specific water capacity in Eq. [C4]
becomes
h 1) LC71
Difference Form of the Water Budget Equation
Equation [C4] is a non-linear partial differential equation in which
the coefficients are functions of the dependent variable. Analytical
solutions to this class of equations exist only for restrictive initial and
boundary conditions. Therefore numerical approximations are usually used
to solve the equation in the general case. When applying numerical
techniques, space and time are divided into discrete intervals and the
dependent variable is considered only at a finite set of points. The sizes
of the space and time increments are chosen sufficiently small to limit the
errors introduced by replacing derivatives with differences.
To apply Eq. [C4] a newtork of elements is superimposed on the flow
region (Fig. Cl). Within this network storage is assumed to occur only at
the nodes. Transport between nodes is assumed steady over a time step and
168
-------
nodes =
capacitor =
conductor =
i-2
i+2
Figure Cl. A network of nodes and elements superimposed on a flow region.
-------
occurs through the linking elements or conductors. A difference form of
Eq. [4] may be written for each of the internal nodes. The resulting set
of equations is solved for the matric potential at each node at discrete
points in time. The continuity equation for node i,j, in Fig. Cl is
k+1 k -- k+1 k
RUPT. . + CRN. . (V . - V. .) = KA(A¥ . . ' + (l-\) M*. . _
i,J i,J i,J i ,J i.J-1 1 ,J-1
- (1-A) 4 .) + rn^ . + (1-M .^ .
k+1 k — k+1
A*r ~ (1"A) 4* + Q) - ""N/"11"
i,J v ' i,J
[C8]
In this formulation, CRN is the node capacity, K is the element conductance,
\ is a weighting factor in time, and the superscript k is a time step counter.
The overbar on each conductance indicates an average conductance over the
element. The simulated storage capacity of a node is the storage capacity
of all surrounding that node (Fig. C2). Accordingly, the capacity at a
node is
Pw6s /i,,' - (l/b+l)
h K \\ln v IK*-. '
At b 4^6 v H'e y ^^-i'a-1 ^-;_T •'^•'••ij.i '--;-n''/~r [C9J
The mass of water flowing to a node in a time step is a function of a
weighted average of the potential difference at the beginning and end of
the time step. Lambda is the weighting factor. Douglas (1956) reported
that the computational stability of equations similar to [C9] is enhanced
if lambda is assigned a value greater then zero. It can be shown (Ames,
1977) that finite-difference solutions to the flow equation will converge
if lambda is greater than 0.5. They may, however, oscillate around the
true solution. The accuracy of the solution depends on the ratio of the
time step to the time constant of the system. If the time constant is
small such as during infiltration, more accurate solutions are obtained as
lambda approaches 1.0.
From Eq. [C6] we see that conductivity is a function of potential.
The mean conductance of an element is some function of the potentials at
either end of the element. Douglas et al. (1959) used an arithmetic mean
of the conductances at the two ends, while Brutsaert (1971) calculated a
harmonic mean conductance and Tongyai (1976) a geometric mean conductance.
170
-------
i-2
Figure C2. Volume considered in calculation of capacity for
internal and boundary nodes.
-------
Each of these formulations may lead to erroneous results under certain
conditions. For example, during infiltration into a dry layer, the
conductivity calculated with the potential of the dry node is quite low.
Both the harmonic mean conductance, K = 2 Kx K2 / (Kx + K2), and the
geometric mean conductance, K = (Kx K2)1/2, weight this low conductivity
too heavily and too little water is allowed to pass into the dry layer.
In this analysis the element conductance was determined by integrating
Darcy's equation over an element. Substituting Eq. [C6] into Darcy's
equation, separating variables, and integrating from i to i+1 yields
~k
Flow between nodes may also be written as
Jw = K. ?¥. - V) [Cll]
where K. is the element conductance. Combining Eqs. [CIO] and [Cll]
~k
Ki =
Therefore,
-k Ye (z. z. a)/2.
'*•>• ' L
-x
X
^\n L\
J J
KB, KC, and KD are calculated similarly.
Solution of the Difference Equation
Both line-successive overrelaxation (LSOR) (Cooley, 1971; Freeze, 1971)
and alternating-direction implicit (ADI) (Rubin, 1968; Selim and Kirkham,
1973; Vaculin et al., 1979) approaches have been employed to successfully
solve Eq. [C8] subject to specified initial and boundary conditions.
Brutsaert (1971) reported, however, that these classical solution schemes
often oscillate and fail to converge. Such numerical problems are worst
when simulating infiltration fronts for which water potential changes
rapidly in time and space. The numerical problems associated with solution
Eq. [C8] are generally attributed to the non-linear coefficients K(f) and
CPN(MO in the flow equation. Classical approaches can be made convergent
by decreasing mesh spacing or time step size in regions of rapid potential
change or by increasing lambda. Cooley (1970) compared the LSOR and ADI
solutions to Eq. [C8] applied to a well drawn-down problem. Both solutions
172
-------
were stable, converged in a similar number of iterations, and gave nearly
the same solution to the investigated problem. The fact that the ADI
method requires a sweep of nodes in each direction means that it requires
more computer time, however. Cooley (1970) reported that the ADI method is
also quite sensitive to the selection of iteration parameters.
Both the LSOR and ADI solutions to Eq. [C8] are iterative procedures.
With lambda equal to 1, a LSOR formulation of Eq. [C8] for nodes oriented in
the i direction can be written as
KA (Yk+:!;>n+1 - 4>k+:!-'n'1"1) + KB (4'k+1''?+1 - vk+1;»n+1
- KC OP' - v) - KD (v ' - M' + g)
,j CC14]
where n is an iteration counter and k is a time step counter. The calculated
values of potential are then adjusted (relaxed) prior to the next iteration
according to the expression
= ,, _ Jc+l.ru
where w is a relaxation parameter between 0 and 2.
The ADI method, on the other hand, utilizes two systems of simultaneous
equations which taken together represent the partial differential equation.
The ADI algorithm requires that the two systems be solved alternately. For
lines of nodes oriented in the j .direction the ADI equation is
KA (4>' - ^,} + (v, _ ^. + g)
T > J~-"- 1>J 1 "J-j J T»J
— — l^+T n+1 /2 1^+1 n+T /*} "~~ !^+T n l^-f*T n
- KC (V. .' - 4/K+i^n-i"1./^ _ KD (V. .' - V. '" + g)
= (CRN. . £n ) (H'. .'n - ^. .) + RUPT. .
for lines of nodes oriented in the i direction the equation is
KA (¥. .% - *P. .' ) + KB (4*._-,'• - f . '. + g)
173
-------
- KC
(CRN + Cn+1)
1 5 J
f¥k+l,n-H _ k-H.n+l + ,
i,j i+l, j
- 4> ) + RUPT.
' » J '>J
[C16]
where t, is an iteration parameter. Equations [C14], [CIS], and [C16] can be
rearranged to the form
A.
.
-l, J
B. 4
-------
_ k-H.n-H/2, + ™ , k+l,n _ k+l,n+l/2 + ,
-
_ [/'(* f\it 9 * iu ' \ l^n /'in* ill > -i. ^
11 ii"^X ^-IT 1-^1-1 y'
= CPN. . (M/k+l.n+1/2 - i^k .) +
»- TC191
In the i direction it was written as
, k-H.n-H/2 _ ^., + fM/, _ , + ,
i,j-l i,j i-1, J i,j
- KC 0^'n+1 - ¥^'n+1) - KD (Yk+^'n+1 - ipk+1»n+1/2 + g)
= CPN. . (YY .' - VY .) + RUPT. .
Equation [C19] (in the form of Eq. [CIS] was then solved for each row, after
which Eq. [C20] was applied to each column. This sequence constitutes one
complete iteration. After each iteration, except the final one, values of
potential were adjusted according to the expression
tea]
The adjusted potentials were used to calculate the conductances and
capacities for each subsequent iteration.
Root Uptake
Water transport from the soil through the plant to the atmosphere is
often described using electrical analogs as portrayed in Fig. C3 (Cowan,
1965; Campbell et al., 1976). Transpiration can then be described as
RUPT = (Y . - I* )/R? = (V - V n)/R = (V - V )/R
xl r 1 ^ xr xr x ^ r xry r
[C22]
175
-------
(A)
(B)
*>
T = transpiration
f = potential
R = resistance
1 = leaf
x = xylem
r = root
s = soil
l +—w—*—w—*
Rr Rs
-*—W—4
?
3 4 V\0—«—W—t
5
-^v^—t
"W—•
Figure C3. (A) Electrical analog of the transpiration stream and
(B) Resistances in the below-ground portion of the
plant-soil-water system.
-------
-2 -1
where: RUPT = water uptake or transpiration rate , kg m s
4* = water potential , j kg
4 -1 -I
R - resistance , m kg s .
The subscripts r, s, 1, and x refer to the root, soil, leaf and xylem,
respectively.
For numerical simulations water uptake is assumed to occur at each
node in the flow network. The total water uptake rate equals the sum of
uptake rates from each of the nodes;
RUPT =11 RUPT. . [C23]
1 j 1>J
Water uptake for each node is (Eq. [C22])
RUPT. . = (V . . - V . .) / (R . . + R . .) [C24]
i,J s i,j xr i,j' s i,j r i,j'
If root permeability is assumed independent of position in the soil and
axial resistances within the root system are assumed negligible, the
resistance at any node may be written as
where R . is the total root resistance and <(>. . is the fraction of roots
rt i, j
around node i,j. The volume of soil considered when calculating <)>. .is
> J
the same volume used during calculation of soil water storage (Fig. C2).
The fraction of roots in the soil about node i,j is then
zi+l/2 xj+l/2 ZT XT
4>, , = (/ / f(x,z) dx dz) / (/ / f(x,z) dx dz) [C26]
'J 00
where (f(x,z) is a function describing the relationship between root density
and distance in the x and z directions, and x_ and ZT are the distances to
flow region boundaries in the x and z directions.
Soil resistance is determined in a manner analogous to that used to
calculate element conductance in a previous section. The rate of water
uptake by a root can be written as
q/A = -k(Y) dY/dr [C27]
177
-------
where: q = uptake rate , kg s
2
A = 2nr, 1 = root surface area, m
r, = root radius
1 = root length
Substituting Eq. [C6] into Eq. [C27] and integrating from the root surface,
r_ , to the mean distance between roots, r~
k ^e
q/(2nl) In (r /r ) = -j^- (C^)""1 - (Jr)"'1) [C28]
s r
Gardner (1960) represented r» as
r2 = (n f(x,z))"1/2 [C29]
The rate of water extraction may also be written as
q/1 = RUPT. ./(f(x,z) ev) [C30]
' > J
where ev is the volume of soil around node i,j. This volume is also the
volume used to calculate the node's water holding capacity. Substituting
Eqs. [C28] and [C29] into [C30] gives
RUPT. ^ = (C)" - (qp)1") / B^. [C3i]
where
B. . = (1-n) In (r2 f (x,z)) / (TT k V f(x,z)
i , j J. s e
ri - zi_1)) [C32]
According to Eq. [C22]
Equation [C31] can be substituted into Eq. [C33] to yield
178
-------
[C34]
Combining Eqs. [C23] and [C24] gives
RUPT = I I (Y . . - V - .) / (R . . - R . .) [C35]
s i,j xr i,' s i, r i,jy
Assuming negligible xlyem resistance, xylem potential, Y , can be assigned
/\ I
a single value for all positions. Equations [C35] can then be solved for
the root xylem potential,
-RUPT + Z Z pP . . / (R . . + R . .))
ij S1'J S1'J ri'J [C36]
Txr I I (1 / (R . . + R . .))
"
A value of root water potential in Eq. [C34] must yet be determined before
RUPT. - can be computed. From Eq. [C22] we have
' > J
»r1,J-RUI>T1,Jllr1.J+»xr
Equations [C24], [C25], [C34], [C36], and [C37] can now be combined with a
value of RUPT to give an expression for root water extraction at node i,j.
The resulting equation is nonlinear, since RUPT. . is a variable in
Eq. [C37]. 1>J
The total transpiration rate is calculated from a value of potential
transpiration and the ratio of actual to potential transpiration,
RUPT = (F) (TP) [C38]
where F is the specified ratio and TP is potential transpiration. Campbell
(1977) gives as an equation for the ratio of actual to potential
transpiration
F = (s + Y((rva + rjs) / rfi)) / (s + Y((rva + r^ I rfi)) [C39]
where: s = slope of saturation vapor density curve , kg m K
y = thermodynamic psychrometer constant , kg m K
179
-------
r = boundary layer resistance to vapor transport , s m
VS.
r = stomatal resistance to vapor transport , s m
r = "no stress" stomatal resistance , s m
r = combined resistance to radiative and _,
convective heat transfer , s m
Stomatal resistance responds to many variables, including leaf water
potential. Lowering leaf water potential has little effect on stomatal
resistance until the potential approaches a "critical" value, M', , at which
point stomates close and stomatal resistance increases markedly over a
narrow range of leaf water potential (Cowan, 1965; Turner, 1974). The
response of stamatal resistance to leaf water potential can be described by
an expression of the form (Tongyai, 1977),
rvs = rvs (1 + (4/l /M/lc)P)' [C40]
Leaf water potential may be calculated from Eq. [C22] as
V1 = M>xr - RUPT(t) Rr [C41]
Actual transpiration can be partitioned during the day with the equation
(Campbell et al., 1976)
RUPT(t) =2.3 RUPT(0.05 + sin4(0.0175(7.5t - 9.75)))/86400 [C42]
A typical graph of Eq. [C42] is presented as Fig. C4. Transpiration occurs
primarily during the daylight hours, with a peak transpiration rate which
is approximately 2.3 times the mean rate and a nighttime rate which is
about 5 percent of the peak rate. Equations [C39], [C40], [C41] and [C42]
can be combined with an estimate of daily potential transpiration to give
an expression for actual transpiration rate at any time of day. This
expression is also non-linear, since RUPT(t) appears as a variable in
Eq. [C41].
NITROGEN BUDGET
The continuity equation in two dimension for a nitrogen specie is
RUPN = IT " 15 CC43]
180
-------
2 i-
o
X
1 -
oo
cu
IB
OL
03
S-
D.
CO
(O
Figure C4. Graph of the function assumed to represent the diurnal nature of transpiration.
A transpiration rate of 7 mm/day was assumed for this case.
-------
-2 -i
where: J.. = mass flux density , kg m s
p = density of water , kg m
W
0 = volumetric water content , m m
t = time , s
RUPN = root uptake of nitrogen , kg m s
Q = other sources or sinks for the specie
under consideration, such as transfer- __ ,
mations or exchange , kg m s
C = solution concentration , kg kg
In this analysis we are concerned with the movement of a solute out of a
band and will assume that the specie under consideration is unreactive, so
Q = 0. Solute flux is a combination of diffusion and mass flow and may be
described by
JM = DVC + J C [C44]
IN W
where: D = apparent diffusion coefficient combining
the effects of diffusion and hyrdodynamic _„ ,
dispersion , kg m s
-2 -1
J = water flux density , kg m s
Under field conditions movement by diffusion within the profile is small
compared to mass flow (Davidson et al . , 1978). Neglecting diffusion
reduces Eq. [C44] to
[C45]
The vector notation (-») indicates that upstream concentrations are used in
Eq. [C45]. If the flow is from node i,j to node i+l,j, the concentration
of node i,j is used and if flow is from node i+l,j to node i,j, instead,
the concentration of node i+l,j is used. If uptake of nitrogen by plant
roots is passive, the process may be described as
RUPN. . = RUPT. . C. . [C46]
1 > J T » J 1 > J
-3 -1
where: RUPN = root uptake of nitrogen , kg m s
182
-------
RUPT = root uptake of water , kg m s
C = solution nitrogen concentration , kg kg
It is recognized that the distribution of flow velocities in pores,
anion exclusion, cation exchange and molecular diffusion all contribute to
the complexity of solute movement in soils. For many practical purposes,
however, solute movement with bulk water is the dominant process affecting
salt flux in unsaturated soil. In such cases, Eq. [C45] adequately describes
the solute transport process. Bresler and Hanks (1969), for example,
simulated one-dimensional chloride movement assuming that convection was the
only significant transport process. They found close agreement between
measured and predicted chloride concentration distributions throughout the
infiltration, redistribution, and evaporation phases for their irrigated
system. Davidson et al. (1978) also found good agreement between predicted
and measured nitrate concentrations using a simple convective flow model.
While diffusion is not explicitly considered in this analysis, numerical
solution techniques generally cause a smearing of solute concentrations
referred to as numerical dispersion. Under certain conditions this numercial
dispersion can be significantly greater then physical dispersion. Numerical
dispersion results primarily from first-order finite differences
approximations to the derivative in Eq. [C43] (Chaudhari, 1971). The
magnitude of the truncation errors in first-order finite difference
approximations depends on the size of both time and space increments and
their ratio (Lantz, 1971). Lantz (1971) suggests that, with a proper
choice of time and space increments, numerical dispersion can be made to
give equivalent concentration profiles to physical dispersion. Chaudhari
(1971) suggests the use of higher-order differences to minimize numerical
dispersion.
A first-order finite difference to the derivative in Eq. [C43] has been
used in this analysis. The ratio of time to space increments is relatively
large, which should lessen numerical dispersion (Chaudhari, 1971). That
numerical dispersion which occurs is assumed not to be significantly
greater than physical dispersion.
NUMERICAL IMPLEMENTATION
The flow region to which Eqs. [Cl] and [C43] are to be applied is shown
in Fig. C5. Symmetry considerations require that the equations be solved
for the region ABCDEF if adjacent furrows are being irrigated and for the
region ABCGHIJF if irrigation is via every other furrow (a common practice
in the irrigated region of central Washington).
The order in which the equations presented in the previous sections
are applied to this analysis is shown in Fig. C6. Initial conditions are
input to the model and root uptake of water and nitrogen are calculated
according to Eqs. [C24] and [C46]. The initial potentials and values of
RUPT. . calculated from Eq. [C24] are then input to Eq. [C8], which is
1 > J
183
-------
CO
Figure C5. Schematic diagram of the flow regions for a homogeneous,
furrow-irrigated soil.
-------
Initialize system: input soil hydraulic properties, grid points,
initial water contents, initial root distrubution
While time is less than t
Update time
calculate root uptake of water Eq. [C23]
calculate root uptake of nitrogen Eq. [C46]
calculate new soil water potentials Eq. [C8]
calculate new nitrogen concentrations Eq. [C43]
Figure C6. Skeletal flow chart of model calculations.
solved for the new water potentials subject to appropriate boundary
conditions. The new potentials are used to calculate the mass of nitrogen
at each node at the end of the time step.
Root Uptake of Water and Nitrogen
Sets of Eqs. [C24], [C25], [C34], [C36], [C37], and [C39], [C40],
[C41], and [C42] are solved to give values for root water uptake and Eq.
[C46] is then used to determine root uptake of nitrogen. The order in
which these equations are solved is presented in Fig. C7.
The non-linear nature of Eq. [C24] and [C38] is handled by assuming
that root and leaf water potentials change slowly enough so that values
from the beginning of each time step can be used during the entire time
step. Simulations in which root and leaf water potential were determined
iteratively showed no discernable differences in the amount or pattern of
root water extraction relative to those calculated when potentials at the
beginning of the time step were used instead.
Values of constants and symbols used in the root uptake equations are
given in Table Cl. The pattern of root water extraction was found to be
185
-------
ROOT UPTAKE SUBROUTINE
initialize system
for al1 rows
for all columns
calculate root density at each node Eq. [C49]
calculate fraction of roots about each node Eq. [C26]
calculate root resistance Eq. [C25]
calculate B for each node Eq. [C32]
input potential transpiration, time, and soil water potential
calculate ratio of actual to potential transpiration Eq. [C39]
calculate transpiration at time t Eq. [C42]
calculate actual transpiration Eq. [C38]
for all internal rows
for all internal columns
calculate soil resistance Eq. [C34]
calculate xylem potential Eq. [C36]
calculate leaf potential Eq. [C41]
for all internal rows
for all internal columns
calculate root water uptake Eq. [C23]
calculate root nitrogen uptake Eq. [C46]
calculate root water potential Eq. [C37]
return
Figure C7. Flow diagram for root uptake subroutine.
186
-------
TABLE Cl. VALUES OF CONSTANTS AND VARIABLES USED IN THE ROOT UPTAKE
SECTION OF THE MODEL
Term Symbol Value
root density function f(x,z) f(x,z) = sech (TTZ)
root radius r, *lE-3 m
4 -1 -1
total root resistance R . 3E+6 m kg s
4 -1 ~1
total leaf resistance R-, 2E+6 m kg s
-3 -1
slope of vapor density curve s 1.5E-3 kg m K
thermodynamic psychrometer constant y 5E-4 kg m K
boundary layer resistance r 50 s m
va
stomatal resistance r 100 s m
combined convective and _,
radiative resistance r 40 s m
e
critical leaf water potential f, -1200 j kg
empirical constant (Eq. 41) P 10
potential transpiration T 7 mm day
The notation 1E-3 means 1(10) .
sensitive to the function used to describe root density. A comparison of
root water extraction patterns using plausible empirical root density
functions is presented by Schnabel (1981). Potato root distribution reports
by Lesczynski and Tanner (1976) guided our choice of an empirical expression
describing the spatial development of a potato root system.
Calculation of Water Potentials
The grid of nodes to which Eq. [Cl] was applied in the calculation of
water potentials is shown in Fig. C8, with the spacings in the x and z
directions being listed in Table C2. The nodes are more closely spaced near
the furrow bottom, where the water potential changes most rapidly during
infiltration. Nodes are also closely spaced midway between furrows,
because the fertilizer was assumed to be placed in this area.
187
-------
*T*~*~
«>• •
«>• •
• i •
• • • <
CX5
CO
•
•
i > •
«•<
-• •-
I •
Figure C8. Node locations on the flow region used in the simulations.
-------
TABLE C2. NODE SPACING IN THE VERTICAL AND HORIZONTAL DIMENSIONS.
Vertical nodes
1
2
3
4
5
6
7
8
9
10
11
12
Distance from
surface
0.00
0.00
0.03
0.08
0.13
0.18
0.20
0.22
0.25
0.35
0.55
0.80
Horizontal
nodes
1
2
3
4
5
6
7
8
9
10
11
12
13
Distance from AF
0.00
0.00
0.02
0.05
0.10
0.25
0.35
0.40
0.45
0.55
0.70
0.90
0.90
The rapid change in water potential at the onset of infiltration
(irrigation) required small time steps in order to keep the solution stable.
An initial time step of one second was used, with the time increment being
increased by a factor of 1.3 each subsequent time step to a maximum of 3600
seconds. The weighting factor lambda was assigned a value of one; all
other values introduced oscillations into the solution.
Equation [C8] was solved with the modified ADI method (Eqs. [C19] and
[C20]). The form of ADI solution used for this analysis differs from the
classical scheme represented by Eqs. [C15], and [C16] in the following ways:
all values of ¥. . are implicit in Eqs. [C19] and [C20] while only the
• > J
values in one direction are implicit in Eqs. [C15] and [C16]; and no
iteration parameters are required to make the solution converge. A value
of 0.8 was used for the adjustment factor in Eq. [C25].
189
-------
The non-linear nature of Eq. [C8] was handled by using values of
potential from the previous iteration to calculate conductance and capacity
values during a given iteration. Iteration was continued until the
potentials of successive iterations agreed within 0.02 J kg"1. No more
than six iterations were required for convergence.
The symmetry lines across which no flow was assumed (AF and DE or IJ)
were simulated by setting horizontal conductances of nodes on the symmetry
lines equal to zero. No-flow conditions were also imposed at boundaries
CB, CG, and GH. A constant potential was imposed on FJ as the lower
boundary condition. Infiltration was simulated by setting the potential
of the nodes at the furrow bottom (AB) equal to the air entry potential.
Depth1 of water in the furrow was controlled by spacing the nodes around
the furrow bottom to achieve the desired cross sectional area for flow.
At the end of the infiltration period the conductances on AB were also
set equal to zero.
Flux of Water and Nitrogen
The mass of nitrogen at each node at the end of each time step was
determined by solving Eq. [C43] for QC. The mass of water moving to the
node was calculated with Darcy's equation, and upstream nitrogen concen-
trations were used in the solution. Drainage of water and nitrogen was
calculated as vertical movement between the bottom two rows of nodes.
190
-------
APPENDIX D
191
-------
PROCEDURE FOR SEDIMENT ANALYSIS OF IRRIGATION RUNOFF SAMPLES
The following description has been provided by Jim Micheal (personal
communication, USBR, Agronomic Soils and Water Laboratory, Ephrata, WA,
1981). Water samples received by the above laboratory were processed
according to the following procedure:
1. After being logged into the lab^and sequentially numbered, all
sample bottles were weighed on a Mettler P1200 top-loading balance.
2. Filter preparation involved sequentially numbering filters and
placing them in a Freas Precision Scientific model 645 drying oven at 110 C
for 1 hour, with air flow of 1 liter per minute. Anhydrous calcium chloride,
as a desiccant, was placed in the oven to aid in the removal of excess
moisture. Upon removal from the oven, filters were placed in a desiccator
containing calcium chloride, to cool for an additional hour before being
weighed on a Mettler A-30 Electronic Analytical balance. Two iypes of
filters were used in the sediment filtration process; Millipore AP type
Prefilters (Cat. No. AP40 047 05) and Millipore HA type 0.45 urn filters
(Cat. No. HAWP 047 00).
3. Filtration was done using the previously mentioned filters placed
on a 47 mm Gelman 4200 Magnetic Filter Funnel assembly with a 300 ml
capacity funnel. Filtrate was collected in a 1000 ml Kimax (No. 27060)
filter flask. Vacuum for the system was provided by a Cenco Hyvac 14
vacuum pump. After most water h'ad passed through the filter, sediment and
remaining water were stirred by swirling. This slurry was then poured into
one or more filter funnels, each fitted with a prefilter. Care was taken
not to overload any one filter for fear of losing sediment. As a result,
some samples required the use of 5 or more filters. Sample bottles were
carefully washed of all traces of sediment by a plastic washbottle. In the
event that the filtrate was cloudy, it was passed through an additional
Millipore 0.45 pm filter. When the filtration process was complete, the
funnel was carefully removed and any remaining sediment was either washed
onto the filter or into another funnel fitted with a filter. To prevent
loss of sediment, filters were then folded in half and the edges sealed by
wetting and pressing together. They were then placed (stacked if more than
one was used) in a pie pan covered with a paper towel. Sediment-laden
filters were then dried for 2 hours at 110 C in the same drying oven used
to dry the empty filters.
After June 1980, evaporation dishes were used for the more heavily
sedimented samples. The procedure consisted of decanting most of the water
through a filter, then washing remaining sediment into filter funnel fitted
with prefilter. A spatula was used to scrape any sediment left on the
prefilter into an evaporating dish (previously weighed in a Gram-atic
192
-------
balance). With most of the sediment in the evaporating dish, the small
amounts that remained on the spatula, in the filter funnel, and in the
sample bottle were washed into another funnel, to be collected on a second
filter. Evaporating dishes were oven-dried for 24 hours, then placed in
the disiccator to cool before weighing on a Gram-atic balance.
4. Weights were obtained on all sample bottles and dried, sediment
filters. The empty sample bottles were allowed to drain following removal
of all sediment. Tare weights for sample bottles were to the nearest 0.1
gram. Following removal from the drying oven, sediment-laden filters were
placed in a desiccator containing anhydrous calcium chloride for at least
one hour to cool. Sediment filters were weighed on a Mettler A-30
Electronic Analytical balance to the nearest 0.1 mg.
5. Calculation of the sediment load for each sample was conducted by
dividing the measured weight of each sediment sample (mg) by the volume of
water (liters) from the corresponding runoff sample.
193
------- |