.tiz;
.. $3.52..

hifize . . . ‘

. . 3 . . I2 .
:1: :17. . . '11:.‘VJ1. «a;
.: 5 . . , 1...}... ......

1v
1.5.93...
int-'34
0. . a.

1:3 : . .973...

.1. 7.... . . 11;... fl.

. m;

, x .. H".

a

 

 

1:33.13 nit . . n .2 t..v..=:
3:21:53. I?! . t 3.3.33.
. ‘31.. v
saw

1.
. 3.. a: .
.A’J 'Y“43A 4. l|4r14$§oi
.9..." shtiiidrz: r x .. up
i . _ . . , . .
.3 h: a I .S
.. ruflbtuhifi. l3...“ . It a . .lq .
S. 1.... 33:. sh."

figs}.

3... :3 .4 i .

.3} :I .
e

 

x33... .. I...
J3... a}... a? .
t ~>...l:‘>os.
1.1:!
It.

a: .
Q
$415515...
:28... '3:
In. 0.. 3...

.2
fatal.
.5!- .o.3rv:vt..:. I.) 9
.24.. 3th 340...... . .
thLlwznw. .ul «135.11.31.
.. V Eiri p. 31.5.1)...
3.... t‘x;x!...2i 35... x 2“ war...
braavi»lrala...!.. V: o
(.16... 51-2-1...
. :L 1.... 3..
«infra...- . , 3,3.
.53.. .33.... . . . yv. 3.1.1.8...
t. it n...’ at:
. 15.... .90

3., . .; :5
.u v. a... .3az.
.> 12...... 33:.

it... . .
........:.....r. x...
I ....
1. . :39}. 6.“... ..
rr.( 2 I!
" ’cxi
.r.§..:¢.....
12.3....
.33.»...
v. .... y...
A... a!

4.6.5:;

 

‘1'.

I39 Htuumen

my:

xi. :1

. x....lr..\u§
- h)? s

L

[I‘Hlfltfl'

d

1 .
I . J.
k... :5 : ,
2.. .5 . 5...; . 1...
Ed . 1. lazilll ..
. .x it... i.
V .. . .v.

 

7
l
I
‘

THESIS .- _-__,_________

 

 

 

 

2 LIBRARY
)00\ Michigan State
University
This is to certify that the

dissertation entitled

MODELING SEDIMENT AND PHOSPHORUS LOADING IN
A SMALL AGRICULTURAL WATERSHED

presented by

Da Ouyang

has been accepted towards fulfillment
of the requirements for

PhD. , Crop and Soil Sciences
degree In

 

 

%«/é%/c/

Major professor

[My /

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

V DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c10lRC/DateDue.p65—p.1s

 

MODELING SEDIMENT AND PHOSPHORUS LOADING [N
A SMALL AGRICULTURAL WATERSHED

By

Da Ouyang

AN ABSTRACT OF A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Crop and Soil Sciences
2001

Professor Jon F. Bartholic

ABSTRACT

MODELING SEDIMENT AND PHOSPHORUS LOADING IN
A SMALL AGRICULTURAL WATERSHED

By

Da Ouyang

Agricultural nonpoint source pollution is considered the leading cause for water
pollution in the United States. Sediment and phosphorus are two major pollutants that are
responsible for water pollutions. While soil erosion degrades soil productivity, it causes
water quality problem through sediment and nutrients. Excessive fertilization,
particularly from phosphorus, leads to eutrophication which deteriorates surface water
quality. Efforts have been made to minimize agricultural nonpoint source pollution by,
for example, implementing best management practices. Controlling agricultural nonpoint
source pollution requires good information and knowledge on identifying the source
areas and quantifying the pollutant loadings. A water monitoring program is helpful but
costly. A scientifically based model can provide an alternative approach to provide a
quantitative estimation on soil erosion, sediment and nutrient loadings, and to help
identify the source areas.

The goal of this research is to investigate various agricultural nonpoint source
pollution models, and develop a 618 based and spatially distributed approach to better
estimate soil erosion, sediment and phosphorus loading in an agricultural watershed

CODICXI.

A small agricultural watershed, Marshall Drain Watershed, was selected as the
study area. This watershed is approximately 400 acres with 90 percent agricultural land
use. It is a subwatershed of the Sycamore Creek watershed, located in Ingham county,
Michigan. Agricultural nonpoint source pollution, particularly sediments, has been
identified as the major cause of water pollution in the watershed. A multi-year water
quality and land use/tillage management monitoring program has been conducted in the
watershed from 1990-1997. Data from this monitoring program are used in this study.

A Spatially Explicit Sediment Delivery Model (SEDMOD) and the modified
Revised Universal Soil Loss Equation (MRUSLE) are used in this study. These two
models are integrated into Sediment and Phosphorus Loading Model (SPLM). It is 618
based and capable of calculating soil erosion, sediment yield and phosphorus loading.
The results showed SPLM estimated sediment and phosphorus loading with an improved
accuracy compared to other models. Input data required to run the model are minimum
and readily available. The results of this research demonstrate the benefits of using a
spatially explicit model combined with GIS technology. SPLM allows users to identify
the source areas and estimate NPS loadings which may lead to a cost-effective watershed

planning and management for minimizing agricultural nonpoint source pollution.

To my wife Sumei and daughter Jessica

iv

ACKNOWLEDGEMENTS

First of all, I want to thank my wife Sumei for her constant support and
encouragement, and my parents who have never given up on me even in those difficult
years in China. Without their support, urging and understanding, this dissertation would
not be possible.

I thank my major advisor and supervisor Dr. Jon Bartholic for supporting me
financially and emotionally, and for offering me an opportunity to work full-time at the
Institute of Water Research while I was completing my dissertation. I also want to thank
my advisory committee members Dr. Richard Harwood, Dr. Delbert Mokma, and Dr.
Stuart Gage for their time, patience and guidance throughout my Ph.D. program.

My thanks are extended to John Suppnick at Michigan Department of
Environmental Quality, who generously provided me all the water quality monitoring
data in this study. I thank Jerry Gri gar, State Agronomist at USDA-NRCS, for his many
helpful discussions about soil erosion models. I also want to express my appreciation to
Bob Hicks at USDA-NRCS Mason Office for sharing his knowledge about the study

watershed.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................. x
LIST OF FIGURES ............................................................................ xii
CHAPTER ONE INTRODUCTION .................................................... 1
1.1 Overview .................................................................................... 1
1.1.1 Problem Statement ............................................................. 1

1.1.2 Research Needs ................................................................. 3

1.2 Research Objectives ....................................................................... 5
1.3 Thesis Organization ........................................................................ 6
CHAPTER TWO LITERATURE REVIEW ........................................... 7
2.1 Soil Erosion Models ...................................................................... 7
2.1.1 USLE - Universal Soil Loss Equation .................................... 7

2.1.1.1 Rfactor in USLE .................................................... 9

2.1.1.2 K factor in USLE .................................................... 10

2.1.1.3 LS factor in USLE ................................................... 11

2.1.1.4 C factor in USLE .................................................... 12

2.1.1.5 P factor in USLE .................................................... 13

2.1.2 RULSE — Revised Universal Soil Loss Equation ........................ 14

2.1.2.1 R factor in RUSLE .................................................. 14

2.1.2.2 K factor in RUSLE .................................................. 15

2.1.2.3 LS factor in RUSLE ................................................ 17

vi

 

2.1.2.4 C factor in RUSLE .................................................. 18

2.1.2.5 P factor in RUSLE ................................................... 20
2.1.3 MUSLE — Modified Universal Soil Loss Equation ..................... 22
2.1.4 WEPP - Water Erosion Prediction Project ............................... 26
2.2 Soil Erosion Modules in Agricultural NPS Models .................................. 28
2.1.1 AGNPS - Agricultural N onpoint Source Pollution Model ............. 28
2.1.2 Ann-AGNPS — Annualized Agricultural Nonpoint Source Pollution
Model ......................................................................... 31
2.1.3 SWAT - Soil and Water Assessment Tool ............................... 32
2.1.4 EPIC — Erosion-Productivity Impact Calculator ......................... 34
2.3 Sediment Delivery Ratios (SDR) ....................................................... 35
2.3.1 Drainage area and SDR ...................................................... 35
2.3.2 Rainfall-Runoff and SDR ................................................... 40
2.3.3 Drainage Area-Runoff and SDR ........................................... 41
2.3.4 Slope, gradient, and relief-length ratio and SDR ........................ 42
2.3.5 Particle size and SDR ........................................................ 43
2.3.6 Loading coefficients and SDR ............................................. 44
2.3.7 Spatially distributed SDR model ........................................... 45
2.3.7.1 Slope gradient ........................................................ 48
2.3.7.2 Slope shape ........................................................... 48
2.3.7.3 Stream proximity .................................................... 49
2.3.7.4 Overland flow ........................................................ 49
2.3.7.5 Soil texture ........................................................... 50

vii

2.3.7.6 Surface roughness ................................................... 50

2.4 Estimating Soil Erosion, Sediment Yield and Phosphorus Load ................... 50
2.5 Phosphorus Index ......................................................................... 51
CHAPTER THREE MATERIALS AND METHODS ................................ 54
3.1 Description of Study Watershed ........................................................ 54
3.2 Data collection ............................................................................. 58

3.2.1 018 data layers ............................................................... 58

3.2.2 Monitoring data ............................................................... 62
3.3 SEDMOD .................................................................................. 69
3.4 Modified RUSLE ......................................................................... 71
3.5 Integration of SEDMOD and Modified RUSLE ..................................... 78
CHAPTER FOUR RESULTS AND DISCUSSION ................................... 80
4.1 Sediment delivery ratio .................................................................. 80
4.2 Soil erosion ................................................................................. 81
4.3 Sediment yield ............................................................................. 86
4.4 Phosphorus loading ....................................................................... 95
CHAPTER FIVE CONCLUSIONS ...................................................... 104

Appendix A Estimates of Manning’s n for overland flow and soil covers based on
n=0.01 for smooth bare soil (Engman, 1985) ............................. 109

Appendix B The program code to calculate soil erosion and sediment yield based
on results from SEDMOD and Modified RULSE ....................... 110

Appendix C Monitoring data from grab samples from Marshall Drain
Watershed ..................................................................... 116

Appendix D Calculated soil-loss ratio (SLR) from Revised Universal Soil Loss
Equation (RUSLE) ........................................................... 1 19

viii

Appendix E Modeling outputs for the calculated soil erosion and sediment load
from SEDMOD and four soil erosion models .......................... 131

BIBLIOGRAPHY .............................................................................. 140

ix

Table 2.1.1

Table 2.3.1

Table 2.5.1

Table 2.5.2

Table 2.5.3

Table 3.2.1

Table 3.2.2

Table 3.2.3

Table 3.2.4

Table 3.3.1

Table 3.3.2

Table 3.4.1

Table 3.4.2

Table 4.1.]

Table 4.2.1

Table 4.3.1

Table 4.3.2

LIST OF TABLES

P values and slope-length limits for contouring ..........................
Estimated sediment delivery ratios for various land uses ...............
Site characteristics and weighting factors used in Phosphorus Index .
Phosphorus loss ratings for soil erosion used in Phosphorus Index
Site vulnerability for phosphorus loss .....................................

Monitored runoff, sediment and phosphorus data in Marshall Drain
Watershed .....................................................................

Monitored rainfall and intensity, and energy-intensity in Marshall
Drain Watershed ..............................................................

Land use and tillage monitoring results (in acres) for Marshall Drain
watershed (1990—1997) ......................................................

Monitored land use and tillage practices by 10-acre grid cells

Manning’s roughness values for various field conditions
(Engman, 1985) ...............................................................

Percent of soils and clay content in Marshall Drain Watershed .......
K and T values for soils in Marshall Drain watershed ..................

An example for calculated SLR in half-month intervals for Com-
Soybean rotation with the prior year condition ..........................

The maximum/minimum/mean values of sediment delivery ratios
over Marshall Drain watershed area (1991-1997) ........................

Estimated soil erosion from four single event models ..................
Model Efficiencies (ME) ....................................................

Monitored sediment yield and its estimation based on spatially
distributed SDR model (SEDMOD) and four soil erosion models

13

45

52

52

53

65

66

67

7O

7O

74

76

81

83

90

93

 

Table 4.3.3

Table 4.4.1

Table 4.4.2

Monitored sediment yield and its estimation based on spatially
lumped SDR model (equation [2.3.4]) and four soil erosion
models ......................................................................... 94

Monitored phosphorus load and its estimation based on sediment
phosphorus content, spatially distributed SDR model (SEDMOD)
and four soil erosion models ................................................ 101

Monitored phosphorus load and its estimation based on sediment

phosphorus content, spatially lumped SDR model (equation [2.3.4])
and four soil erosion models ................................................ 102

xi

 

Figure 2.1.1
Figure 2.3.1
Figure 2.3.2

Figure 3.1.1

Figure 3.2.1
Figure 3.2.2
Figure 3.2.3
Figure 3.5.1
Figure 4.1.]
Figure 4.2.1

Figure 4.3.1

Figure 4.3.2

Figure 4.3.3

Figure 4.3.4
Figure 4.4.1
Figure 4.4.2

Figure 4.4.3

Figure 4.4.4

LIST OF FIGURES

Examples of three typical slope shapes ....................................
Relationship between the delivery ratio and watershed size ............
The relationship between the delivery ratio and the distance ..........

Study area - Marshall Drain watershed in south of Mason,
Ingham County, Michigan ...................................................

10-acre grid cells in Marshall Drain watershed ..........................
Soils in Marshall Drain watershed ..........................................
Digital Elevation in Marshall Drain watershed ...........................
Integrated Sediment and Phosphorus Loading Model ..................
Estimated sediment delivery ratio in Marshall Drain watershed ......
Estimated soil erosion in Marshall Drain watershed .....................

Monitoring sediment load and estimated sediment load from the
modified RUSLE with SEDMOD and SDR Model [2.3.4] ............

Measured sediment load and estimated from SEDMOD for delivery
ratio, and MU SLE, AGNPS, and USLE for soil erosion ...............

Measured sediment load and estimated from Equation [2.3.4] for
delivery ratio, and MUSLE, AGNPS, and USLE for soil erosion

Estimated sediment yield in Marshall Drain watershed .................
Monitored sediment load and total phosphorus load ....................
Monitoring and estimated phosphorus loading ...........................

Measured phosphorus load and estimated from SEDMOD for
delivery ratio, and three soil erosion models ..............................

Measured phosphorus load and estimated from Eq. [2.3.4] for
delivery ratio, and three soil erosion models ..............................

xii

25

37

38

57

59

60

61

79

84

85

87

88

88

92

97

98

99

99

Figure 4.4.5 Estimated phosphorus load in Marshall Drain watershed ............... 100

xiii

CHAPTER ONE

INTRODUCTION

1.1 Overview

1.1.1 Problem Statement

Since the 19705, following the passage of the Clean Water Act (CW A), great progress
has been made in controlling point source pollution, especially from industrial facilities
and municipal discharges. Nonpoint source (NPS) pollution has now become the leading
cause of surface water degradation in the United States. Nonpoint source pollution is
generally referred to the pollution resulting from diffuse sources. The single major
nonpoint source pollution comes from agricultural cropland. According to the U. S.
Environmental Protection Agency’s (U SEPA) National Water Quality Inventory reported
to Congress, agricultural nonpoint source pollution is responsible for the impairment of
60 percent of the impaired river miles and 50 percent of lake acres in the United States
(USEPA 1994). Although progress has been made over the past few years, agricultural
nonpoint source pollution has been continuously considered as the No. 1 source for water
pollution in rivers and lakes according to a series of EPA’s 1994, 1996 and 1998 reports
for the National Water Quality Inventory (USEPA 1994, 1996, 1998, 2000).

The primary water quality problems from agricultural nonpoint source pollution are
sediment and nutrients. Soil erosion not only degrades agricultural productivity,
sediment-bound pollutants such as phosphorus also cause the major pollution concern in

surface waters. Contribution of nonpoint source pollution from agricultural land use has

been estimated at 64% of total suspended sediment and 76% of total phosphorus (Duda
and Johnson, 1985). Overenrichment of nutrients in freshwater stimulates algal and
rooted aquatic plant growth, and results in oxygen depletion, fish kills, odor problem and
consequently eutrophication (Lee, 1971). Eutrophication impairs the use of surface
waters for recreation, fisheries, industry and drinking (Sharpley and Meyer, 1994;
Sharpley et al. 1994). Environmental and economic impacts for controlling agricultural
nonpoint source pollution are tremendous. It is estimated that the economic damage to
surface water quality caused by sediment and nutrients from agricultural cropland ranges
from 2.2 to 7 billion dollars each year in the United States (Lovejoy et al., 1997).

Soil erosion and sediment delivery are the key processes controlling NPS pollution in
agricultural watersheds. Not all soil erosion will be delivered to waterways due to
deposition. Actually, only a portion of soil loss from agricultural land is transported to a
river. The process of soil erosion involves three steps: detachment of soil particles,
transport, and deposition. Soil erosion by water can be categorized into sheet, rill, gully
or channel, and stream bank erosion. When raindrops fall on soil surface and detach soil
particles, the water runs over the soil surface and removes particles in a thin layer, which
causes sheet erosion. Rill erosion occurs where water concentrates to form small channels.
Erosion from large channels due to large water concentration causes gully erosion (Foster
1986). Sheet and rill erosion are the main sources of erosion from agricultural land.
Widely used soil erosion model, Universal Soil Loss Equation (USLE), was developed
primarily for predicting sheet and rill erosion.

Phosphorus is one of the major pollutants from agricultural nonpoint sources.

Although there are many forms of phosphorus, the particulate phosphorus or sediment

attached phosphorus is the most common form in which phosphorus components move
with eroded soil particles. Previous studies suggest that high phosphorus losses from
croplands are associated with intensive row crop agriculture, fertilizer additions and

tillage, etc. (Nelson, and Logan 1983).

1.1.2 Research Needs

Because of the diffuse nature of NPS pollution, challenges remain in identifying,
assessing and controlling NPS pollution. The processes of NPS pollution and
effectiveness of various NPS control technologies are not fully understood. While
numerous models have been developed in the past three decades, few has been evolved as
national standardized procedure for NPS pollution assessment. Among the US.
government agencies, for example, USEPA and the US. Department of Agriculture
(USDA) use their own different models/ approaches/ procedures for nonpoint source
pollution assessment. Most of nonpoint source pollution models either require extensive
input data, or lack of integration of new technologies available today such as geographic
information system (GIS) which helps spatial analysis, or lack of necessary accuracy in
assessing sedimentation. Especially, for estimating sediment delivery ratio and sediment
load, most models are spatially lumped models, which lack capabilities to identify source
areas or high risk areas. Developing a spatially distributed approach to better estimate soil
erosion and sediment/phosphorus loading with minimum and readily available data sets is
needed.

A single storm event loading model for estimating nonpoint source pollution is

essential to determine the allocation of pollutant loadings among point and nonpoint

 

sources to meet the Total Maximum Daily Load (TMDL) standard. The federal
regulations require the states to identify the waters that require TMDL and rank/prioritize
those impaired waters (U SEPA 1999). The so-called 303(d) list is the list of impaired
waters (stream segments and lakes) that are submitted to USEPA by the states under
Section 303(d) of the Clean Water Act. Development of spatially distributed model with
an ability to estimate sediment load for single storm events will help to better identify and
prioritize the areas with impaired waters.

Some NPS models have been developed for assessing nonpoint source pollution
such as Agricultural Nonpoint Source Pollution Model (AGNPS) (Young, et a1. 1978),
Soil and Water Assessment Tool (SWAT) (Arnold et a]. 1996), among others. These
models can estimate NPS pollution such as soil loss, sediment, nutrient, and pesticide
losses. The major drawbacks for using these models include extensive input data
requirement and lack of accuracy for quantitative estimation of nonpoint source pollution
in agricultural watersheds. Studies from Michigan Department of Environmental Quality
(MDEQ) showed that estimated average annual sediment loading using AGNPS was 88%
- 94% higher than the monitored data in their study watershed (Suppnick 1999). In
addition, Needham (1999) has concluded from his study that Annualized AGNPS (ANN-
AGNPS) is incapable of accurately estimating NPS loadings. Further research is clearly
needed to improve the applicability and the accuracy of the modeling approaches.

Walling (1983) pointed out that two major problems existed in the current
sediment delivery and soil erosion models: temporal lumping and spatial lumping.
Temporal lumping models aggregate the individual storms and attempt to average out the

variations among them. Spatial lumping models, on the other hand, provide a single value

for the entire watershed without considering the variations on local conditions. In this
study, spatially explicit sediment delivery ratio model is integrated with the modified
Revised Universal Soil Loss Equation for estimating sediment and phosphorus load,

which takes into account the temporal and spatial variations in the watershed.

1.2 Research Objectives

The goal of this research was to investigate various agricultural nonpoint source
pollution models, and develop a GIS based and spatially distributed approach to better
estimate soil erosion, sediment and phosphorus loading on a watershed basis.

Objectives of this study are (l) to investigate the various NPS models for soil
erosion and sediment delivery calculations; (2) to modify the Revised Universal Soil Loss
Equation (RUSLE) so that it is suitable to estimate soil erosion for single storm events; (3)
integrate the Modified RUSLE (MRUSLE) and Spatially Explicit Sediment Delivery
MODel (SEDMOD) for estimating sediment yield and phosphorus loading; and (4)
compare and evaluate this integrated spatial modeling approach with other selected
models using monitoring data available in the study watershed.

This study provides an alternative modeling approach for assessing NPS pollution
and identifying its source areas in agricultural watersheds using existing data. The
outcomes of this study may help watershed planners, resource managers and local
governmental and federal agencies and field staff to prioritize the critical areas needing
technical assistance, and implement best management practices to minimize soil erosion,

sediment and nutrient loading and its impacts on water quality.

1.3 Thesis Organization

This thesis is divided into five chapters. Chapter one gives an overview of nonpoint
source pollution problems that exist today, research needs and objectives. Chapter two
presents the literature review regarding various NSP models including major agricultural
NPS models, soil erosion models, sediment delivery ratio models, and sediment yield and
phosphorus loading estimation methods. Chapter three describes the methods and data
used in this study, including description of the study watershed, data source, and data
processing for soil erosion, sediment delivery and phosphorus loading calculations.
Chapter four provides the research results as well as discussion on the significance of
these results including comparison of results from several different models with the water
monitoring data in the study watershed. Chapter five provides with conclusions drawn

from this study and commendations for future study.

CHAPTER TWO

LITERATURE REVIEW

A number of models have been developed to calculate the sediment delivery ratio
and sediment yield. Many nonpoint source pollution models provide more comprehensive
simulations on hydrology, sedimentation, and biological and chemical processes. They
can generally be grouped into two catalogs. One is called statistical or empirical models
such as the Universal Soil Loss Equation (USLE). These kinds of models are statistically
established based on observed data, and are usually easier to use and computationally
efficient. The other kinds of models can be called parametric, deterministic, or physically
based models, such as Water Erosion Prediction Project (WEPP). These models are
developed based on the fundamental hydrological and sedimentological processes. They
may provide detailed simulation with many different sub-components, but usually require
extensive data input and are not easy to use. The following sections will describe a few
agricultural nonpoint source pollution models including both empirically and physically
based models. The soil erosion and sediment delivery components used in the NPS

models are also reviewed.

2.1 Soil Erosion Models
2.1.1 USLE —- Universal Soil Loss Equation

Research on soil erosion and its effect on agricultural productivity started in
19305. During 19405 and 19505, research scientists began to develop a quantitative

procedure for estimating soil loss in the Corn Belt in the United States. Several factors

were introduced to an early soil loss equation, in which slope and practice were primarily
considered. A rainfall factor was later added in the erosion equation (W ischmeier and
Smith 1978). In 1954, the US. Department of Agriculture, Agricultural Research Service
(ARS) established the National Runoff and Soil Loss Data Center at Purdue University to
locate, assemble, and consolidate all available data throughout the United States. More
than 10,000 plot-years of basic runoff and soil loss data were then collected from US.
federal-state cooperative research projects in 49 US. locations. Based on the data
assembled at the Data Center and studies conducted by other soil scientists, Wischmeier,
Smith, and others developed the Universal Soil Loss Equation (USLE) using six factors.
An agriculture handbook describing USLE was published in 1965 (Agriculture Handbook
282) and revised in 1978 (Agriculture Handbook 537) (W ischmeier and Smith 1978).
USLE has been widely used as the major conservation planning tool in the United States
and other countries in the world.

After undergoing several changes in the form of soil loss equations, Wischmeier
and Smith (1978) have established a general format of Universal Soil Loss Equation
(U SLE) which is the product of six factors:

A=R*K*LS*C*P [2.1.1]
Where A = estimated average soil loss in tons per acre per year

R = rainfall-runoff erosivity factor

K = soil erodibility factor

L = slope length factor

S = slope steepness factor

C = cover-management factor

P = support practice factor

2.1.1.1 Rfactor in USLE

R is the rainfall-runoff erosivity factor, which is a measure of the erosive force of
rainfall with a consideration of runoff effect. The energy of a storm is relative to both
volume and intensity of rainfall and runoff. R factor for a given time is a sum of the
energy-intensity (ED values from all storms in that time period. The value of E1 equals
the product of total storm energy (E) times the maximum 30—minute intensity (130). Rain
showers of less than 0.5 in. are omitted from the erosion index computation, unless at
least 0.25 in. of rain fell in 15 minutes (Wischmeier and Smith 1978). Average annual
values of E1 computed from 22-year station rainfall records were used to plot an
isoerodent map for R factor in USLE. The isoerodents connect points with equal rainfall
erosivity.

Total storm energy (E) used to compute R factor in USLE was calculated from the
following equation:

E=916+33110gioI I<=3in /hr. [2.1.2]

E = 1074 I > 3 in/hr. [2.1.3]
Where E = kinetic energy in foot-tons per acre-inch

I = intensity in inches per hour
Energy-Intensity (EI) for a storm is then computed as follows:

E1 = E * 130 [2.1.4]
Where E1 = Energy-Intensity in foot-tons per acre per hour.

130 = maximum 30-min intensity in inches per hour.

2.1.1.2 K factor in USLE

K is the soil erodibility factor, a rate of soil loss per rainfall erosion index unit as
measured on a unit plot which is 72.6 ft. long with a uniform slope of 9 slope, in
continuous fallow, tilled up and down the slope (W ischmeier and Smith 1978). K
represents the effect of soil properties and soil profile characteristics on soil 1055. Some
soils are more erodible than others. Soil erodibility factor is affected by several factors
including soil texture, organic matter, and permeability. It is also interrelated with other
factors such as the rainfall-runoff factor and the cover-management factor. Organic
matter reduces soil erodibility through increased infiltration (reduced runoff) which
reduces the susceptibility of soil to detachment (Grigar and Davis 1995). Different soils
have different K values. For example, clay soils have low K values because they are
resistant to detachment. Sandy soils are easily detached, but usually have low runoff, and
thus may have low K values. Soils with high contents of silt have high K values because
they are easily detached and also produce high rates of runoff.

In USLE application, a simple way to determine K factor is to find the value from
soil erodibility nomograph developed by Wischmeier and others in 1971. The nomograph
represents the influence of five soil and soil-profile parameters on K value: percent
modified silt (0.002-0.1 mm), percent modified sand (0.1-2 mm), percent organic matter
(OM), and classes for structure and permeability. For soils that have less than 70 percent
silt faction, the following equation can be used to approximate the nomograph:

K = [2.1 * 10“4 (12 — 0M)‘-14 + 3.25 (s — 2) + 2.5 (p — 3)] / 100 [2.1.5]

Where K = soil erodibility factor in ton 0 acre - h (hundreds of acre - ft-tonf 0 in) ‘1

10

 

 

 

M = product of percent modified silt (0.002-0.1 mm) * (% silt + % sand)
OM = percent organic matter

5 = soil structure code used in soil classification, and

p = profile-permeability class.

The nomograph is suited particularly well when used for soils with less

aggregated and medium-textures surface in the Midwest (Renard, et a1 1997).

2.1.1.3 LScfactor in USLE
L is the slope length factor and S is the slope steepness factor. L factor and S
factor are usually considered together as LS factor to reflect the influence of topography
on erosion. L is defined as the horizontal distance from the origin of overland flow to the
point where either (1) the slope gradient decreases enough that deposition begins or (2)
runoff becomes concentrated in a defined channel (W ischmeier and Smith 1978). Slope
length can be measured by pacing or measuring in the field. L factor represents the ratio
of soil loss in a particular field to that from the standard plot. It can be expressed as:
L = (M 72.6)m [2.1.6]
Where L = slope length factor
k = slope length in ft.
m = variable slope-length exponent.
The value of m varies with percent slope. m equals 0.5 for percent slope >= 5; 0.4
for slopes between 3.5 to 4.5 percent; 0.3 for slopes of 1 to 3 percent; and 0.2 for slopes

less than 1 percent (W ischmeier and Smith 1978).

11

S factor reflects the influence of slope gradient on erosion. SIOpe can be estimated
in the field by use of a clinometer or similar device, or from contour maps. The LS factor
is a ratio of soil loss per unit area from an actual field slope to that from the standard unit
plot with 72.6-ft long and 9% slope. Values of LS factor can be found in the LS factor
tables. For agricultural landuse, a LS factor table has been developed and is provided in
Agricultural handbook 537 for row-cropped agricultural and other moderately
consolidated soil conditions with little-to-moderate cover. LS factor in USLE was derived
from the equation:

LS = (1/ 72.6)'n (65.41 5in20 + 4.56 sin 0 + 0.065) [2.1.7]

Where 0 = angle of slope. k and m are defined as above.

2.1.1.4 C factor in USLE

C is the cover-management factor. The C factor is used to reflect the effect of
cropping and management practices on erosion rates. It is a ratio of soil loss under a
specific condition to that from conditions defined in USLE plot, i.e. clean-tilled,
continuous fallow. It indicates that how much soil loss can be reduced by different
conditions such as land cover, crop sequence, and management practices. C factor is
considered the one that can be adjusted by implementing conservation tillage to reduce
the overall soil erosion. The soil loss ratios used in USLE to determine C factor was
derived from analysis of about 250,000 plot soil 1055 data. USLE provides a guideline
with tabulated data for determining C factor. Tabulated data contain soil loss ratios for
various cropping system and management at different crop stages. Because effectiveness

of surface cover on soil loss is different over time, six crop-stage periods are used: Period

12

F (rough fallow); Period SB (seedbed); Period 1 (establishment); Period 2 (development);
Period 3 (maturing crop); and Period 4 (residue or stubble). Erosion-index distribution
data were also considered to determine C factor for individual crop-stage periods since it

affects the effectiveness of management practices on soil erosion (W ischmeier and Smith

1978).

2.1.1.5 P factor in USLE

P factor is the support practice factor, a ratio of soil loss with a specific support
practice such as contouring, stripcropping and terracing to that with straight row farming
up-and-down slope. P factor considers practices that modify the flow pattern, grade, or
direction of surface runoff and reduce the amount and rate of runoff. It does not consider
improved tillage practices such as no-till and other conservation tillage, which are
considered in C factor (Foster et al., 1997). The following table lists the P values with

various slope-length conditions for contouring.

Table 2.1.1 P values and slope-length limits for contouring

 

 

 

 

 

 

 

 

Land slope percent P value Maximum length (ft)
' 1 - 2 0.60 400
3 — 5 0.50 300
6 — 8 0.50 300
9 - 12 0.60 120
13 — 16 0.70 80
17 - 20 0.80 60
l 21 - 25 0.90 50

 

 

 

 

 

(from Agriculture Handbook 537, Wischmeier and Smith, 1978)

13

2.1.2 RUSLE — Revised Universal Soil Loss Equation

Recent research, computing technology and new information have led to the
development of a revised Universal Soil Loss Equation, namely RUSLE. The revised
USLE or RUSLE has retained the same form of the equation with same six factors.
However, procedures for calculating these factors have been refined. The key changes in
RUSLE include deriving some of the main factors such as C and R from sub-factors with
15—day intervals, which better reflects the impact of temporal variations in climate,
surface conditions and management practices. K factor in RUSLE also takes into account
subsurface rock fragment effects on infiltration, and seasonal variation, which reflects the
influence of freeze-thaw process. Complex slope segments are considered in RUSLE and
better guideline for determining slope length (L) was provided. RUSLE also gives a
detail and better description of conservation practices to calculate P factor. The following

sections describe some details for computing the six factors used in RUSLE.

2.1.2.1 R factor in RUSLE

Similar to R factor used in USLE, R factor in RUSLE is also calculated from the

sum of energy intensity value (EI). However, a different formula was used to compute E,

the kinetic energy.

e=em[l—aexp(-bi)] [2.1.8]

where e = the kinetic energy

em“ = a maximum unit energy as intensity approaches infinity;

i = total storm rainfall intensity;

a, b = coefficients.

l4

Brown and Foster (1987) recommended that parameters in the question can be
applied as the following equation based on their analysis.

e= 1099 [ l —0.72 exp (-l.27 i)] [2.1.9]
where e = kinetic energy in ft - tonf - acre“l - in'1

i = total storm rainfall intensity in in - h'l

An average rainfall erosivity R value over a long-term period of time can be

calculated:

R=[: (E130),]/N [2.1.10]

i-l
Where (E130), = E130 for storm i,
j = number of storms,

N = year period.

2.1.2.2 K factor in RUSLE

RUSLE has considered the effect of rock fragments and seasonal variations in K
values. Effects of rock fragments include the rocks’ surface coverage that reduces soil
erosion, and subsurface component that increases soil erosion due to reduction in water
infiltration. Surface cover is counted for in the C factor as surface cover similar to crop
residue. The subsurface effect on K value is considered by adjusting permeability classes.

Soil erodibility factor is affected by freeze—thaw cycles. Soils are most susceptible
to erosion during thawing periods while soil erodibility gradually decreases during the
growing season. The K value at any given date can be calculated based on maximum and
minimum K values which need to be first computed from the nomograph and rainfall-

runoff erosivity factor R:

15

Km = (3.0 —— 0.005 R) Kn0m [2.1.11]

Kmin=Km/(8.6-0.019 R) [2.1.12]
Where Km, Km = maximum and minimum K values respectively;

m = K value from nomograph;

R = rainfall-runoff erosivity factor.

The time when maximum and minimum K values occur can be determined from
the following relationships:

tm=154-0.44R [2.1.13]

tm=tw+FF [2.1.14]
where tum, tum. = julien dates for maximum and minimum K;

R = rainfall-runoff erosivity factor;

FF = frost-free period or growing season in days.

Seasonal K values (Ki) can be calculated from the following relationships which
represent two scenarios:

Scenario 1: tmax < tmin
When tm < t, < tmin

K; = K1W (Km,-In / Km) " ((ti — tmax) / At) [2.1.15]
When ti < tum or ti > tmin when average temperature Tm, > 27° F:

K; = Km exp [0.009 (t, - rm,“ + 3656 )] [2.1.16]
Where8= l if(ti-tmin)SO; 8=0if(ti-tmin)>0andforTav527° F,Ki=Kmin_
Scenario 2: tmax > tmin
When t,,,[.,,,,>ti>tmin ,

For average temperature Tav > 27° F:

16

K, = Km“ exp [0.009 (t, - tmin )] [2.1.17]
For average temperature Tav S 27' F: K; = Km“,

When t, > tm or t; < twin ,
Ki=Km(Km/Km)"((Ti—Tm+365 8)/At) [2.1.18]

Where6=lif(ti-tmin)30; 5=Olf(ti-tmin)>0.

2.1.2.3 LS factor in RUSLE

The topographic LS in RUSLE can be computed for both simple slopes similar to
USLE as well as for complex hillsides with a number of slope segments (Renard et al.
1997; McCool et a1. 1991). Slope steepness is calculated from slope angles as:

S= 10.8 sin0+0.03 S<9% [2.1.19]

S = 16.8 sin 0 - 0.50 S 2 9% [2.1.20]
For slopes that are shorter than let., the following equation is used:

5 = 3.0 (sin a)” + 0.56 [2.1.21]
Where S = slope steepness factor,

0 = slope angle.

Foster and Wischmeier ( 1974) suggested that a complex or irregular slope can be
broken into a number of segments with similar characteristics. An effective segment of
topographic factor LS can be computed from the following equation:

L8, = S, xm [ i"n+1 — (i - 1) "”1] /(7.26) m [2.1.22]
Where LS, = effective LS for ith segment,

m = slope-length exponent (usually assumed equal to 0.5)

x = length in ft. of each segment.

17

2.1.2.4 C factor in RUSLE

While USLE calculates the Soil Loss Ratio (SLR) on various crop stages for C
factor, RUSLE computes the SLR on a 15-day time step. When a management event
occurs during the half-month period, the SLR can be calculated at smaller time segments
(Yoder et al. 1997). Soil Loss Ratio (SLR) is calculated based on five subfactors which
contain cropping and management variables affecting soil erosion:

SLR = PLU * CC * SC * SR * SM [2.1.23]
Where SLR = soil loss ratio

PLU = prior-land-use subfactor,

CC = canopy-cover subfactor,

SC = surface-cover subfactor,

SR = surface-roughness subfactor, and

SM = soil-moisture subfactor.

Prior-land-use subfactor (PLU) represents the effects of crop residual from prior
crops and prior tillage practices on soil erosion. It is calculated as:

PLU = c, * c, * exp[(-cu, * B...) + (c,s * Bu, / cf “5] [2.1.24]
Where Cf = surface-soil-consolidation factor,

Cb = relative effectiveness of subsurface residue in consolidation,

B(JIJr = mass density of live and dead roots incorporated surface residue in the
upper inch of soil (lb/acre-in.),

Bus = mass density of incorporated surface residue in the upper inch of soil

(lb/acre-in),

l8

C“; = impact of soil consolidation on the effectiveness of incorporated residue,

cm and cus are calibration coefficients indicating the impacts of the subsurface
residues.

Canopy cover subfactor (CC) represents the effects of energy reduction from
interception of vegetative canopy. The effect is estimated as:

CC=1-Fc*exp(-0.1*H) [2.1.25]
Where CC = canopy cover subfactor (0 — 1),

Fc = fraction of land surface covered by canopy, and

H = distance that raindrops fall after striking the canopy.

Subsurface cover subfactor (SC) is an important subfactor in determining SLR
which represents the effects of subsurface cover such as crop residues, rocks and other
materials that reduce the transportation capacity of runoff water (Foster 1982). It can be
computed as:

SC = exp [ -b * Sp * (0.24 / R..) 0.08] [2.1.26]
Where b = empirical coefficient,

Sp = percentage of land area covered by surface cover, and

Ru = surface roughness.

Surface roughness subfactor (SR) represents the effects of roughness and
disturbance of the surface on soil erosion by reducing the flow velocity, and trapping
water and sediments. This subfactor is computed as:

SR = exp [ -0.66 (Rn — 0.24)] [2.1.27]

Soil moisture subfactor (SM) considers the influence of antecedent soil moisture

on the infiltration and runoff. SM is equal to 1 when the soil profile is close to the field

19

capacity. Since soil moisture is usually high during the crop season in most of the

continental United States, the soil moisture does not need to adjusted in those regions.
After the Soil-1.055 Ratio (SLR) is calculated, the C factor is then obtained by

summing SLRs for each time interval which is weighted by El distribution for that time

period over the annual EI:

C = (: SLR, * Eli ) / EL [2.1.28]

i=1

Where C = C factor in RUSLE,
SLR, = Soil Loss Ratio for the time period i,
E1; = percentage of annual EI for the time period i,
n = number of time periods, and

EL = sum of the E1 percentage for the entire time period.

2.1.2.5 P factor in RUSLE

Support practice factor P in RUSLE represents the effect of specific support
practices such as contouring, stripcropping and terracing on soil erosion. It is the ratio of
soil loss under those support practices to that with traditional upslope and downslope
tillage. Improved tillage practices such as no—till are not taken into account in P factor
since they are already considered in C factor (Foster et al. 1997).

For contouring practice, the P factor is calculated by considering ridge height and

storm erosivity. The base value of P factor can be calculated as follows:

Pb = a (5m — sc)b + Pmb sc < sm [2.1.29]
Pb = c (sc — Sm)d + P...b s, 2 5,, [2.1.30]
Pb=1 sc_>_se [2.1.31]

20

Where Pb = base values of the P factor for contouring,

Sm = slope at which contouring has its greatest effectiveness,

sc = slope for which a value of Pb is desired,

se = slope steepness above which contouring is ineffective,

Pmb = minimum P value for a given ridge height.

The P factor for contouring is then computed as:

P=1—(1—Pb)(1—Pm)l(1-Pmb) [2.1.32]

P factor for stripping practices is calculated from empirical erosion-deposition
model (Renard and Foster 1983). It is computed based on the following relationship:

P,=1-B/gp [2.1.33]
Where P, = value of P factor for contouring,

B = credit for deposition, and

gp = sediment load at the end of slope that would occur if the strips caused no
deposition.

Similarly, P factor for terracing is determined as

P=1—B(l—P,) [2.1.34]
Where P = value of P factor for conservation planning,

B = credit for deposition, and

Py = sediment delivery factor.

2.1.3 MUSLE — Modified Universal Soil Loss Equation

21

USLE was initially designed to predict soil loss for a long-term annual average
soil loss. It is generally not recommended for a short period of time or a single storm
event (W ischmeier 1976). In many cases, however, it is important to know the soil loss
and sediment delivery in a specific time or a storm event. Several researchers have
conducted studies to modify USLE to add its capability to predict sediment yield and soil
loss for a single storm event (Foster et al 1977; and Williams and Bemdt, 1977). One of
the modified versions was developed by Williams and Bemdt in 1977 by substituting the
rainfall erosivity factor R with a runoff effect term, which can be used to predict sediment
yield for a single event:

Y=11.8(Q*qp)0'56KCPLS [2.1.35]
Where Y = sediment yield from an individual storm in metric ton,

Q = storm runoff volume in m3,

qp = peak runoff rate in m3/sec,

K = soil-erodibility factor,

LS = slope length and gradient factor,

C = crop management factor, and

P = erosion-control-practice factor.

The model developers suggested that Q represents the energy for the detachment
process and qp defines sediment transport in this modified USLE (MUSLE). And thus
introducing storm runoff volume (Q) and the peak runoff rate (qp) in the model eliminates
the need to calculate the sediment delivery ratio in order to estimate sediment yield.

Q and qp are computed from runoff model from SCS equation:

Q = (R - 0.2s)2 / (R + 0.85) [2.1.36]

22

q, = b, * Q ”2 [2.1.37]

Where Q = daily runoff,
. R = daily rainfall,

s = retention parameter, and

b1, b2 = constants determined from plotting peak and volume values on log paper.

MUSLE was tested on 26 watersheds in Texas using monthly and annually
sediment data. It was shown that the accuracy of MU SLE generally increases as
watershed area decrease. The model can not be used to determine where deposition
occurs since it does not have sediment routing function (Williams and Bemdt 1977).

Almost at the same time, Onstad, Foster and Meyer have proposed another
version of modification on USLE to predict soil erosion for single storm events (Onstad
and Foster 1974; Foster et al. 1977). They also focused on modifying the rainfall
erosivity term R factor in USLE. They took into account both storm rainfall factor and
storm runoff (runoff volume and peak rate) in erosivity factor. The energy term is given
as a function of rainfall and runoff energy:

W = 0.5 R,. + 15 Q qp “3 [2.1.38]
Where W = energy term which substitutes R factor,

R5. = storm rainfall factor, EI units,

Q = storm runoff volume, in.

q1) = storm peak runoff rate, in. per hour.

It was assumed that detachment is evenly divided by rainfall and runoff. It
considers the detachment and transport capacities for each compartment (calculating unit)

of the watersheds.

23

Foster et al. (1977) proposed a similar form of erosivity term based on runoff
erosivity and runoff ersovity:

R... = 0.5 Rst + 0.5 a vu 0' p, "3 [2.1.39]
Where Rm = modified R factor,

Rs] = storm energy (E130),

01 = coefficient,

Vu = runoff volume,

a p“ = peak runoff rate

The study showed that the modified USLE improved the soil loss estimation for
specific storms compared to USLE. They split USLE into two separate terms to reflect
different sources: rill erosion and interrill erosion, which also contributed to improvement
of accuracy. They suggested that research is needed to define a better relationship to
computer runoff erosivity than the one they suggested (Foster et al. 1977).

Young et al. (1987) used another modified version of USLE in Agricultural
Nonpoint Source Pollution Model (AGNPS) which is a model for a single storm event.
The modified USLE takes slope shape into consideration and introduced a slope shape
factor (SSF). The slope shape factors represent the difference of concave and convex
slopes, and were determined from the chart developed by Wischmeier and Smith (1978).
Soil erosion from single storm events is calculated from the following equation which is
used in AGNPS.

E=EI*K5*Lf*Cf*Pf*SSF [2.1.40]
Where E = soil loss in tons/acre;

E1 = rainfall energy-intensity in hundred foot-ton inch/acre hour;

24

Ks = soil erodibility factor in ton-acre hour/hundred-acre foot-ton inches;

1.; = slope-length factor;

L; = slope-steepness factor;

Cf = cover and management factor;

Pf = support practice factor, and

SSF = a calculated factor to adjust for slope shape.

Variables in this equation EI, Ks, 14', Cf, and Pf are defined as same as USLE. The
slope shape factors were calculated based on gradients for the upper third, the middle
third, and the lower third segments on a 75-foot slope.

Because a different shape of a slope affects the erosion/deposition process, a slope
shape factor is introduced in this equation. An upwardly concave slope may trap runoff
water and sediment, which transports less sediment than upwardly convex slope (Figure

2.1.1).

Concave Convex Uniform

Figure 2.1.1 Examples of three typical slope shapes

In this modified USLE, slope shape factor has a multiplication factor of 1.00, 1.30,

and 0.88 for uniform, convex and concave slope respectively.

2.1.4 WEPP — Water Erosion Prediction Project

25

Different from USLE and RUSLE which are empirical models, the Water Erosion
Prediction Project (WEPP) is a continuous simulation computer program which predicts
soil loss and sediment deposition from overland flow on hillslope, soil loss and sediment
deposition from concentrated flow in small channels, and sediment deposition in
impoundments (USDA 1995). WEPP not only has a component to calculate soil erosion,
it also contains several other components including stochastic weather generator, a
hydrology component, a daily water balance component, a plant growth and residue
decomposition component, and an irrigation component.

WEPP can be used for different scales: hillslope profiles (tens of meters) and
small watersheds (hundreds of meters). For hillslope profiles, WEPP predicts the soil
particle detachment by raindrop impact and transport by sheet flow on interrill areas, and
soil particle detachment, transport and deposition by concentrated flow in rill areas. As an
extension of the WEEP hillslope model, the WEPP watershed model is capable of
identifying zones of sediment deposition and detachment within constructed channels and
representing the effects of agricultural management practices on erosion and deposition
processes.

Erosion model used for hillslope in WEPP is a steady-state sediment continuity
equation which represents the movement of sediment in a rill:

dG

—- = D + D.- 2.1.41

dx f I I
where x = distance downslope in meters,

G = sediment load in kg - s'l - rn'l

Df = rill erosion rate in kg - s'l - m’2

D = interrill sediment delivery to the rill in kg - s'1 - m’2

26

A positive value of rill erosion Df represents detachment while a negative value
represents deposition. The soil loss computed from this equation is expressed in soil loss
per unit land area. Sediment load and sediment transport capacity are calculated on a unit
rill width basis. Sediment transport capacity is calculated using the following equation:

n=hnm 010]
Where T6 = transport capacity,

k. = transport coefficient (m 0'5 - 52 kg“), and

If = hydraulic shear acting on the soil (Pa).

Sediment load is calculated as the load per unit time:

G=GsT°c[w/Rs] [2.1.43]
Where G = sediment load in kg - 5'1 per unit width,

G. = sediment load that is normalized to the transport capacity at the end of the
uniform slope,

Tcc = transport capacity at the end of the uniform slope,

w = rill width in meters,

R, = spacing of the rill in meters.

The total sediment load for the entire storm event is the product of sediment load
per unit time (G) and effective storm runoff duration.

WEPP is written in ANSI FORTRAN 77 and running under MS-DOS operating

system. The windows version is currently under beta test.

2.2 Soil Erosion Modules in Agricultural NPS Models

2.2.1 AGNPS -— Agricultural Nonpoint Source Pollution Model

27

Agricultural Nonpoint Source Pollution Model (AGNPS) developed by Young et
al. (1987) is a process-based single event model, used to simulate sediment and nutrient
transport from agricultural watersheds. This model uses a distributed approach which
divides a watershed into uniform square areas or cells. Input data of approximately 22
parameters are required for each cell. Outputs from AGNPS include volume and peak
runoff, sediment yield, sediment-attached and soluble nitrogen and phosphorus, COD
(soluble chemical oxygen demand yield). Results are provided in a summarized report
and detailed information for each cell. Sediment and nutrient transport are broken down
into five particle size classes: clay, silt, small aggregates, large aggregates, and sand.

AGNPS has four basic components: hydrology; erosion; sediment transport; and
transport of nitrogen, phosphorus, and chemical oxygen demand. The model can be used
to compare the effects of various management practices such as cropping system, and
fertilizer application method and timing on water quality within a watershed. As
described in the previous section, a modified USLE was used to simulate the soil erosion
for a single event. Methods described in Technical Release 55 (TR-55) published by the
Soil Conservation Service (SCS, 1986) is used in the hydrology module to calculate
overland runoff and peak flow.

Sediment routing through the watershed is modeled on a per-cell and per particle
size basis from the headwaters to its outlet. Simulation of sediment transportation and

deposition is derived from the steady state continuity equation:
0. (x) = Q. (0) + 0.1 Ax / L. - j 000 w dx [2.2.1]
0

Where Q, (x) = sediment discharge at the downstream end of the channel reach in

pounds per second,

28

Q,- (0) = sediment discharge at the upstream end of the channel reach in pounds
per second,

Q] = lateral sediment inflow rate in pounds per secound,

x = downslope distance in feet,

Lr = reach length in feet,

D(x) = sediment deposition rate at point x in pounds per second-square foot, and

W = channel width in feet.

Deposition is computed as:

D (i) = Vss / I q * (q. - g.) ] [2.2-2]
Where D (i) = sediment deposition rate at point i between points x and 0 in pounds per
second-square foot,

Vss = particle fall velocity in feet per second,

Q = runoff rate in cubic feet per second-foot,

q8 = sediment flow rate in pounds per second—foot, and

gs: effective sediment transport capacity in pounds per second-foot.

Effective sediment transport capacity (g,) is calculated from the modified stream
power equation (Young et al., 1987):

g,=g*ic*r*v,2/Vss [22m
where r] = effective transport factor,

K = transport capacity factor,

I = shear stress in pounds per square foot, and

Vc = average channel velocity in feet per second.

29

Sediment discharge is calculated for the period when eroded sediment moves
from the upland to the channel, and the period when sediment moves within the channel.

Sediment attached nutrient load is calculated from a submodel used in CREAMS
(The Chemical, Runoff, and Erosion from Agricultural Management Systems) (Frere et
al., 1980):

SED’ = SOIL * SED * ER * 0.892 [2.2.4]
Where SED’ = N or P transported by the sediment in pounds per acre,

SOIL = N or P concentration in the soil,

SED = sediment yield in kilograms per hectare, and

ER = enrichment ratio for N or P.

The current version of AGNPS program is v5.0 which is written in FORTRAN
running on DOS operating system. In order to run the program, the following parameters
for each cell are required: cell number, cell division, receiving cell number, receiving cell
division, flow direction, SCS curve number, land slope (%), slope shape, slope length
(ft.), overland Manning’s n, USLE K factor, C factor, P factor, surface condition constant,
COD factor, soil texture, fertilizer indicator, pesticide indicator, point source indicator,

additional erosion source, and impoundment indicator.

2.2.2 Ann-AGNPS — Annualized Agricultural N onpoint Source Pollution Model
Annualized Agricultural Nonpoint Source Pollution model (Ann-AGNPS) is an

extension of AGNPS from simulating a single storm event to a yearly basis. It is a batch-

process, and continuous simulation model for surface runoff pollutant loading. Ann-

AGNPS is capable of simulating amounts of water, sediment, chemicals (nutrients and

30

pesticides) leaving from one area (cell) and their movement to another location within the
watershed on a daily basis.

The Revised Universal Soil Loss Equation (RUSLE) replaced the USLE in Ann-
AGNPS for soil erosion module. Upstream runoff volume and weighted rainfall including
rainfall, snowmelt, and irrigation are used to calculate runoff curve number. Sediment
loads are calculated for within bank flow and out of bank flow from three sediment
sources: sheet and rill erosion, gully erosion, and bed and bank erosion. Similar to
AGNPS, the sediment is routed by five particle classes: clay, silt, small aggregate, large
aggregate, and sand. Sediment attached and soluble compounds for nitrogen, phosphorus,
and organic carbon are also modeled. A decay function is used based on the reach travel
time, and water temperature.

Ann-AGNPS is still under development. Few studies using Ann-AGNPS have
been reported or published. Needham (1994) used Ann-AGNPS with a geographic
information system (GIS) interface. He concluded that the Ann-AGNPS model can not
accurately predict quantitative contributions for NSP pollutions. It is only capable of

providing qualitative assessment in identifying and ranking the contributing areas.

2.2.3 SWAT - Soil and Water Assessment Tool
Soil and Water Assessment Tool (SWAT) developed by the USDA and Texas A
& M University (Arnold et al., 1996) is a physically based and continuous time model

that provides daily simulation on impacts of management on water, sediment and

31

 

agricultural chemicals. SWAT has five major components: hydrology, weather,
sedimentation, crop growth model, and nutrients.

SWAT uses a weather generator which simulates daily precipitation, daily
maximum and minimum air temperature, solar radiation, wind speed, and relative
humidity. Hydrology module simulates surface runoff, percolation, lateral surface flow,
groundwater flow, evapotranspiration, and snow melt. The hydrology model is based on

the water balance model:

SW: = SW + : (Ri — Q] — ET] — P] - QRi) [2..25]

i=1
Where SW = soil water content — 15-bar water content,

t = time in days,

R = daily precipitation (mm),

Q = daily runoff (mm),

ET = daily evapotranspiration (mm),

P = daily percolation (mm), and

QR = return flow (mm).

Runoff volume is calculated using the SCS curve number equation (U SDA-SCS,
1972) and peak runoff rate is calculated from the modified Rational formula or the SCS
TR-55 method (U SDA-SCS 1986). The Modified Universal Soil Loss Equation
developed by Williams and Bemdt (1977) is used to calculate sediment yield for each
sub-basin. This equation is described in previous section.

Nitrogen (organic N) transported with sediment is calculated using a loading
function developed by McElroy et al. (1976) and modified by Williams and Hann (1978):

YON = 0.001 (Y) (CON) (ER) [2.2.6]

32

Where YON = organic N runoff at the sub-basin outlet in kg/ha,

CON = concentration of organic N in the top soil layer in g/t,

Y = sediment yield in t/ha, and

ER = enrichment ratio.

Similarly, phosphorus (P) transport by sediment is calculated a loading function:

YP = 0.01 (Y) (Cp) (ER) [2.2.7]
Where YP = sediment phase of P 1055 in runoff in kg/ha,

CD = concentration of P in the top soil layer in g/t,

Y = sediment yield in t/ha, and

ER = the enrichment ratio.

A GIS interface in Geographic Resources Analysis Support System (GRASS) has
been developed for SWAT input and output. The program is run in the Unix operating
system. Windows version of SWAT has also been developed. Like other physically based
models, SWAT requires tremendous input data sets to run the model, including crop,
nutrient, pesticide, irrigation data as well as lake water quality data, and information

about ponds and reservoirs.

2.2.4 EPIC — Erosion-Productivity Impact Calculator

Erosion-Productivity Impact Calculator (EPIC) developed in 19805 (Williams et
al., 1984) is a continuous simulation model that can be used to assess the impact of
management practices on agricultural production, and soil and water resources. There are

10 major components in EPIC model: weather simulation, hydrology, erosion-

33

sedimentation, nutrient cycling, pesticide fate, plant growth, soil temperature, tillage,
economics, and plant environment control. EPIC model can be used for a field-size area,
up to 100 hectares in which weather, soils, and management systems are assumed to be
homogeneous.

Erosion model used in EPIC simulates the erosion caused by rainfall, runoff, and
irrigation. It provides three versions of soil loss equations: original Universal Soil Loss
Equation (W ischmeier and Smith 1978) using energy-intensity term (EI) substitutes R
factor for single events, modified USLE developed by Williams (1975), and by Onstad
and Foster (1975).

Among many equations used in the model, rainfall energy (in metric units) is
calculated using the following equation:

RE = AR (12.1 + 8.9 log ( AR /At )) [2.2.8]
Where RE = rainfall energy, and

AR = rainfall amount in mm during a time At interval in hours.

The USLE rainfall energy factor E1 is then computed by multiplying the energy
term by the maximum 0.5-hour rainfall intensity:

E1 = R [12.1 + 8.9 (log rp — 0.434)].(r05) / 1000 [2.2.9]
Where E1 = energy-intensity,

R = daily rainfall amount in mm,

rp = peak rainfall rate in mm/h, and

r05 = maximum 0.5-hour rainfall intensity in mm/h.

34

EPIC model requires users to have considerable amount of technical knowledge
to install and maintain the model. In addition, extensive data collection and conversion to

EPIC format may be required in order to run the model.

2.3 Sediment Delivery Ratios (SDR)

By definition, sediment delivery ratio (SDR) is the ratio of the amount of
sediment that actually leaves an area to the total soil loss in that area. The expression for
computing sediment delivery ratio can be written in the following form:

SDR = SY / E [2.3.1]
where SDR = sediment delivery ratio

SY = sediment yield

E = gross erosion per unit area above a measuring point.

Water quality monitoring including sediment measurement is expensive and time
consuming. In most cases, sediment yield is not measured, but estimated from soil
erosion and sediment delivery ratio. The following sections describe some

models/approaches used to estimate sediment delivery ratios.

2.3.1 Drainage area and SDR (SDR curves):

Relationships between SDR and other factors have been established as curves.
Watersheds with large drainage area and fields with long distances to streams have low
sediment delivery ratios. This is because large areas have more chances to trap soil
particles, thus the chance of soil particles reaching the water channel system is low.

Some researchers suggested that SDR is closely related to the power of -0.2 to the

35

drainage area or the distance to the stream, others suggested the power of -0.1 and -0.3 in
the function (Gianessi et al., 1986; and Roehl 1962). Figure 2.3.1 shows that a
relationship can be established between sediment delivery ratio and drainage area while
Figure 2.3.2 shows another relationship between sediment delivery ratio and distance

from field boundary to a stream.

36

 

.32. £665 6:6 6:2,: EB: .5666» a 8 E2658 :6: 65
we 3:565 65 :0 women Eat a can can. Bea—av Eon—Gen 5233 9:95:23.— ...0d 959...

.fi . may—4.5.0

0000. 0000. coon OOON coo. CON com com com 00. .
fi_____..____ _...__.___.____ _

 

 

11
.Q":

 

 

 

l
~mQhQ

80$
37

.32. .260”. E95 .3.» nos—22m? new 638 boa—mu 65 59.33 Enmcezflom N5." 959".

 

 

 

use. .624 «06595
000. 00. o.

Edadd a u —-dddu- H u —-a-uu -

 

11.4 . .

o'

I

IJILJLl r

.0.

 

0’. “01103 Manuao tualfllpas

Ifiii’n'l 1
o
g

 

 

 

38

The relationships have been generalized as curves called SDR curves. The SDR
curves include SDR vs. drainage area and SDR vs. distance. The drainage area method is
most often and widely used in estimating the sediment delivery ratios in previous
research.

Renfro (1975) developed an equation relating SDR with drainage area. It is based
on Maner’s (1962) equation and the sediment yields observed in 14 watersheds in the
Blackland Prairie Area in Texas. The model shows a good relationship between SDR
and drainage area (R 2 = 0.92). The model can be written as follows:

log(SDR) = 1.7935 - 0.14191 log (A) [2.3.2]
where A = drainage area in km 2

Vanoni (1975) used data from 300 watersheds throughout the world to develop a
model by the power function. This model is considered a more generalized one to
estimate sediment delivery ratio.

SDR = 0.42 A ”-125 [2.3.3]
where A = drainage area in square miles.

The USDA SCS (1979) developed a SDR model based on the data from
Blackland Prairie, Texas. A power function was derived from the graphed data points:

SDR = 0.51 A *1“ [2.3.4]
where A = drainage area in square miles.

Haith and Tubbs (1981) suggested a similar equation for sediment delivery ratio
based on watershed area:

T5,, = 0.47 A 4“” [2.3.5]

39

Where TS]. = transport factor, or sediment delivery ratio, and

A = watershed area in square kilometers.

2.3.2 Rainfall—runoff and SDR

Water is the vehicle for sediment transport. Rainfall and runoff are the driving
forces of sediment delivery. A humid watershed usually has a higher SDR due to more
rainfall. SDR is also associated with rainfall pattern. A longer duration rainfall event
with less intensity has a lower SDR than a short rainfall event with higher intensity. Land
use/land cover is another factor affecting SDR. A watershed with good vegetation cover
has a low SDR because vegetation slows down the runoff rate and allows the eroded soil
particles to deposit. The rainfall factor of the USLE reflects the energy used in soil
detachment while the runoff factor used in the Modified USLE (MUSLE) reflects the
energy used in both sediment transport and detachment.

A SDR model which is used in the Soil and Water Assessment Tool (SWAT)
(Arnold et al., 1996) takes runoff factor into account. The primary form of the SDR
model is

SDR = (q p / r q, ) 0'56 [2.3.6]
where q p = peak runoff rate in mm/hr.

rep = peak rainfall excess rate in mm/hr

= peak rainfall rate(rp)- the average infiltration rate(f).
The average infiltration rate in mm/hr can be estimated by the following equation
f=(R-Q)/DUR [2.3.7]

where R = rainfall in mm.

40

 

Q = runoff volume in mm.
DUR = duration of a rainfall event in hr.
= 4.605 (R / rp)

Therefore, SWAT-SDR model can be re-written as follows:

SDR = ((qp I r,)/ (0.782845 + 0.217155 Q / R )) 0'56 [2.3.8]

This model is developed for estimating sediment delivery for a single event. The
factors qp and rp are peak runoff and peak rainfall in mm/hr for an event, respectively.
Factors Q and R are runoff and rainfall volumes in mm for an event, respectively. Other
units such as in/hr and inch(es) can be used, and can be canceled if used consistently. In
this study, this model was used to estimate SDR based on long-term average rainfall and

runoff data.

2.3.3 Drainage Area-Runoff and SDR
Dency and Bolten (1976) suggested general watershed sediment yield equations

relating deposits in 800 reservoirs to drainage area size and mean annual runoff. The
equations are
S = 1280 Q 0'46 (1.43 - 0.26 log A) for areas where runoff is less than 2 in. [2.3.9]
5 = 1958 e “5’ Q (1.43 - 0.2616g A) for other areas. [2.3.10]
where S = sediment yield (tons per square mile per year)

Q = runoff (in.),

A = watershed area (square miles).

41

2.3.4 Slope, gradient, and relief-length ratio and SDR:

SDR is affected by topographic features of the watershed. A watershed with short
and steep slopes will deliver more sediment to a channel than a watershed with a long and
flat landscape. Shape of a watershed also affects SDR. A narrow watershed may have a
high SDR. The feature of watershed shape can be expressed by relief-length ratio. Relief
of a watershed is defined as the difference of the elevations in the watershed divide and
outlet. Watershed length is the distance of the two points measured parallel to the main
stem drainage from the watershed divide to the point of sediment yield measurement
(SCS, 1971). Relief-length ratio is used as a physiographic characteristic which affects
sediment delivery ratio.

Williams and Bemdt (1972) used slope of the main stream channel to predict
sediment delivery ratio. The model is written as:

SDR = 0.627 SLP 0‘03 [2.3.11]
Where SLP = % slope of main stream channel.

Maner’s studies (1958) suggested that SDR was better correlated with relief and
maximum length of a watershed expressed as relief-length ratio (R/L) than with other
factors. Renfro 1975 modified the model (R2 = 0.97) as follows:

log (SDR) = 2.94259 + 0.82362 log (R/L) [2.3.12]
where R = relief of a watershed, defined as the difference in elevation between the
average elevation of the watershed divide and the watershed outlet.

L = maximum length of a watershed, measured approximately parallel to mainstream

drainage.

42

 

Williams (1977) found the sediment delivery ratio is correlated with drainage area,
relief-length ratio, and runoff curve numbers. He developed a model based on sediment
yield data for 15 Texas basins. The model is expressed as follows:

SDR = 1.366 x 10 '“ (DA) “-0998 (ZL) 03629 (CN) 5-444 [2.3.13]
where DA = drainage area in kmz,

ZL = relief-length ratio in m/km,

CN = long-term average SCS curve number.

2.3.5 Particle size and SDR

SDR is also affected by texture of the sediment materials. Texture of the eroded
materials is associated with sources of erosion. Coarse materials are usually produced by
streamka and gully erosion, while fine materials are often from sheet and rill erosion.
Less energy is needed to transport fine particles (i.e. silt and clay) than coarse materials
(i.e. sands). Thus, sands are more likely deposited in the transport process, while eroded
silt and clay particles are more easily transported downstream. As a result, sediment
containing high clay content will have a high delivery ratio.

Walling (1983) suggested that sediment delivery ratio may be calculated from
proportions of clay in the sediment and in the soil.

SDR (%) = C soil (%) / C sed (%) [2.3.14]
where C so“ (%) = content of clay in the soil (%).

C sod (%) = content of clay in sediment (%).

43

2.3.6 Loading Coefficients and SDR
A more simplistic approach to estimate sediment delivery ratio as well as nutrient
loading is to use a loading coefficient based on the type of land use. Delwiche and Haith
(1983) have established a loading function for predicting runoff, sediment, and nutrient
losses from complex watersheds. The function was based on input data including
watershed landuse, soil information, daily precipitation and temperature records and
rainfall erosivities. They proposed several separate models for different types of land uses:
urban, cropland, forest, and barnyard. These models provide predicted runoff, sediment
load, and nutrient load for each runoff event and for each source in a watershed. The
model for cropland runoff was given as follows:
L8,“ = 0.001 Cs“, X1“ TS]I Ak [2.3.15]
Where LS.“ = solid compound of nutrient export from area k on day t,
Cstk = nutrient concentration in sediment in mg/kg,
X1“ = soil loss in tons/ha,
T5;‘ = solid phase transport factor or delivery ratio,
A. = area of source k, and
0.001 = conversion factor.
The term for soil 1055 (X... ) is computed using USLE modified by Haith and
Tubbs (1981) for single storm events. The transport factor or sediment delivery ratio is
defined by the equation described in the previous section.
Based on Wisconsin watershed data, Novotny and Chesters (1989) estimated

sediment delivery ratios for various land uses (Table 2.3.1).

Table 2.3.1 Estimated sediment delivery ratios for various land uses

 

 

 

 

 

 

 

 

 

 

Land use Impervious Area Sediment Delivery

(%) Ratio (%)

Agriculture < 5 1 - 30

Developing construction < 5 20 — 50

Low density residential, < 20 < 10

unsewered

Parks < 10 < 3

Medium density residential, 30 - 50 30 — 70

partially sewered

Medium density residential, 30 - 50 70 — 100

sewered

Commercial high density > 50 100

residential, sewered

 

 

 

As variations may exist even in the same type of land use, it needs to refine the
sediment delivery ratio for a specific watershed. The above table shows that the sediment
delivery ratio from agricultural land range 1% to 30%. This range may be too great to
apply to a specific watershed to estimate sediment delivery ratio quantitatively. However,
it may provide a general comparison between different land uses.

Rast and Lee (1983) did an extensive literature review regarding nutrient export
coefficients for nonpoint source loading to the lakes. They collected data for generalizing
nutrient export coefficients from various previous studies and government research
documents. It is reported that phosphorus export coefficient is from 0.03 to 0.07 (g P /

m2 - yr) from agricultural and rural areas.

2.3.7 Spatially Distributed SDR
Most empirical models for sediment delivery ratios are spatially lumped models.
That is, one single value of sediment delivery ratio is applied to the entire watershed.

Spatial diversity on local topographic, land use, and soil conditions are not taken into

45

consideration in the most SDR models. Variations on those factors are assumed to be
averaged out within the watershed. On the other hand, spatially distributed sediment
delivery ratio models take into account the local conditions that affect sediment delivery
process. Spatially Explicit Delivery MODel (SEDMOD) is one of few spatially
distributed SDR models available today. This model was newly developed in late 19905
and has not been widely tested for many different watersheds.

SEDMOD is implemented with Arc/Info GIS and can be used to calculate a site-
specific delivery ratio for nonpoint source pollutants (Fraser, et al., 1998; Fraser, 1999). It
takes into account six important parameters that affect sediment transport. These six
parameters are flow-path slope gradient, flow-path slope shape, flow-path hydraulic
roughness, stream proximity, soil texture, and overland flow. SEDMOD uses a cell-based
model with GRID in Arc/Info GIS. A raster GIS layer or a grid is first created for each of
six parameters to represent their effects on sediment transport process. A linear weighting
model was used to estimate the delivery potential which created a composite raster GIS
layer. The delivery potential (DP) is expressed as a percentage.

DP = SGr SGW + SS,SSw + SRr SRw + SPr SPw + STr STw + OFr OFw [2.3.16]

Where DP = Delivery Potential
SG = Flow-Path Slope Gradient,
SS = Flow-Path Slope Shape,
SR = Flow-Path Surface Roughness,
SP = Stream Proximity,
ST = Soil Texture,

OF = Overland Flow Index, and

46

for the subscript letter r and w:
XX, = Normalized rating for parameter XX, ranging from 0 — 100, and
XXw = Relative weight of parameter XX, ranging from 0 — 1.
The weights assigned to these six parameters are normalized so they sum to l.
The composite layer for delivery potential has a range of 0 - 100. The composite layer is
then sealed to have a mean value same as that predicted using a spatially lumped
sediment delivery model proposed by American Society of Civil Engineers (ASCE,
1975). ASCE suggested that sediment delivery ratio varies approximately with —l/8 th
power of drainage area, i.e.
DR.=c Ab [2.3.17]
Where DR. = Delivery Ratio in percentage,
c = A constant for a “small watershed” as a “point erosion source” with 100%
delivery ratio,
A = Drainage area in square miles, and
b = A constant, -1/8
SEDMOD assumed that an average sized 22 meters long plot, usually designed
for Universal Soil Loss Equation (USLE), as a “point erosion source.” Average
percentage delivery ratio for a watershed is calculated as follows:
DR = DP + (DRa - u(DP)) [2.3.18]
Where DR = Final delivery ratio layer, expressed in percentage,
DP = Delivery potential from composite layer,
DRa = Spatially lumped delivery ratio,

u(DP) = Mean value of delivery potential

47

The delivery ratios for individual cells in this adjusted layer have values
distributed around the average delivery ratio, which reflects the magnitude of the six
delivery parameters. It gives a spatially distributed delivery ratio index for a watershed,
instead of one delivery ratio for the entire watershed if using spatially lumped model
alone.

A number of parameters are required to run SEDMOD. These parameters include

local topographic and soil conditions which are described below.

2.3.7.1 Slope Gradient

Digital Elevation Model (DEM) and a stream network grid were used to derive
this parameter. SEDMOD calculates slope gradient averaged over each cell’s unique
flow-path to the stream. Mean flow-path gradient was used in a power function to

calculate SEDMOD slope parameter.

2.3.7.2 Slope Shape

A convex slope with decreasing slope (divergence) has a greater sediment
delivery ratio than uniform slope and concave slope. A concave slope with increasing
slope (convergence) has greater deposition than convex slope. SEDMOD uses DEM data
to derive the slope shape parameter from the profile curvature, which is derived from the
coefficients of a 3x3 cell polynomial surface. Flow-path curvature values were

normalized from 0 for extremely concave to 100 for extremely convex.

48

2.3.7.3 Stream Proximity
Due to the deposition process, sediments are less likely transported to a water
body from sites far away from a stream than from sites near the stream. SEDMOD uses
an exponential decay function to estimate the stream proximity parameter. The decay
function is based on a 50 percent sediment reduction for every 100 meters distance.
Proximity factor = e "0006391) [2.3.19]

Where D = distance to a stream, in meters

2.3.7.4 Overland Flow

SEDMOD uses a topographically based moisture index (Beven and Kirkby, 1979)
to derive the overland flow parameter. It provides a relative measure of soil saturation for
each cell that may be subject to overland flow conditions. The index was computed as
follows in Arc/Info GRID:

M1 = Ln[A/(C T, tan 01)] [2.3.20]
Where M1 = Moisture Index,

A = Contributing area,

C = Grid cell length,

Ti = Local depth integrated soil hydraulic conductivity, and

tan a = Local slope gradient.

The index shows that sites with large contributing area, low soil transmissivity,

small slope gradient are more likely to become saturated and generate overland flow.

49

2.3.7.5 Soil Texture

Clay content (%) in the upper soil horizon was used as a factor in the SEDMOD
to represent differences in transport of soil particles with different size. Compared to
sandy soils, clay and silt soil particles are more easily detached and transported by
moving water. Fine textured soil particles may also carry higher concentration of
nutrients such as phosphorus than course particles (Nelson and Logan, 1983). An
enrichment ratio is used to calculate the percentage of soil fraction in transported

sediment to that in parent soil.

2.3.7.6 Surface Roughness

Surface coverage (vegetation/residues) can reduce surface runoff and sediment
delivery. The surface roughness is determined as a hydraulic roughness coefficient or
Manning’s n that reflects the effect of surface conditions on the erosion and transport of

sediment.

2.4 Soil Erosion, Sediment Yield and Phosphorus Loading

Sediment yield is different from soil erosion in that sediment yield is only a
portion of gross soil erosion, that is, erosion minus deposition. Soil erosion calculated
from the Universal Soil Loss Equation or Revised USLE does not consider deposition
and only for soil loss from sheet and rill erosion. Sediment yield can be estimated from
soil erosion and sediment delivery ratio:

SY = A * SDR [2.4.1]

Where SY = sediment yield,

50

A = soil loss, and

SDR = sediment delivery ratio in percent.

SDR can be calculated from various models described in previous sections. Soil
loss can be calculated from USLE or RULSE. Modified USLE (MUSLE) developed by
Williams (1977) was attempted to eliminate the need to calculate sediment delivery ratio
and can be used to estimate sediment yield directly for single storms.

Two sediment delivery ratio models including SEDMOD, a spatially distributed
sediment delivery ratio model is used for estimate sediment delivery ratios. Four soil
erosion models including the use of modified RUSLE which considers temporal variation
with 15-day time intervals for single storm events are used for soil erosion. Together,
sediment delivery ratio and soil erosion are used to calculate sediment yield. The results
are compared with ground truth monitoring data.

Phosphorus loading is calculated based on sediment yield and phosphorus content
in sediment (PC). The phosphorus load is calculated as:

PL = SY * PC [2.4.2]
Where PL = phosphorus load,

SY = sediment yield, and

PC = phosphorus content in sediment.

Phosphorus content in sediment can be determined from the monitored sediment

and phosphorus loading data.

2.5 Phosphorus Index
In addition to the method for estimating phosphorus loading based on sediment

and phosphorus content, some other methods have been developed in previous studies for

51

assessing phosphorus losses from agricultural runoff. USDA-NRCS has developed a
phosphorus indexing tool to assess phosphorus loss from agricultural land (Lemunyon
and Gilbert, 1993; Stevens et al., 1993). Phosphorus Index uses a number of parameters
including soil, hydrology, and land management that affect phosphorus availability,
retention, management, movement, and uptake. A weighting factor is assigned to each

site characteristic parameters to determine the site vulnerability for phosphorus 1055:

Table 2.5.1 Site characteristics and weighting factors used in Phosphorus Index

 

 

 

 

 

 

 

 

 

 

 

Site characteristics Weighting factors
Soil erosion 1.5
Irrigation erosion 1.5
Runoff class 0.5
Soil P test 1.0
P fertilizer application rate 0.75
P fertilizer application method 0.5
Organic P source application rate 1.0
Organic P source application method 1.0

 

A phosphorus loss rating value is given to each site characteristic. It reflects the
level of the severity with each site characteristic for phosphorus 1055. Five levels are used
in the ratings: none, low, medium, high, and very high with a value of 0, l, 2, 4, 8,

respectively. For example, the following ratings are used for Soil Erosion:

Table 2.5.2 Phosphorus 1055 ratings for soil erosion used in Phosphorus Index

 

 

 

 

 

 

 

 

Soil Erosion Phosphorus Loss Rating
None (0) 0
Low ( < 5 tons/acre) 1
Medium (5 -10 tons/acre) 2
High (10 —15 tons/acre) 4
Very High ( > 15 tons/acre) 8

 

52

 

 

The site vulnerability for phosphorus loss is calculated by multiplying the
weighting factor and rating values to get the weighted value for each site characteristic,
and then summing the weighted values. Site vulnerability is assessed by comparing the

weighted value with site vulnerability chart provided below:

Table 2.5.3 Site Vulnerability for Phosphorus Loss

Total of wei ratin values Site
< 8 Low

8 - 14 Medium
15 — 32
> 32 V Hi

 

This phosphorus assessment tool (Phosphorus Index) provides an easy-to-use
method for assessing agricultural phosphorus 1055. However, cautions must be used when
applying it on specific sites as the weighting factors and ratings are arbitrarily assigned
although they are based on professional judgment. Furthermore, phosphorus index does
not provide a quantitative estimate for phosphorus loading. It may provide the first tier

analysis for phosphorus loss assessment.

53

CHAPTER THREE

MATERIALS AND METHODS

3.1 Description of Study Watershed

A small agricultural watershed, Marshall Drain Watershed, was selected as the
study area (Figure 3.1.1). Note that most images in this dissertation are presented in
color. Marshall Drain is a sub-watershed of Sycamore Creek watershed. It is located in
Vevay township, south of Mason, Ingham County, Michigan. Marshall Drain
subwatershed drains to Sycamore Creek which is a tributary to the Red Cedar River
flowing into the Grand River, and discharges eventually into Lake Michigan. The
drainage area for the Marshall Drain watershed is approximately 400 acres with 90%
agricultural lands. Total acreage of the entire Sycamore Creek watershed is 67,740 acres,
of which about 70% are agricultural landuse (Suppnick 1999). Problems identified in the
Sycamore Creek Watershed from previous studies (USDA 1990) are sediment from soil
erosion, phosphorus, nitrate and pesticides, which cause sedimentation and turbidity
problems, nuisance algae growth, and groundwater contamination. Sycamore Creek
watershed was on the 303(d) list due to low dissolved oxygen problem, and was required
to develop a Total Maximum Daily Load (TMDL) to reduce the sediment load by one
half (Suppnick 1999).

Sycamore Creek watershed has been selected as a Section 319 project in
Michigan. Section 319 of the Clean Water Act was enacted by Congress in 1987, as a
national program specifically aimed at controlling NPS pollution. USEPA developed the

Section 319 National Monitoring Program which funded 22 projects in 18 states for

54

consistently monitoring water quality and land management for 6 to 10 years. Sycamore
Creek watershed is one of the watersheds for water quality monitoring program funded
by Section 319 grants. Water quality and land management monitoring data collected in
Sycamore Creek watershed during 1990-1997 were used in this study.

In an earlier study conducted in Sycamore Creek watershed, it was estimated that
about 75% of the cropland is considered as critical soil erosion areas (within a one-half
mile distance of a stream or drain). Approximately 1,800 acres exceed 12 tons/acre and
13,000 acres range from 8 to 10 tons/acre of soil loss per year (USDA 1990). Prior
analysis of water samples from the stream indicated that concentration of suspended
solids, total phosphorus and nitrate increased dramatically during runoff events.
Phosphorus concentrations during wet weather ranged from 0.04—0.71 mg/L in Sycamore
Creek (USDA, 1990). It is considered that phosphorus levels in surface waters greater
than 0.01 mg/L (10 ug/l) have been associated with increased algae growth in streams
and rivers (Foth and Ellis 1997). The pollutants including sediments and phosphorus are
primarily from agricultural land and thus agricultural nonpoint source pollution is a major
concern in the watershed. In a TMDL case study conducted in Sycamore Creek '
watershed, it was further identified that among the pollutants, sediments are most
responsible to water quality degradation (U SEPA 1992).

The reasons Marshall Drain sub-watershed in Sycamore Creek Watershed was chosen
as the study area include:
1. The watershed is predominated by agriculture with 90% of agricultural land use. This
is an ideal watershed for studying nonpoint source pollutions in an agricultural

watershed context.

55

2. Water quality is an environmental concern in the watershed. Because of water quality
problem, this watershed is on the 303(d) list which requires the development of Total
Maximum Daily Load (TMDL) for pollution reduction.

3. Nonpoint source pollution, particularly sediment, is the major cause of impaired
water quality. No point source discharge and no bank erosion occurs in the watershed.
Therefore, this avoids the complication of multiple discharges for assessing
agricultural nonpoint source pollution.

4. The Water Quality Monitoring project has conducted intensive water quality
measurements since 1990 by the Michigan Department of Environmental Quality
(MDEQ). Land use and tillage management monitoring data have also been collected
by the Natural Resources Conservation Service (NRCS), US. Department of
Agriculture (USDA) during the same period of time. This provides an excellent data

set for agricultural NPS modeling.

56

‘5

S: x.

 

Ingham County, MI Vevay township in Ingham County, MI

4555.5

___._ — '=-- Emir.“ [il
L‘ I ”\, = h‘ UL]

 

 

 

Figure 3.1.1 Study area — Marshall Drain Watershed in south of Mason,
Ingham County, Michigan

57

3.2 Data Collection

3.2.1 GIS Data Layers

Most agricultural NPS models require extensive data inputs. Even with an aid of
Geographical Information System (GIS), there are quite a few data layers that need to be
created before the model can be run. Fortunately, using Sediment and Phosphorus
Loading Model (SPLM) which is built on SEDMOD and the modified RUSLE requires
minimum data sets. All GIS data layers used in this study are readily available or can be
easily created. Digital base maps used include roads, streams, county boundary and soils.
The data were obtained from Michigan Resource Information System (MIRIS). Soil clay
content and soil erodibility factor K obtained from USDA-NRCS were added to soil data
layer by joining the tabular data with the GIS coverage. Soil data layer was calibrated by
Center for Remote Sensing, Michigan State University.

Digital Elevation Model (DEM) data with 30-meter resolution was obtained from
the US. Geological Survey (USGS). These data were used to generate watershed
boundary using Arc/Info command WATERSHED which creates a watershed boundary
similar to that manually drawn from USGS topographic map. DEM was also used to
calculate the slope steepness and slope length, both of which are needed for estimating
soil erosion.

For land use and tillage management, a GIS coverage file with cells of
approximately 10 acres in size was created using Arc/Info commands GENERATE and
FISHNET. Land use and tillage management data were added to this data layer according

to the monitoring data in Table 3.2.4.

58

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

14
15 19
25
31 36
51
5 54 55 57

 

 

Figure 3.2.1 10-acre grid cells in Marshall Drain Watershed

59

 

Figure 3.2.2 Soils in Marshall Drain Watershed

 

Figure 3.2.3 Digital Elevation in Marshall Drain Watershed

61

3.2.2 Monitoring Data

This study used data collected from previous projects conducted in Sycamore
Creek Watershed. From 1990 to 1997, the Michigan Department of Environmental
Quality, Surface Water Quality Division conducted a multi-year water quality monitoring
program in Sycamore Creek watershed. Pollutant loadings, land use and land cover, and
rainfall intensity had been measured and recorded for the spring season from after snow

melt until crop canopy (Suppnick, 1999).

3.2.2.1 Spring Storm Monitoring

Spring storms were monitored using automatic samplers, bubbler flow meters and
recording rain gages in the three subwatersheds. Automatic samplers were used to collect
samples during runoff events. Sampling began each year from March or April when frost
was out of the ground and freezing weather was minimal to July or August when crop
canopy had a complete ground coverage. Samples were collected every 30 minutes to 6
hours after a 1-2 inch increase in stage. More samples were collected during the first part

of the storm than the end of the storm.

3.2.2.2 Spring Grab Sampling

Grab samples were also collected during the spring storm sampling period from
March/April to July/August each year (1990-1997). Grab samples were collected on a
regular basis every week except that the first three years grab samples were collected

occasionally during non-runoff conditions.

62

3.2.2.3 Sample Analysis

Because of laboratory expenses, not all collected samples were used for
laboratory analysis. About 8 samples (some up to 22 samples) for each storm were kept
for analysis. ‘

At least one sample was used for the rising limb of the hydrograph and near the
hydrograph peak, and at least two samples for the descending limb of the hydrograph.
More samples were used for periods with higher suspended solid loads. Samples were
submitted to laboratory for analysis within 24-48 hours after storm events. Samples were
analyzed by the Michigan Department of Environmental Quality laboratory in Lansing,
Michigan. Chemical analysis included suspended solids, total phosphorus, ortho-
phosphorus, Kjeldahl nitrogen, ammonia nitrogen, nitrite, nitrite plus nitrate, chemical
oxygen demand (COD) and turbidity.

Total pollutant load for each storm was estimated from the interval method
(Richards and Holloway, 1987) as follows. Bedload was not included in the calculation.

L=Zci*q[*t;*k [3.2.1]
Where L = storm load (kg);

c = pollutant concentration (mg/l);

q = instantaneous flow associated with the sample (l/sec);
t = time increment for the interval (minutes);

k = unit conversion constant; and

i = individual sample

63

Table 3.2.1

Monitored runoff, sediment and phosphorus data in Marshall Drain

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Watershed

Date Peak Flow Runoff Runoff Suspended Total

(mm/dd/yy) (CFS) Volume Volume Solids Phosphorus
(in) (Cubic ft.) (ton) (kg)

4/20/90 2.28 0.05 69317 0.000 0.000
5/4/90 0.45 0.01 87 l 1 0.002 0.018
5/17/90 2.28 0.05 75287 0.082 0.378
3/27/91 3.82 0.08 125730 1.280 3.180
4/4/91 0.53 0.00 6000 0.000 0.000
4/9/91 2.76 0.05 72220 0.562 1 .420
4/15/91 2.64 0.04 56865 0.144 0.490
4/19/91 2.98 0.06 88749 0.292 1.090
4/23/91 1.52 0.02 23589 0.043 0.199
4/24/92 9.3 0.26 405000 3.940 10. 100
7/ 14/92 1 0.03 41300 0.050 0.220
4/19/93 16 0.32 488508 13.073 28
5/4/93 0.64 0.00 5637 0.016 0.060
6/7/93 2.1 0.06 84710 0.138 0.580
6/ 14/93 0.47 0.00 5499 0.012 0.050
6/19/93 1 1.4 0.18 273000 5.654 12.750
4/12/94 4.93 0.08 116225 0.035 0.220
4/30/94 1 .05 0.01 8202 0.027 0. 150
7/20/94 0.49 0.00 2688 0.1 14 0.240
4/26/95 1 .28 0.01 20427 0.054 0.220
7/4/95 0.48 0.01 14418 0.005 0.050
5/10/96 4.07 0.06 86039 0.647 2.400
5/21/96 3.63 0.02 32168 0.179 1
6/18/96 26.5 0.44 670000 10.846 21.9
4/5/97 3.75 0.06 96123 0.051 0.460
5/5/97 1.27 0.01 16124 0.034 0.150
5/19/97 2.89 0.03 39436 0.051 0.270
6/23/97 1 .69 0.02 31462 0.032 0.190

 

 

 

 

 

 

 

3.2.2.4 Rainfall Monitoring

To get more accurate rainfall data in the watershed, a Belfort rain gauge was
installed in the watershed. The rain gauge was calibrated and checked to determine that it
operated appropriately. Recorded rainfall data were used to calculate rainfall intensity,
and EI (Energy-Intensity), one factor to calculate soil erosion. Monitored rainfall and

intensity in Marshall Drain Watershed are listed in Table 3.2.2.

Table 3.2.2 Monitored rainfall and intensity, and E1 in Marshall Drain Watershed

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Date Total rainfall Max. 30-min. intensity EI

(mm/dd/yy) (in) (in/30 min.)

4/20/90 0.65 0.08 0.58
5/4/90 0.50 0.13 0.76
5/17/90 0.78 0.20 1.93
3/27/91 0.85 0.47 6.20
4/4/91 0.35 0.05 0.18
4/9/91 0.82 0.45 6.18
4/15/91 0.72 0.14 1.67
4/19/91 0.83 0.10 0.94
4/23/91 0.40 0.24 1.26
4/24/92 1.60 0.39 5.90
7/14/92 1.15 0.35 6.00
4/19/93 1.6 0.17 3.593
5/4/93 0.37 0.24 1.524
6/7/93 1.61 0.30 6.337
6/ 14/93 0.45 0.30 2.254
6/19/93 2.14 0.65 17.60
4/12/94 0.90 0.18 1.89
4/30/94 0.48 0.07 0.347
7/20/94 0.97 0.22 8.69
4/26/95 0.64 0.075 0.50
7/4/95 0.83 0.10 10.50
5/10/96 1.14 0.70 4.98
5/21/96 0.9 0.25 3.35
6/18/96 3.18 0.34 14.7
4/5/97 0.50 0.07 0.40
5/5/97 0.56 0.13 0.86
5/19/97 0.61 0.25 2.07
6/23/97 0.57 0.30 2.72

 

 

 

 

 

65

3.2.2.5 Landuse and Tillage Management Data

Landuse and tillage management practices were recorded by staff of the Natural

Resources Conservation Service of the US. Department of Agriculture. The data cover

the same period for water monitoring from 1990-1997. The 10-acre grids were created in

the watershed to reflect the landuse and tillage practices used in those years (Figure 3.2.4).

The monitoring data are listed in the following tables (Table 3.2.3 and Table 3.2.4).

Table 3.2.3 Land Use and Tillage Monitoring Results (in acres) for Marshall Drain
watershed (1990-1997)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Landuse/Tillage 1990 1991 1992 1993 1994 1995 1996 1997
management

B 0 0 190 0 0 0 0 10
C 140 210 20 150 110 0 0 0
NB 0 0 40 0 180 110 170 110
NC 0 0 0 30 0 150 130 20
NW 0 0 110 0 30 20 170
P 170 80 90 80 80 80 60 50
WH 70 80 30 0 0 0 0 0
W0 20 20 20 20 20 20 0 0
R 10 30
G 0 10 10 10 10 10 10 10
P=established Pasture, meadow, hay, alfalfa Wo= Woods

C=Conventional till corn G: Active gravel mining

NC= No-till com
WH= Wheat

NW = No-till Wheat

66

NB = No-till Soybeans
R= Low Density Residential development
B= Conventional till Soybeans

Table 3.2.4 Monitored land use and tillage practices by lO-acre grid cells

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

7 h C
m pppppmmmmmmmwmmmmmmmmmmmmmmaRR...
h
m PppppmmmmmmmmmmmmmmmmmmmmmmMWRM
a% ppppPMMPmmmmpmmmmmmmmmmmmmmmmmm
m
h
mm pppppmmpmmmmpmmmmmmmmmmmmmmmmmc
m
:03 1“
mW PPPPPCCPCCCCPCCCCCCCCCCCCCCMWWW
w
mm PPPPPBBPWWBBBBBBBBBBBPPPWWWMMMB
w PPPPPCCPMMCCPMMMCCCMMCCCMMMWMMC
m PPPPPWMPPPMMPPPPMMWPPM%%PPPMMMC
#
w mnunmwmnmommanunmnmwwmnnasnnnwm

 

 

67

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

41 C C B NW h C NB NC R
42 P P P P P P P B
43 P P P P P P P B
44 C C B NWh C NB NWH NC
45 C C B NW h C NB NWH NC
46 C C WH NC NB NWH NC NB
47 C C B NWh C NB NC NB
48 C C B NWh C NB NC NB
49 C C C NWh C NB NC NB
50 C C C NWh C NB NC NB
51 C C C NWh C NB NC NB
52 C C NB C NC B WH NC
53 C C WH NC NB NWH NC NB
54 C C B NW h C NB NC NB
55 C C C NWh C NB NC NB
56 C C C NW h C NB NC NB
57 C C C NWh C NB NC NB
58 C C C NWh C NB NC NB
P=established Pasture, meadow, hay, alfalfa Wo= Woods

C=Conventional till corn G: Active gravel mining

NC= No-till corn

WH= Wheat

NWH: No-till Wheat

NA= No-till Alfalfa (year of planting)

68

G1: Inactive gravel mining

R: Low Density Residential development
B= Conventional till Soybeans

NB = No-till Soybeans

3.3 SEDMOD Model

Basic GIS data layers required by SEDMOD are generally readily available. Basic
data layers include clay content, DEM, and land use/land cover. Optional data layers
include streams which can be derived from DEM if not available, and saturated soil
transmissivity in inches per hr. A number of secondary data layers are derived from the
basic data layers.

Vegetation roughness is derived from land cover. The roughness layer is created
by reclassifying the landuse/land cover layer using Manning’s roughness obtained from
literature. Table 3.3.1 lists the roughness coefficients for some field conditions. More
detail data about Manning’s roughness values are included in the appendix. DEM is used
to derive a number of data layers including slope and slope shape, and cell travel distance.
Soil moisture index is derived from DEM and soils.

When the secondary data layers are generated, raster data layers for the six
parameters used in SEDMOD are then created in Arc/Info GRID to compute the
sediment delivery ratio layer in the watershed. The calculations are based on data
processing with GRID, which uses an algorithm for cell-based raster data layers. GRID is
a powerful tool used for spatial modeling.

Soil clay content data layer is created using ArcView GIS by joining the soil clay
content in tabular data (Table 3.3.2) with the digital soil coverage file. Percentage of clay
in the upper soil horizon is used as a factor to represent the differences in particle size

transport (Fraser 1999).

69

Table 3.3.1 Manning’s roughness values for various field conditions (Engman, 1985)

 

Field condition

Manning’s Roughness value

 

 

 

 

 

 

 

 

Fallow Smooth, rain packed 0.01-0.03
Medium, freshly disked 0.1-0.3
Rough turn plowed 0.4-0.7
Cropped Grass and pasture 0.05-0.15
Clover 0.08—0.25
Small grain 0.1-0.4
Row crops 0.07—0.2

 

 

 

Table 3.3.2 Percent of soils and clay content in Marshall Drain Watershed

 

 

 

 

 

 

 

Soil Name Percent (%) of land Clay (%)
(Soil Symbol)
Colwood (Co) 45 5-26
Marlette (Ma) 38 10-18
Capacn (Ca) 5 10-18
Aubbeenaubbee (An) 4 8-15
Owosso (Ow) 3 5-18

 

 

 

 

7O

 

3.4 Modified RUSLE (MRUSLE)

Inherited from USLE, RUSLE was designed to estimate the average annual soil
1055. However, it can be modified to estimate soil loss for single events (Onstad and
Foster, 1974; Foster, 2000). Modification of RUSLE is similar to that for USLE made by
Onstad and Foster (1974) which modified the USLE’s R factor by accounting for both
rainfall and runoff effects on erosion. The modified RUSLE can be expressed as follows:

ERO=(EI)m*K*LS*SLR*P [3.4.1]
Where BRO = 1055 erosion for a single storm event, tons/acre,

(EI)m = Modified Erosion Index term with consideration of rainfall and runoff
effects,

K = soil erodibility factor,

LS = slope length-steepness factor,

SLR = soil loss ratio, and

P = support practice factor.

The modified Erosion Index term is calculated using the following equation:

(EDm=a*EI+b*Q*qp [3.4.2]
where (EDm = energy term with consideration of rainfall and runoff effects,

E1 = Energy-Intensity,

Q = storm runoff volume, in.

qp = storm peak runoff rate, cubic ft. per second (CPS), and

a, b = coefficients. (a = 0.0475 and b = 0.825 were used in the study watershed)

71

In addition to the modified RUSLE, other forms of modification are also
calculated for comparison purpose. They include equation [2. 1.38] proposed by Onstad
and Foster (1974) which accounts for both rainfall and runoff effects, and equation
[2.1.40] developed by Young et al. (1978) which is used in AGNPS and adjusted by a
slope shape factor. The original USLE’s EI term without modification is also used in soil

erosion calculations and used to compare with other models.

3.4.1 Slope Shape Factor

The slope shape factor (SSF) in AGNPS’ modified USLE is used to reflect the
difference of slope shapes affecting net erosion/deposition at a specific location. Average
empirical multiplier coefficients are used in AGNPS’ modified USLE equation: 1.30 for
convex slopes meaning the erosion is 30% more than the uniform slopes; and 0.8 for
concave slopes meaning that erosion is 20% less than the uniform slopes due to the
deposition that may occur in that type of slopes. Concave and convex slopes are
determined using a command of Arc/Info GIS called CURVATURE with the digital
elevation model data (DEM). Slope curvature is calculated as the second derivative of the
surface, i.e. the change of rate in the slope gradient or the slope of the slope. It affects the
acceleration/deceleration of water flow, which influences erosion/deposition process
(Moore et al. 1991; Zeverbergan and Thorne 1987).

Running CURVATURE command on DEM data will get a new grid with the
information of 510pe curvature. A positive value indicates a convex slope and a negative
value indicates a concave slope while the zero value means no change on slope, i.e. an

uniform slope. After the curvature layer is created, a slope shape factor layer is generated

72

by assigning the slope shape factor multiplier to grid cells. Then the slope shape factor

layer is used to adjust USLE for estimating soil loss.

3.4.2 LS factor calculation

L factor and 8 factor are usually considered together to combine the effect of
slope-steepness and slope-length, which basically reflects the terrain on a given site. For
this study, a model developed by Moore and Burch (1985) is used to compute LS factor.
They developed an equation to compute length-slope factor:

LS = (A5 / 22.13)In (sin B / 0.0896)u [3.4.3]
where: m = 0.4 —0.6 and n =1.2 —1.3.

LS = computed LS factor.

A5 = specific catchment area, i.e. the upslope contributing area per unit width of
contour (or rill), in m2 / m. It is calculated in Arc/Info using the function called
FLOWACCUMULATION multiplied by the squared cell size and divided by the cell
size.

For 30 meter resolution DEM grids, As can be calculated as:

A5 = calculated flowaccumulation * 30 * 30 / 30 (for cell size = 30 m) [3.4.4]

B = 510pe angle in degrees. It is calculated in Arc/Info using the function called
SLOPE with option PERCENTRISE which is 100 times Tan B. Then B is calculated
using “Atan” function in Arc/Info.

Tan B = slope (in percentrise)/ 100 [3.4.5]

B=Atan (Tan B)

73

LS factor is calculated with Arc/Info GIS which creates a separate LS factor data

layer for the modified RUSLE.

3.4.3 Other factors in Modified RUSLE

The following table lists the main soils in the watershed with soil name, soil
symbol, K factor and soil loss tolerance level T. According to the official definition of
soil loss tolerable level, T value is defined as the maximum amount of soil loss in tons
per acre per year, which can be tolerated and still permit a high level of crop productivity

to be sustained economically and indefinitely (W ischmeier and Smith, 1978; Renard, et al.

1997).

Table 3.4.1 K and T values for soils in Marshall Drain Watershed

 

 

 

 

 

 

 

 

 

 

Soil Name K T

(Soil Symbol)
Colwood (Co) 0.28 5
Marlette (Ma) 0.24 5
Capacn (Ca) 0.32 5
Aubbeenaubbee (An) 0.24 5
Owosso (Ow) 0.24 5

 

RUSLE computerized program is used to calculate the soil loss ratio (SLR) sub-
factor. The option “Time varying scenario” was chosen. So that it is computed for the soil
loss ratio sub-factor in half-month time period. The SLR values for the half-month period
when the storm event occurred are used. SLR is calculated for each crop rotation in the
study watershed. An example of the output of SLR is listed in the following table for the

two-year crop rotation with Com-Soybean. More results are listed in the appendix.

74

SLR is calculated using the equation [2.1.23] based on the five sub-factors: the
prior-land-use subfactor (PLU), the canopy-cover subfactor (CC), the surface-cover
subfactor (SC), the surface—roughness subfactor (SR), and the soil-moisture subfactor
(SM). After the SLR is calculated for the all crop rotations, it is then joined with the GIS
coverage to create a new data layer for C factor in the watershed. Since different crops
were grown in the watershed each year, a separate C factor data layer is created for each

year for each 10-acre grid cells.

75

Table 3.4.2 An example for calculated SLR in half-month intervals for Corn-Soybean

rotation with the prior year condition

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Dates % PLU CC SC SR SM SLR % EI
Residue
Cover
10/15/1 25 0.643 1 0.463 0.536 1 0.159 0.2
10/16 - 10/31/1 24 0.654 1 0.478 0.555 1 0.173 2
11/1 - 11/15/1 22 0.672 1 0.503 0.593 1 0.201 2
11/16 - 11/30/1 21 0.682 1 0.519 0.632 1 0.224 1
12/1 - 12/15/1 20 0.686 1 0.525 0.666 1 0.24 1
12/16 - 12/31/1 20 0.686 1 0.527 0.694 1 0.251 0
1/1 - 1/15/2 20 0.684 1 0.527 0.714 1 0.258 0
1/ 16 - 1/31/2 20 0.682 1 0.526 0.73 1 0.262 1
2/1 - 2/15/2 20 0.68 1 0.526 0.744 1 0.266 1
2/ 16 - 2/28/2 20 0.678 1 0.526 0.756 1 0.27 1
3/1 - 3/15/2 20 0.677 1 0.529 0.77 1 0.276 1
3/16 — 3/31/2 19 0.681 1 0.54 0.79 1 0.291 2
4/1 — 4/15/2 18 0.69 1 0.559 0.812 1 0.313 2
4/ 16 — 4/30/2 16 0.704 1 0.588 0.834 1 0.345 3
5/1 - 5/4/2 15 0.715 1 0.613 0.847 1 0.371 1.1
5/5 - 5/14/2 6 0.74 0.983 0.814 0.883 1 0.523 2.7
5/15/2 2 0.733 0.967 0.925 0.889 1 0.582 0.3
5/ 16 - 5/31/2 2 0.732 0.938 0.933 0.897 1 0.574 7
6/1 - 6/4/2 2 0.719 0.908 0.942 0.906 1 0.557 2.4
6/5 - 6/15/2 1 0.693 0.761 0.969 0.889 1 0.454 6.6
6/ 16 -— 6/30/2 1 0.625 0.51 0.974 0.904 1 0.281 9
7/1 - 7/15/2 1 0.538 0.35 0.98 0.918 1 0.17 9
7/16 — 7/31/2 0 0.505 0.267 0.984 0.93 1 0.124 10
8/1 - 8/15/2 0 0.512 0.267 0.987 0.941 1 0.127 10
8/ 16 — 8/31/2 0 0.518 0.297 0.99 0.95 1 0.144 9
9/1 - 9/15/2 0 0.522 0.399 0.992 0.958 1 0.198 7
9/ 16 — 9/30/2 0 0.524 0.487 0.994 0.965 1 0.245 6
10/1 - 10/14/2 0 0.524 0.487 0.994 0.969 1 0.246 2.8
10/15/2 93 0.527 1 0.04 0.971 1 0.021 0.2
10/16 - 10/31/2 92 0.534 1 0.041 0.973 1 0.021 2
11/1 - 11/15/2 54 0.44 1 0.195 0.457 1 0.039 2
11/ 16 — 11/30/2 53 0.452 1 0.2 0.502 1 0.045 1
12/1 - 12/15/2 52 0.457 1 0.202 0.544 1 0.05 1
12/16 - 12/31/2 52 0.459 1 0.201 0.578 1 0.053 0
1/1 - 1/15/3 51 0.458 1 0.2 0.604 1 0.055 0
1/ 16 — 1/31/3 51 0.457 1 0.2 0.624 1 0.057 1
2/1 - 2/15/3 51 0.456 1 0.199 0.641 1 0.058 1

 

76

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/16 — 2/28/3 51 0.455 1 0.198 0.657 1 0.059 1
3/1 - 3/15/3 51 0.455 1 0.199 0.676 1 0.061 1
3/16 - 3/31/3 50 0.462 1 0.203 0.702 1 0.066 2
4/1 - 4/15/3 48 0.474 1 0.21 1 0.732 1 0.073 2
4/16 — 4/30/3 46 0.493 1 0.226 0.762 1 0.085 3
5/1 - 5/14/3 43 0.519 1 0.247 0.788 1 0.101 3.7
5/15/3 28 0.515 1 0.399 0.824 1 0.169 0.3
5/16 - 5/31/3 26 0.536 0.97 0.423 0.837 1 0.184 7
6/1 - 6/14/3 23 0.572 0.862 0.468 0.859 1 0.198 8.4
/15/3 17 0.573 0.791 0.57 0.825 1 0.213 0.6
6/16 — 6/30/3 16 0.59 0.692 0.595 0.84 1 0.204 9
7/1 - 7/15/3 14 0.606 0.466 0.637 0.864 1 0.155 9
7/16 — 7/31/3 12 0.59 0.205 0.674 0.883 1 0.072 10
8/1 - 8/15/3 12 0.597 0.144 0.672 0.901 1 0.052 10
8/16 - 8/31/3 24 0.621 0.215 0.44 0.916 1 0.054 9
9/1 - 9/15/3 43 0.642 0.486 0.228 0.93 _ 1 0.066 7
9/16 — 9/30/3 45 0.656 0.67 0.214 0.942 1 0.089 6
10/1 - 10/9/3 41 0.662 0.703 0.249 0.948 1 0.1 1 1.8
10/10 - 10/14/3 70 0.671 1 0.089 0.951 1 0.057 1

 

Note:

SLR=PLU*CC*SC*SR*SM

SLR = Soil loss ratio; PLU = the prior-land-use subfactor;
CC = the canopy-cover subfactor;
SR = the surface-roughness subfactor, and SM = the soil-moisture subfactor.

SC = the surface-cover subfactor;

77

 

3.5 Integration of SEDMOD and Modified RUSLE

The spatially explicit sediment delivery model SEDMOD has its own menu
driven interface written in Are Marco Language (AML) which is run in Arc/Info
environment. Because this model only calculates the sediment delivery, a soil erosion
model with the modified Revised Universal Soil Loss Equation is added. This integrated
model (Sediment and Phosphorus Loading Model - SPLM) is capable of computing
sediment delivery ratio, soil erosion for single storm events, sediment yield and
phosphorus loading if the phosphorus content in sediment is known. Results from this
integrated model are compared with other combinations of sediment delivery ratio model
and soil erosion models in the next chapter.

The integrated model SPLM is built on SEDMOD and the modified RUSLE. Data
required to run SEMOD and the modified RUSLE are then required to run SPLM. The
data layers and structure of SPLM are shown in Figure 3.5.1. Some factors such as C
factor in soil erosion model are computed using the RUSLE program. A GIS data layer
for C factor is then created by joining the tabular data with a coverage file for the
watershed. Slope length and slope steepness factor (L8) is computed from DEM data
layer directly.

Because SPLM is run under Arc/Info GIS, input data sets in common data format
such as DEM and soils which are readily available can be used. The results can be used
for a variety of spatial analysis and representation. The integration for SPLM is also

written in AML and can be run in Arc/Info environment.

78

22.3% .68: 65.86.. aaofiaoi .25 2.65.8.6. 8.28.... E.» 656E

0:68..

maocamoca

> \\ BEER \
© I\.....4_\|N\ as \

\ .266. x

\ \ £3596 :8

\ E250 >20

\ __ow

 

 

 

 

 

 

 

 

 

 

 

 

 

\\co_mo.w __om\A ,

 

 

 

 

 

22> .8568 I

 

 

 

 

 

 

 

 

 

\ .263 o
Nongomzucmq
[ism 526$
mmoccmaom
|1\ Lo>oo ucm._\

 

 

 

79

CHAPTER FOUR

RESULTS AND DISCUSSION

4.1 Sediment Delivery Ratios

Two models were used to estimate sediment delivery ratio in Marshall Drain
Watershed. One is the newly developed spatially distributed model SEDMOD; the
second was the spatially lumped empirical model equation [2.3.4]. The reason for
choosing equation [2.3.4] for comparison is that it was found in a pervious study
conducted in Saginaw Bay watershed of Michigan, that this model provides more
accurate results than other spatially lumped statistical models that were tested in the study
watershed (Ouyang and Bartholic 1997).

Sediment delivery ratios were calculated for 1991 — 1997 using SEDMOD, which
reflected the spatial variations and year-to-year different land uses. Table 4.1.1 lists the
maximum, minimum and mean sediment delivery ratios over the grid cells in the
watershed. Although these statistical data for sediment delivery ratio are similar over the
years, there are some spatial variations of sediment delivery ratios in watershed due to the
changes in land use. Figure 4.1.1 shows that sediment delivery ratios in the study
watershed based on 1993 and 1996 data. The spatial distribution of sediment delivery
ratios are slightly different from year to year due to land use change. Higher delivery
ratios give indications of areas that have potential to contribute more sediment.

Sediment delivery ratios were also calculated using an empirical model with
drainage-area-based SDR [2.3.4] which is a spatially and temporally lumped model.

According to the equation, the calculated sediment delivery ratio is about 0.537 which is

80

relatively high due to the small size of the watershed in which sediments have a short

distance to travel to water systems.

Table 4.1.1 The Maximum/Minimum/Mean values of sediment delivery ratios over

Marshall Drain Watershed area (1991-1997)

 

 

 

 

 

 

 

 

 

 

 

 

 

1991 1992 1993 1994 1995 1996 1997
Min. 0.23 0.25 0.25 0.24 0.22 0.21 0.21
Max. 1.00 1.00 1.00 1.00 1.00 1.00 0.99
Mean 0.365 0.364 0.364 0.364 0.365 0.365 0.365
4.2. Soil Erosion

 

Four models were used to calculate soil erosion. They are the original form of
USLE with El term substituting R factor for the single events (equation [2.1.1]); modified
USLE used in AGNPS (equation [2.1.40]); modified USLE by Onstad and Foster (1974)
with a modified energy term W (equation [2.1.38]); and modified RUSLE (MRUSLE),
equation [3.3.1]. Soil erosion estimated from these four models for 25 storm events are
listed in Table 4.2. 1.

The results show the original USLE gave a higher estimate for soil erosion than
the other three models, followed by AGNPS which is adjusted by a slope shape factor
(SSF). The Onstad-Foster model and MRULSE estimated soil erosion with an average of
8.5 tons per acre per storm and 4 tons per acre per storm, respectively. The original
USLE predicts soil erosion ranging from 0.43 to 68.24 tons per acre for storm events with
an average of 14.39 tons per acre per storm, while AGNPS estimates soil erosion with an
average soil loss of 13.28 tons per acre per storm.

Because no direct measurements were made for soil erosion from the fields in the

Marshall Drain watershed, soil erosions estimated from these four models were carried

81

into the comparisons made for sediment yield which the monitored data is available.

Figure 4.2.1 shows the soil erosion in two major storms occurring in 1993 and 1996.

82

Table 4.2.1 Estimated soil erosion from four single event models

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Date MRUSLE Onstad-Foster AGNPS USLE
(tons) (tons) (tons) (tons)
3/27/91 1.106 6.710 11.439 12.394
4/4/91 0.025 0.231 0.398 0.431
4/9/91 0.960 7.715 13.657 14.798
4/15/91 0.383 2.245 3.691 3.999
4/19/91 0.616 2.097 2.856 3.098
4/23/91 0.261 2.193 3.829 4.152
4/24/92 5.650 9.936 13.333 14.438
7/14/92 0.731 7.266 13.103 14.276
4/19/93 42.082 32.677 31.657 34.528
5/4/93 0.839 8.669 16.756 17.194
6f7/93 3.178 26.51 1 46.516 50.765
6/14/93 0.869 9.072 16.545 18.054
6/19/93 9.740 37.225 62.571 68.240
4/12/94 0.413 1.247 1.811 1.959
4/30/94 0.036 0.315 0.548 0.592
7/20/94 0.281 2.953 5.456 5.903
4/26/95 0.105 0.771 1.273 1.383
7/4/95 0.854 8.947 16.419 17.844
5/10/96 1.209 7.587 13.110 14.154
5/21/96 0.750 5.872 10.462 1 1.309
6/18/96 29.156 28.281 38.701 41.770
4/5/97 0.389 0.721 0.675 0.730
5/5/97 0.089 0.775 1.374 1.473
5/19/97 0.247 1.715 3.000 3.204
6/23/97 0.175 1.566 2.830 3.023

 

83

 

ES 89 Ba 82 :6 88.:
8:896; 565 Ears. s 8:3. 52.8 Eaéuam confirm F. 3 659”.

 

852935 £90 .3523). E co_wo._m __om UmumEzwm fimé 9:9".

as:
com A I
8m-8~ I
8m - on. I
8? - 8. I

 

as:
com A
com - 8“ I
8“ - on.
o? - 2: I
2: - 8 W1.
1 on nv U
238 5.45 :8

 

85

4.3 Sediment Yield

Sediment yield was calculated using the form of equation [2.4.1], i.e. the product
of soil erosion and sediment delivery ratio. The results are generated from a combination
of using two sediment delivery ratio models and four soil erosion models, which produce
8 estimates for sediment yield for each storm event. The integrated model, Sediment and
Phosphorus Loading Model (SPLM) is based on spatially explicit sediment delivery ratio
mode SEDMOD and the Modified RUSLE (MRUSLE). The outcomes of SPLM are
compared with results from other 7 combinations of sediment delivery ratio models and
soil erosion models.

Table 4.3.3 lists the monitored sediment yield and its estimation based on four
soil erosion models and spatially distributed SDR model (SEDMOD). Table 4.3.4 lists
the results from the four soil erosion models and an empirical SDR model [2.3.4].

The model results and monitoring data are plotted on Figure 4.3.1- 4.3.3. The line
shown on the figure is 1:1, or a perfect fitting line. Points above the 1:1 line indicate that
the model overestimated sediment yield while below the line means the model
underestimated the sediment yield. Statistical data for the correlation between the
measured sediment yield and estimated sediment yield from these 6 models are included

in Table 4.3.2 and Table 4.3.3.

86

25

 

 

 

 

 

 

 

 

 

 

 

 

x
a; 20
r:
2
2 A
o
3'. 15 x
g ASPLM
- xuodol [2.3.4]
8
‘0 10
i
.E
'95
111 5 w
X A
A
o l I l l
O 5 1O 15 20 25
Measured Sediment Yield (tons)

Figure 4.3.1 Monitoring sediment load and estimated sediment load from the modified
RUSLE with SEDMOD (SPLM) and SDR Model [2.3.4]

87

 

8

 

N
OI
V

 

 

 

\

 

‘ / e Onstad-Foster
I I AGNPS

A USLE

 

 

 

 

 

 

01

Estimated Sediment Load (tons)
3 a:
0
o
O
I

 

 

O

 

l' j

5 10 15 20 25 30
Measured Sediment Load (tone)

0

Figure 4.3.2 Measured sediment load and estimated from SEDMOD for delivery ratio,
and three soil erosion models.

 

 

8818
I

 

 

to
or
H

 

O Onstad-Foster

A
! / I AGNPS
4 . A USLE

 

 

 

 

d .5
o 0"

Estimated Sediment Load (tone)
8
I

 

 

0 10 20 30 40
Measured Sediment Load (tons)

Figure 4.3.3 Measured sediment load and estimated from Equation [2.3.4] for delivery
ratio, and three soil erosion models.

88

Figure 4.3.1 shows the results using the modified RUSLE for soil erosion, and
SEDMOD and Equation [2.3.4] for sediment delivery ratios. Statistical data in Table
4.3.1 show that the modified RUSLE with SEDMOD (i.e. SPLM) has the best results
with the highest R square (0.95) and the lowest standard error (0.85). With the modified
RULSE, empirical sediment delivery ratio model [2.3.4] also has a fairly good fit with
monitoring data with a R square of 0.95, but has a higher standard error (2.87). Onstad-
Foster version of modified USLE model has R square of 0.62 with SEDMOD which is
similar to that with Equation [2.3.4]. AGNPS version of modified USLE which takes into
account a slope shape factor does not make much difference with the original USLE, both
of which have a poor correlation between the monitoring data and modeling results. They
have generally overestimated the sediment yield with both sediment delivery ratio models
(Figure 4.3.2 — 4.3.3). It indicates these models have systematic errors.

From the results in this watershed, it is shown that SPLM (the modified RUSLE
combined with SEDMOD) provides the most accurate estimates.

To compare the model results and monitoring data, we also used one of the most
useful methods to evaluate the modeling results, which is called Model Efficiency (ME).
Model efficiency was first used by Nash and Sutchliffe ( 1970) and later used by many
researchers in water related modeling (Green and Stephenson, 1986; Risse et al. 1993;
and Rapp et al. 2001). The model efficiency is defined as follows:

‘2 (Qm- — (2.642

ME = 1 — ‘j‘ _ [4.3.1]
2(Qmi "' Qm)A2

i=1

 

Where ME = model efficiency,

89

Qmi = measured value of event i,
Qci = computed valued of event 1,
a; = mean of the measured values, and

n = number of events.
The higher the ME value, the better results a model provides. A model provides
perfect results when MB is l. The ME values for the four soil erosion models with two

sediment delivery ratio models are listed in Table 4.3.2.

Table 4.3.1 Model Efficiencies (ME)

 

 

 

 

SEDMOD Eq. [2.3.4]
MRULSE Onstad- AGNPS USLE MRULSE Onstad- AGNPS USLE
Foster Foster
ME 0.93 -0.05 -1.98 -2.77 0.58 -1.03 -5.97 -7.47

 

 

 

 

 

 

 

 

 

The model efficiency for SPLM is highest (ME=0.93) when SEDMOD is used
with the modified RUSLE (MRUSLE). It is better than Eq. [2.3.4] when it is used with
the modified RUSLE (ME = 0.58) and also better than other models. This is consistent
with statistical analysis data that SPLM (using SEDMOD with the modified RUSLE) has
the highest correlation coefficient (R2 = 0.95) and the lowest standard error (SE = 0.85).

Figure 4.3.4 shows the sediment yield on cell-by-cell basis for two major storms
in 1993 and 1996. The results are from SPLM (i.e. SEDMOD and the modified RUSLE).
The output map illustrates the high sediment contributing areas in the watershed. By
identifying the high contributing areas, it helps prioritize the implementation of Best
Management Practices (BMPs) to reduce soil erosion. Equation [2.3.4] is a spatially

lumped model and does not have the capability to provide spatial representation of

90

 

sedimentation. SPLM by integrating SEDMOD, a spatially distributed sediment delivery
ratio model, with the modified RUSLE, provides a new alternative modeling approach

which is able to identify the risk areas with an improved accuracy.

91

8:296; £90 .3526: 5 26568 8558 4.3. 659".

85 as:
coax I coma I
8m-8~ I ooméom I
8289' 828.-
on; -2: I
02 -8 (Vin.
om uv ,Fil.

2381.85.8w

92

 
 

Table 4.3.2

Monitored sediment yield and its estimation based on spatially distributed
SDR model (SEDMOD) and four soil erosion models

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Date Monitored SPLM Onstad-Foster AGNPS USLE
Sediment (tons) (tons) (tons) (tons)
Yield
(tons)
3/27/91 1.280 0.385 2.573 3.961 4.31 1
4/4/91 0 0.009 0.088 0.138 0.150
4/9/91 0.562 0.334 2.957 4.727 5.145
4/15/91 0.144 0.133 0.860 1.277 1.390
4/19/91 0.292 0.219 0.820 1.006 1.099
4/23/91 0.043 0.093 0.858 1.348 1.473
4/24/92 3.940 2.190 4.247 5.102 ‘ 5.597
7/14/92 0.050 0.276 3.022 4.885 5.385
4/19/93 13.073 16.140 13.818 11.961 13.242
5/4/93 0.016 0.322 3.672 5.962 6.606
6f7/93 0.138 1.215 11.176 17.522 19.406
6/14/93 0.012 0.332 3.824 6.232 6.903
6/19/93 5.654 3.708 15.624 23.480 25.979
4/12/94 0.035 0.156 0.518 0.676 0.738
4/30/94 0.027 0.013 0.124 0.195 0.212
7/20/94 0.1 14 0.104 1.210 2.007 2.193
4/26/95 0.054 0.042 0.339 0.502 0.551
7/4/95 0.005 0.332 3.835 6.314 6.937
5/10/96 0.647 0.462 3.200 4.981 5.415
5/21/96 0.179 0.289 2.499 4.007 4.365
6/18/96 10.846 1 1.423 12.216 15.056 16.364
4/5/97 0.051 0.161 0.328 0.27 6 0.301
5/5/97 0.034 0.030 0.286 0.456 0.493
5/19/97 0.051 0.076 0.581 0.921 0.985
6/23/97 0.032 0.055 0.537 0.879 0.940
R2 0.95 0.62 0.37 0.37
Standard Error 0.85 2.87 4.86 5.38

 

93

 

Table 4.3.3 Monitored sediment yield and its estimation based on spatially lumped
SDR model (equation [2.3.4]) and four soil erosion models.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Date Monitored Modified Onstad- AGNPS USLE
Sediment RUSLE Foster (tons) (tons)
Yield (tons) (tons)
(tons)
3/27/1991 1.230 0.594 3.603 6.143 6.656
4/4/1991 0 0.013 0.124 0.214 0.231
4/9/1991 0562 0.516 4.143 7.334 7.947
/15/1991 0144 0.206 1.206 1.982 2.147
4/19/1991 0292 0.331 1.126 1.534 1.664
4/23/1991 0043 0.140 1.178 2.056 2.230
4/24/ 1992 3940 3.034 5.336 7.160 7.753
7/14/1992 0050 0.393 3.902 7.036 7.666
4/19/1993 13-073 22.598 17.548 17.000 18.542
5/4/1993 0.016 0.451 4.655 8.998 9.233
6/7/1993 0133 1.707 14.236 24.979 27.261
6/14/1993 0012 0.467 4.872 8.885 9.695
45/19/1993 5 .654 5.230 19.990 33.601 36.645
4/12/ 1994 0035 0.222 0.670 0.973 1.052
4/30/1994 0027 0.019 0.169 0.294 0.318
7/20/1994 0.1 14 0.151 1.586 2.930 3.170
4/26/ 1995 0.054 0.056 0.414 0.684 0.743
7/4/1995 0005 0.459 4.805 8.817 9.582
5/10/1996 0.647 0.649 4.074 7.040 7.601
5/21/ 1996 0.179 0.403 3.153 5.618 6.073
6/18/1996 10.346 15.657 15.187 20.782 22.430
4/5/ 1997 0051 0.209 0.387 0.362 0.392
5/5/1997 0034 0.048 0.416 0.738 0.791
5/19/1997 0051 0.133 0.921 1.611 1.721
6/23/1997 0032 0.094 0.841 1.520 1.623
R2 0.95 0.61 0.36 0.36
Standard Error 1.20 3.66 6.94 7.56

 

94

 

4.4 Phosphorus Loading

Once the sediment yield was calculated, one could estimate phosphorus loading
based on sediment load and phosphorus—to-sediment ratio or phosphorus content in
sediment. A previous study has shown that sediment attached phosphorus is the major
form in phosphorus loading (Nelson and Logan, 1983). Figure 4.4.1 shows the sediment
load and total phosphorus load from the monitoring data in the study watershed. It is
illustrated that a good linear relationship exists between phosphorus loading and sediment
loading (R2 = 0.99). Thus, the phosphorus content in sediment, or a phosphorus-to-
sediment ratio can be obtained from the slope of the line. Based on the monitoring data in
the study watershed, the phosphorus-to-sediment ratio is about 2.13 kg phosphorus per
ton sediment. The results of phosphorus loading estimated from this method (equation
[2.4.2]) are listed in Table 4.4.1 and Table 4.4.2.

Since phosphorus loading is based on the sediment yield and phosphorus content
in sediment, its accuracy relies on the accuracy from both factors. Similar to the model
efficiency for estimating sediment yield, SPLM has the highest model efficiency (ME =
0.91), followed by the case when the modified RUSLE is used with Eq. [2.3.4] which has
the model efficiency of 0.54.

Figure 4.4.2 shows that monitoring phosphorus loading and estimated phosphorus
loading along with the 1:1 line. Points are distributed in the both sides of the 1:1 line,
indicating that the model does not systematically overestimate or underestimate the
phosphorus loading. The correlation coefficient for SPLM is 0.936 with the standard

error of 2.12. It is better than the modified RUSLE with Eq. [2.3.4] which has a

95

correlation efficient of 0.935 and standard error of 2.95. Results from other models are
plotted on Figures 4.3.3 and 4.3.4, which show the similar trends as sediment loads.

The advantage of using SPLM based on SEDMOD with the modified RUSLE is
that not only can it provide more accurate results, it also provides the spatial
representation for helping identify the high contributing areas. This information can help
prioritize the efforts for more efficient nonpoint source pollution control.

Similarly to results of sediment yield, SPLM based on SEDMOD with the
modified RUSLE provides more accurate results than other models tested in this study.
Figure 4.4.5 shows phosphorus loading in two major storms in 1993 and 1996. The
output map helps identify the potential high phosphorus contributing areas in the
watershed.

The fact that phosphorus load is highly correlated with sediment load suggests

that management practices that are used to control sediment may control phosphorus

loading as well.

96

 

 

01)
O

y = 2.1343x
R2 = 0.9943 /

 

l
\

 

-5 N
0| 0

 

 

Phosphorus load (kg)

—l
U! 0
\

 

 

 

/
O i r

O 5 1 0 1 5
Sediment load (ton)

 

Figure 4.4.1 Monitored sediment load and total phosphorus load

97

 

 

a 8
l
i
\

 

A .
x / ASPLM

:1: Eq. [2.3.4]

 

 

 

 

 

Estimated Phosphorus Load (kg)
3 8 8
\ 1

\

 

 

 

 

o
I:
b

fi

10 20 3O 4O 50 60
Measured Phosphorus Load (kg)

0

Figure 4.4.2 Monitoring and estimated phosphorus loading from the modified RUSLE
with SEDMOD and SDR Model [2.3.4]

98

 

 

 

 

 

. A / O Onstad-Foster
I

I AGMDS
’ l A USLE

 

 

 

8 8 8 8 8
L‘—

 

..s
o

 

Estimated Phosphorus Load (kgj

 

O r T a 1 r
0 10 20 30 40 50 60

manured Phosphorus Load (kg)

 

Figure 4.4.3 Measured phosphorus load and estimated from SEDMOD for delivery
ratio, and three soil erosion models.

 

 

 

5‘88
.J
l
l
l

 

 

 

 

 

e Onstad-Foster

5 / IAGNPS
‘f/ AUSLE
./

 

 

 

 

 

8888
i
1
l
l
1

3V.
1

 

 

 

Estimated Phosphorus Load (kg)
N
O

.s
O

 

 

 

O

 

20 30 40 50 60 70 80 90
Meanured Phosphorus Load (kg)

O
.s
C

Figure 4.4.4 Measured phosphorus load and estimated from Eq. [2.3.4] for delivery
ratio, and three soil erosion models.

99

856.63 can :25: 6. .aeaa 63.. .662 a BREE me... 2:9”.

.2. .9:

MN r F mN - _. I

F - .3 F - we I
0.0 r md 0.0 .. m6

no - no WM.

no - to - .

F o - o D

938a

 

 

100

Table 4.4.1

Monitored phosphorus load and its estimation based on sediment

phosphorus content, spatially distributed SDR model (SEDMOD) and four soil erosion

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

models
Date Monitored SPLM Onstad—Foster AGNPS USLE
PhOSphorus (kg) (kg) (kg) (kg)
Load
(kg)
3/27/91 3.18 0.822 5.492 8.454 9.201
4/4/91 0 0.019 0.188 0.295 0.320
4/9/91 1.42 0.713 6.31 1 10.089 10.981
4/15/91 0.49 0.284 1.835 2.726 2.967
4/19/91 1.09 0.467 1.750 2.147 2.346
4/23/91 0.199 0.198 1.831 2.877 3.144
4/24/92 10.1 4.674 9.064 10.889 1 1.946
7/14/92 0.22 0.589 6.450 10.426 1 1.493
4/19/93 28 34.448 29.492 25.528 28.262
5/4/93 0.06 0.687 7.837 12.725 14.099
6/7/93 0.58 2.593 23.853 37.397 41.418
6/14/93 0.05 0.709 8.162 13.301 14.733
6/19/93 12.75 7.914 33.346 50.113 55.447
4/12/94 0.22 0.333 1 . 106 1.443 1.575
4/30/94 0.15 0.028 0.265 0.416 0.452
7/20/94 0.24 0.222 2.583 4.284 4.681
4/26/95 0.22 0.090 0.724 1.071 1.176
7/4/95 0.05 0.709 8.185 13.476 14.806
5/10/96 2.4 0.986 6.830 10.631 1 1.557
5/21/96 1 0.617 5.334 8.552 9.316
6/18/96 21.9 24.380 26.073 32.134 34.926
4/5/97 0.46 0.344 0.700 0.589 0.642
5/5/97 0.15 0.064 0.610 0.973 1.052
5/19/97 0.27 0.162 1.240 1.966 2.102
6/23/97 0.19 0.1 17 1.146 1.876 2.006
RT 0.94 0.62 0.38 0.38
Standard Error 2.12 6.09 10.34 11.45

 

 

 

 

 

101

 

Table 4.4.2
phosphorus content, spatially lumped SDR model (equation [2.3.4]) and four soil erosion

Monitored phosphorus load and its estimation based on sediment

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

models.
Date Monitored Modified Onstad- AGNPS USLE
Phosphorus RUSLE Foster (kg) (kg)
Load (kg) (kg)
(kg)
3/27/1991 3.13 1.268 7.690 13.111 14.206
4/4/1991 0 0.028 0.265 0.457 0.493
4/9/1991 1-42 1.101 8.842 15.653 16.961
4/15/1991 0.49 0.440 2.574 4.230 4.582
4/19/1991 1-09 0.706 2.403 3.274 3.551
4/23/1991 0199 0.299 2.514 4.388 4.759
4/24/1992 10.1 6.475 1 1.389 15.282 16.547
7/ 14/ 1992 0-22 0.839 8.328 15.017 16.362
4/19/1993 28 48.231 37.453 36.283 39.574
5/4/1993 0-06 0.963 9.935 19.204 19.706
6/7/1993 058 3.643 30.384 53.313 58.183
6/14/1993 005 0.997 10.398 18.963 20.692
6/19/1993 12.75 11.162 42.665 71.715 78.211
4“ 211994 0.22 0.474 1 .430 2.077 2.245
4/30/1994 0.15 0.041 0.361 0.627 0.679
7/20/1994 0.24 0.322 3.385 6.253 6.766
4/26/ 1995 0.22 0.120 0.884 1.460 1.586
7/4/1995 0-05 0.980 10.255 18.818 20.451
5/10/1996 2.4 1.385 8.695 15.025 16.223
5/21/ 1996 1 0.860 6.729 1 1.990 12.962
6/ 1 8/ 1996 21.9 33.417 32.414 44.355 47.872
4/5/1997 0-46 0.446 0.826 0.773 0.837
5/5/1997 015 0.102 0.888 1.575 1.688
5/19/1997 027 0.284 1.966 3.438 3.673
6/23/1997 0- 19 0.201 1.795 3.244 3.464
R2 0.94 0.62 0.37 0.37
Standard Error 2.95 7.76 14.75 16.07

 

 

 

 

 

 

102

 

4.5 Assessing Annual Sediment and Phosphorus Loading

The Revised Universal Soil Loss Equation (RUSLE) was used to predict average
annual soil loss. The modified RUSLE (MRUSLE) was used to estimate soil loss for
single events. Sediment and Phosphorus Loading Model (SPLM) is developed based on
SEDMOD and MRUSLE, which is a single event model for estimating sediment and
phosphorus loads. When the original RUSLE was used, in place of the modified RUSLE,
SPLM can be used to estimate annual sediment and phosphorus loads. Unfortunately,
there were no data available in the study watershed to verify sediment and phosphorus
loading on an annual basis. SPLM for annual sediment and phosphorus loading
assessment needs to be tested in the future in other watersheds where annual data are

available or collected.

103

CHAPTER FIVE

CONCLUSIONS

Although great progress has been made in controlling point source pollution in the
past three decades, nonpoint source pollution, particularly from agriculture, is considered
the leading cause for water pollution in the United States. Sediment and phosphorus are
two major pollutants that are responsible for water pollution. Soil erosion degrades soil
productivity and causes water quality problems through sediments and nutrients.
Excessive fertilization, particularly from phosphorus, causes surface water deterioration.
Efforts have been made to minimize agricultural nonpoint source pollution such as
implementing the best management practices. Controlling agricultural nonpoint source
pollution requires the information and knowledge on identifying the source areas and
quantifying the loading. Water monitoring program is helpful but costly. Science based
modeling provides an alternative to provide a quantitative estimation on soil erosion,
sediment and nutrient loadings, and helps identify the source areas.

The goal of this research is to investigate various agricultural nonpoint source
pollution modeling options, and develop a GIS based and spatially distributed approach
for estimating soil erosion, sediment and phosphorus loading in agricultural watershed
context.

Spatially lumped and temporally lumped models that are often used in soil erosion
and sediment delivery generally can not provide good site-specific and event-specific

estimation for the varying conditions. Most agricultural NPS models generally require

104

extensive data inputs which are often not readily available in many cases. This
dissertation has investigated and evaluated various commonly used models with ground
truth water monitoring data. A spatially explicit sediment delivery model (SEDMOD) has
been integrated with the modified Revised Universal Soil Loss Equation (MRUSLE).
This integrated model led to the development of Sediment and Phosphorus Loading
Model - SPLM which is capable of calculating soil erosion, sediment delivery ratios,
sediment yield and phosphorus loading. Because it is also integrated with Geographic
Information System, it has automated spatial data processing and provided spatial
representation of NPS pollution in the watershed context. From the output maps of the
model, high risk areas can be identified to help prioritize and/or initialize the best
management practice implementation.

In this study watershed, the spatially explicit sediment delivery model not only
provides information for spatial analysis, bUt also improves the accuracy of modeling
compared to the spatially lumped model. The new model shows less systemic errors than
spatially lumped models which most often overestimated the sediment. The integrated
model SPLM requires only a few data layers (i.e. soils, DEM, streams, surface
roughness) which are readily available or can be easily created.

For the soil erosion sub-model used in SPLM, both rainfall and runoff effects on
soil erosion have been taken into account in energy term. The accuracy of the modified
Revised Universal Soil Loss Equation (MRUSLE) model is improved compared to the
original equation which primarily considers the rainfall energy. When the energy term is

modified, the Revised Universal Soil Loss Equation is capable of estimating soil erosion

105

for single storm events although the model was initially designed for predicting long term
average annual soil loss.

Phosphorus loading is calculated based on sediment yield and phosphorus-
sediment ratio in SPLM. The water quality monitoring data showed that sediment
attached phosphorus is the major form of phosphorus that is transported to stream system.
Phosphorus loading is found to be directly related to sediment yield. In this watershed,
total phosphorus is highly correlated with sediment yield (R2 = 0.99). This suggests that
management practices that are used to control sediment may control phosphorus loading
as well.

Among the combined uses of several sediment delivery ratio models and soil
erosion models, the integrated model with SEMOD and the modified RUSLE (i.e.
SPLM) has been proven to provide the most accurate estimation on sediment yield and
phosphorus loading in the study watershed. SPLM has a user-friendly menu driven
interface and requires a minimum input data set.

In summary, the development of Sediment and Phosphorus Loading Model
(SPLM) based the spatially explicit sediment delivery model and the modified RUSLE,
provides an easy-to—use tool for agricultural nonpoint source pollution assessment.
Compared to other models used in the study watershed, SPIM estimates soil erosion,
sediment delivery ratio, and sediment and phosphorus loading on a watershed basis with
an improved accuracy. It has demonstrated the benefits of using spatially explicit model
combined with GIS technology. This allows users to identify the source areas and
estimate NPS loadings which may lead to a cost-effective watershed management

planning for minimizing agricultural nonpoint source pollution.

106

 

5.1 Recommendations for Future Study

Although SPLM was used to estimate soil erosion, sediment load, and phosphorus
load for a single storm, it could be used to estimate annual average soil erosion, sediment
load, and phosphorus by simply using Revised Universal Soil Loss Equation (RUSLE) to
replace the modified RUSLE. Because RULSE can predict annual average soil erosion,
when used with SEDMOD, it will provide an estimation of annual average sediment load
and phosphorus load. This has yet to be tested in other watersheds where annual data are
available.

Like spatially explicit sediment delivery model (SEDMOD), Sediment and
Phosphorus Loading Model (SPLM) is a newly developed model and has not been widely
tested on other watersheds. Further calibrations and verifications may be needed for its

application in other watersheds with different conditions.

107

 

APPENDICES

108

Appendix A Estimates of Manning’s n for overland flow and soil covers based on

n=0.01 for smooth bare soil (Engman, 1985)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Treatment Manning’s n
Comstalk residue applied to 1 ton/acre 0.020
fallow surface 2 tons/acre 0.040
4 tons/acre 0.070
Comstalk residue disk- 1 ton/acre 0.012
harrow incorporated 2 tons/acre 0.020
4 tons/acre 0.023
Wheat straw mulch 0.25 ton/acre 0.015
0.5 ton/acre 0.018
1 ton/acre 0.032
2 tons/acre 0.070
4 tons/acre 0.074
Grass Sparse 0.015
Poor 0.023
Fair 0.032
Good 0.046
Excellent 0.074
Dense 0.150
Very dense 0.400
Small grain (20% to full maturity)
Across slope Poor stand 0.018
Moderate stand 0.023
Good stand 0.032
Dense 0.046
Upslope and downslope Poor stand 0.012
Moderate stand 0.015
Good stand 0.023
Dense 0.032

 

 

 

109

 

Appendx B The program code to calculate soil erosion and sediment yield based on
results from SEDMOD and Modified RULSE

/****************************************************************

/* File name: erosion_sediment.aml *
/* *
/* Purpose: *
/* This program to calculate the soil erosion and *
/* sediment yield based on the sediment delivery ratio *
/* computed from SEDMOD and the modified Revised Soil *
/* Loss Equation. *
/* *
/* Usage: under Arc/Info prompt, *
/* Arc: &run erosion_sediment.aml input.data output.file *
/* *
/* Written by Da Ouyang *
/* Date: August 2000 *
/* *

*

/***************************************************************

&args input_file output_file

&if [null %input_file%] or [null %output_fi1e%] &then
&return &warning Specify the file names. — Usage: PORG.aml
IN_fi1e OUT_file

/******‘k‘k*‘k'k'k'k'k************************'k'k************************

/* open the input file to read - outside of the loop - open one
/*

/****************************************************************

&s my_in := [open %input_file% openstat -r]

&if %openstat% ne 0 &then
&return &error The input data file can not open to read

/****************************************************************

/* read the records from the input file, skip the first two lines
/* outside of the loop
/*

/**'k*************************************************************

&s linel
&s line2

[read %my_in% readstat]
[read %my_in% readstat]

/* open the file to write records - outside of the loop
/* open once
&s my_out := [open %output_file% openstat —w]

&if %openstat% ne 0 &then

110

&return &error The file can not open to write

/* open the file to read the monitoring data for comparsion
&s monitoring = marshall.data
&s my_data := [open %monitoring% openstat —r]
&if %openstat% ne 0 &then
&return &error The monitoring data file can not open to read

&s count = l

/* check if it reaches the end of file, which returns 102
&do &while %readstat% ne 102

/* After read 25 storms, there is a sperator
&if %count% eq 26 &then &do

&s terms := [read %my_in% readstat]

&s count = 0
&end

&if %count% ne 0 &then &do

&do I := 1 &t0 4
&s record%I% := [read %my_in% readstat]
&end

/* To save the records, the file has to be closed before return
&if %readstat% eq 102 &then &do

&s writestat = [write %my;out% [quote %terms%]]
/* &s writestat = [write %my_out% ------------------ ]
&s closestat := [close -all]
/* &system perl out.cgi
&return &inform It has reached the END OF FILE - DONE
&end

/* read the monitoring data
&s measured_data := [read %my_data% readstat]

/* To read again, the file has to be closed then open again
to read
&if %readstat% eq 102 &then &do

&s writestat = [write %my_out% [quote %terms%]]
/* as writestat = [write %my_out% ----------------- ]

&s closestat := [close %my_data%]

&s my_data := [open %monitoring% openstat -r]

&s measured_data := [read %my_data% readstat]
&end

111

 

 

/********************‘k'k*************************************

/* Erosion calculation for marsahll watershed

/*

/* &if [exists soil_loss -GRID] &then kill soil_loss
/* &sv wshed = [response 'Enter watershed name '1
/*

/**********************************************************‘k

&sv wshed = Marshall

/* &sv yearl - [response ’Enter the year YY ']

/* &sv monthl [response ’Enter the month MM ’]
/* &sv datel [response ’Enter the date DD ’ 1
/* &sv ei [response 'Enter the El value ’]

&sv yearl = %record1%
&sv monthl = %record2%
&sv datel = %record3%
&sv ei = %record4%
&type
&type ---> Erosion calculation for %wshed% watershed
%year1%%monthl%%date1% <---
&type
&if %date1% < 16 &then
&sv date2 = 1
&if %date1% > 15 &then
&sv date2 = 2

/*****************************~k*******************************

/* define the data file path
/*

/*************************************************************

&sv ls = /home/mydata/1s/%wshed%_ls

&sv sdr = /home/mydata/sdr/%wshed%/sdr%year1%/del_ratio
&sv k = /home/mydata/k/k%month1%%date2%

&sv c = /home/mydata/c/c%year1%%month1%%date2%

&if %:program% <> GRID &then grid

&messages &on

/*********************'k***************************************

/* soil_loss = E * K * C * LS * P ( in tons / acre / storm)
/*

/* where E = energy term

/*

/*******'k*****************************************************

112

/*************************************************************
/*

/* define a name for different storms

/*

/**********************‘k******************************‘k*******

&sv erosion = er%year1%%monthl%%datel%
&sv sediment sy%yearl%%month1%%datel%
&sv erosion2 = e%year1%%month1%%datel%
&sv sedimentZ = s%yearl%%month1%%date1%
&sv sum_er = te%year1%%monthl%%date1%
&sv sumbsy = ts%year1%%month1%%date1%

 

&if [exists %erosion% -GRID] &then kill %erosion%
%erosion% = %ei% * %k% * %c% * %ls% * 1

describe %erosion%

/*********************************************************

/* convert short ton to kg, convert square meters to acre
/*

/*********************************************‘k'k'k'k'k'k'k'k‘k'k'k'k

&if [exists %erosion2% -GRID] &then kill %erosion2%
%erosion2% = 907 * %erosion% * %grd$dx% * %grd$dy% / 4047

DOCELL

%sum_er% += %erosion2%

 

END
&s resultl = [show %sum_er%]

&type The total soil loss is %result1% Kg from this storm event

/**************************************************************

/* write a record in the output file
/* single quote is needed if the record contains a space
/*

/**************************************************************

&s sperator = <-—-%year1%%month1%%datel%—-->

&s writestat [write %my_out% [quote %sperator%]]

&s writestat [write %my_out% [quote %result1%]]

113

/**************************************************************

/* calculate sediment yield based on soil erosion

/*

/* sediment delivery ratio needs to be divided by 100
/*

/*********************************'k****************************

&if [exists %sediment% -GRID] &then kill %sediment%
%sediment% = %erosion% * (float(%sdr%) / 100)

describe %sediment%
/* converted a ton to kilograms, converted square memters to acre
&if [exists %sediment2% -GRID] &then kill %sediment2%
%sediment2% = 907 * %sediment% * %grd$dx% * %grd$dy% / 4047
DOCELL

%sum_sy% += %sediment2%
END

&s result2 = [show %sum_sy%]

&type The total sediment yield is %result2% Kg from this storm
event

/* Write a record in the output file
&s delta = %result2% - %measured_data%
&if %measured_data% eq 0 &then
&s delta2 = NA
&e1se
&s delta2 = %delta% / %measured_data%

&s writestat = [write %my_out% [quote %result2%, %measured_data%,
%delta%, %delta2%]]

/* delete variables
&dv %sumpsy% %sumher% %resultl% %resu1t2% %delta% %delta2%

kill %erosion% all

kill %erosion2% all

kill %sediment% all

kill %sediment2% all

&end /* end of &if %count% eq 0 &then &do

&s count = %count% + 1

114

 

 

/********************************************************

/* End the loop of reading input file
/*

/*******************************************************at

&end /* end of &do &while

/* close the output file
&s closestat := [close -all]

&return

115

Appendix C Monitoring data from grab samples from Marshall Drain watershed

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DATE TIME Flow COD O-P T-P Residue
(018) (mg/1) (mg/1) (mg/1) (mg/1)

5/2/1990 15 :45 0.07 8 0.03 4
5/9/1990 1 1:25 0.02 8 0.029 5
5/14/1990 14:48 0.02 7 0.023 4
6/5/1990 15 :33 0.2 7 0.032 4
6/19/1990 1 1:55 0.07 6 0.04 4
1/30/1991 13:23 5 0.004 0.053 9

/29/1991 10:30 0.3 13 0.011 0.041 8
Afl/1992 8: 15 0.4 10 0.002 0.014 4
4/27/1992 14:00 1 17 0.066 7
4/30/1992 13: 10 0.6 16 0.01 0.025 4
5/11/1992 9:00 0.3 5 0.012 0.017 4
5/24/1992 8: 15 0.2 10 0.013 0.03 4
7/13/1992 13:55 0.07 5 0.022 0.046 10
04/26/93 13:00 0.39 12 0.060 21
05/03/93 1 1 :40 0.41 6 0.012 0.028 4
05/10/93 08:20 0.3 1 1 0.01 1 0.021 4
05/17/93 08:45 0.18 7 0.010 0.023 4
05/24/93 13:40 0.15 10 0.019 0.041 5
06/02/93 13:50 0. 15 6 0.005 0.030 4
06/14/93 12:57 0.45 16 0.012 0.075 6
06/21/93 14:00 3.8 22 0.112 4
06/28/93 09:48 0.32 14 0.037 0.074 1 1
07/06/93 12:20 0.28 1 1 0.028 0.042 7
07/13/93 1 1 :35 0.32 12 0.020 0.036 4
07/19/93 16: 10 0.26 9 0.200 0.032 4
07/29/93 13:30 0.25 6 0.019 0.033 4
04/08/94 12:00 6 0.008 0.016 4
04/14/94 12:00 0.5 18 0.012 0.032 4
04/19/94 12:00 0.05 12 0.01 0.03 4
04/26/94 12:00 0.26 13 0.006 0.029 4
05/03/94 12:00 0.45 5 0.01 0.023 4
05/10/94 12:00 0.34 20 0.01 0.018 4
05/17/94 12:00 0.23 8 0.06 0.02 4

 

116

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

05/24/94 12:00 0.2 1 1 0.014 0.05 5
06/01/94 12:00 0.16 6 0.024 0.045 6
06/07/94 12:00 0.13 1 1 0.032 0.058 8
06/14/94 12:00 0.12 1 1 0.036 0.064 9
06/21/94 12:00 0.07 5 0.48 0.074 8
/28/94 12:00 0.36 15 0.035 0.062 4
07/05/94 12:00 0.24 28 0.001 0.105 6
07/12/94 12:00 0.12 10 0.027 0.041 6
07/19/94 12:00 0.1 1 7 0.027 0.039 7
07/26/94 12:00 0.23 1 1 0.03 0.041 9
08/02/94 12:00 7 0.027 0.049 4
08/09/94 12:00 1 1 0.022 0.038 5
08/16/94 12:00 23 0.077 0.105 4
08/26/94 12:00 1 2 0.027 0.039 4
08/30/94 1 2:00 8 0.02 0.025 5
04/04/95 12:00 8 0.007 0.032 4
04/12/95 09:57 22 0.029 0.086 4
04/17/95 16: 10 8 0.009 0.022 5
04/26/95 10:05 0.4 9 0.006 0.016 4
05/01/95 08:26 0.35 13 0.005 0.014 4
05/08/95 12:30 0.3 1 8 0.006 0.019 4
05/18/95 15:39 0.23 8 0.01 1 0.032 4
05/24/95 09:50 0.36 29 0.079 0.16 29
06/01/95 1 1:45 0.18 1 1 0.027 0.048 6
06/08/95 09:25 0.16 1 1 0.027 0.054 6
/15/95 10:00 0.12 5 0.021 0.052 6
06/21/95 10:00 0.1 1 24 0.03 0.46 7
06/28/95 09:52 0.13 10 0.032 0.127 49
07/06/95 12:30 0.28 1 1 0.026 0.047 4
07/13/95 09:30 0.1 8 15 0.028 0.077 24
07/20/95 09:30 0.22 5 0.033 0.072 8
07/27/95 09:20 0.28 14 0.035 0.06 4
08/1 1/95 09:00 0.038 0.063 6
08/18/95 1 1:00 0.046 0.07 4
08/28/95 09: 15 5 0.035 0.07 6
4/4/ 1996 8:50 0.45 18 0.01 1 0.021 4
4/9/1996 9:16 0.34 10 0.01 0.026 4
4/18/1996 12:00 0.38 8 0.053 0.041 4
4/23/1996 9:30 0.41 5 0.01 1 0.028 4
4/30/1996 8:41 0.38 24 0.032 0.08 13

 

117

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/7/1996 9:45 0.38 10 0.01 1 0.021 4
5/14/1996 1 1:30 0.49 26 0.023 0.065 4
5/22/1996 10:27 1.02 21 0.016 0.061 13
5/30/1996 9:30 0.45 13 0.013 0.056 4
6/4/1996 9: 20 0.38 6 0.013 0.026 4
6/1 1/1996 10:30 0.41 10 0.01 1 0.035 4
6/20/1996 21 :07 1.48 23 0.074 0. 123 8
6/25/ 1996 1 1 :00 0.72 12 0.028 0.046 4
7/2/1996 1 1 :00 0.49 15 0.016 0.047 4
7/10/1996 10:00 0.41 12 0.017 0.034 4
7/16/1996 10:00 0.41 10 0.016 0.036 5
7/24/1996 1 l :20 0.30 10 0.019 0.042 7
7/31/1996 7:50 0.38 15 0.02 0.043 1 1
8/6/1996 13:00 0.41 6 0.023 0.038 15
8/13/1996 10:30 0.30 5 0.012 0.037 8
4/1/1997 10:50 0.85 9 0.011 0.017 4
4/7/1997 13:06 0.80 18 0.006 0.018 4
4/13/1997 12:00 0.60 16 0.004 0.015 4
4/22/1997 13:50 0.45 5 0.007 0.015 4
4/29/1997 12:00 0.35 5 0.004 0.02 4
5/13/1997 12:51 0.48 16 0.002 0.021 7
5/28/1997 9:20 0.25 21 0.006 0.092 6
6/4/1997 9:20 0.23 18 0.02 0.039 4
6/10/1997 9: 1 1 0.10 24 0.008 0.1 5
6/17/1997 8:40 0.20 17 0.013 0.029 4
6/30/1997 9:21 0.30 15 0.016 0.049 12
7/8/1997 14:45 0.30 19 0.029 0.045 7
7/15/1997 15:00 0.30 12 0.021 0.038 4
7/22/1997 10:30 0.28 14 0.025 0.045 4
7/29/1997 10:30 0.20 7 0.029 0.043 10
8/5/1997 9:44 0.23 8 0.023 0.037 7
8/1 1/1997 1 1:07 0.18 10 0.024 0.046 4

 

 

118

 

.4. _

Appendix D Calculated soil-loss ratio (SLR) from Revised Universal Soil Loss

Equation

RUSLE Version: SWCSl.05pre_d

surface cover

1/2 crop: c

---Date -------- Field Operation ----- Res. Add. (#lA)-
4/20/2 cult;prim-swp 6-12 N

5/5/2 planter;db1.dsk.op N

5/15/2 anhyd applic; disk N

6/5/2 cult;row-mu1t sweepN

10/15/2 harvest 7000

- Inputs for C-Factor -

Input File: 1CC

corn grain following corn grain

spring cult.
row cult. 1x
125 bu. yield
zone 102b

City code: 22002 GRAND RAPIDS MI
adjust for soil moisture depletion: NO
% surface covered by rock fragments: 0

function; B-value code:

orn;125bu 90day 30'

(1) normal conditions

senescence code: NO

------- New Growth Set--—---

2/2 crop: corn;125bu 90day 30' senescence code: NO
-—-Date -------- Field Operation ----- Res. Add. (#lA) -------- New Growth Set ------
4/20/3 cult;prim-swp 6-12 N
5/5/3 planter;dbl.dsk.op N
5/15/3 anhyd applic; disk N
6/5/3 cult;row-mult sweepN
10/15/3 harvest 7000
— Results By Crops -
C
crop start date end date %EI factor
corn;125bu 90day 30' 4/20/2 4/20/3 100.0 0 08
corn;125bu 90day 30' 4/20/3 4/20/4 100.0 0 08

------------------------------------------- Rotation C Factor = 0.08 -----

crop # 1/2: c

operation

- Results By Operations -

orn;125bu 90day 30'

previous crop:

cult;prim-swp 6-12 N
p1anter;dbl.dsk.op N
anhyd applic; disk N
cult;row-mu1t sweepN
harvest

corn;125bu 90day 30'

% res. cover op. date
after op. date next op. SLR %EI
53 4/20/2 5/5/2 0.054 3 3
44 5/5/2 5/15/2 0.072 2 7
28 5/15/2 6/5/2 0 123 9.7
19 6/5/2 10/15/2 0.088 69.4
93 10/15/2 4/20/3 0.02 15.0
= 0.08 ------- Crop C Factor = 0.08 ---------

———————————— Rotation C Factor

119

 

 

crop # 2/2: corn;125bu 90day 30' previous crop: corn;125bu 90day 30'
% res. cover op. date
operation after op. date next op. SLR %EI
cult;prim-swp 6-12 N 53 4/20/3 5/5/3 0.054 3.3
planter;db1.dsk.op N 44 5/5/3 5/15/3 0.072 2.7
anhyd applic; disk N 28 5/15/3 6/5/3 0.123 9.7
cult;row-mult sweepN 19 6/5/3 10/15/3 0.088 69.4
harvest 93 10/15/3 4/20/4 0.02 15.0
------------ Rotation C Factor = 0.08 -———--- Crop C Factor = 0.08 --—------
- Results By lS-Day Period - Page 1
% res.
cover plu * cc * sc * sr * sm = SLR %EI
4/20/2 corn;125bu 90day 30' cult;prim-swp 6-12 N
4/20 - 4/30/2 51 0.358 1 0.189 0.774 1 0.052 2.2
5/1 - 5/4/2 50 0.374 1 0.198 0.788 1 0.058 1.1
5/5/2 corn;125bu 90day 30' planter;dbl.dsk.op N
5/5 - 5/14/2 43 0.371 0.983 0.241 0.815 1 0.072 2.7
5/15/2 corn;125bu 90day 30' anhyd applic; disk N
5/15/2 28 0.341 0.967 0.404 0.819 1 0.109 0.3
5/16 - 5/31/2 26 0 359 0.938 0.428 0.832 1 0.12 7.0
6/1 - 6/4/2 24 0.375 0.908 0.458 0.847 1 0.132 2.4
6/5/2 corn;125bu 90day 30' cult;row-mult sweepN
6/5 - 6/15/2 18 0.362 0.761 0.562 0.791 1 0.122 6.6
6/16 - 6/30/2 16 0.353 0.51 0.601 0.819 1 0.088 9.0
7/1 - 7/15/2 13 0.328 0.35 0.642 0.845 1 0.062 9.0
7/16 - 7/31/2 12 0.329 0.267 0.679 0.867 1 0.052 10.0
8/1 - 8/15/2 10 0.353 0.267 0.714 0.887 1 0.06 10.0
8/16 - 8/31/2 9 0.377 0.297 0.749 0.905 1 0.076 9.0
9/1 - 9/15/2 7 0.399 0.399 0.783 0.92 1 0.115 7.0
9/16 - 9/30/2 6 0.414 0.487 0.807 0.933 1 0.152 6.0
10/1 - 10/14/2 6 0.424 0.487 0.824 0.942 1 0.16 2.8
10/15/2 corn;125bu 90day 30' harvest
10/15/2 93 0.431 1 0.041 0.945 1 0.017 0.2
10/16 - 10/31/2 93 0.44 1 0.042 0.948 1 0.017 2.0
11/1 - 11/15/2 91 0.456 1 0.043 0.954 1 0.019 2.0
11/16 - 11/30/2 91 0.465 1 0.044 0.959 1 0.02 1.0
12/1 - 12/15/2 90 0.468 1 0.045 0.963 1 0.02 1.0
12/16 - 12/31/2 90 0.468 1 0.045 0.967 1 0.02 0.0
1/1 - 1/15/3 90 0.465 1 0.045 0.969 1 0.02 0.0
1/16 - 1/31/3 90 0.463 1 0.045 0.971 1 0.02 1.0
2/1 - 2/15/3 90 0.46 1 0.045 0.972 1 0.02 1.0
2/16 - 2/28/3 90 0.458 1 0.045 0.974 1 0.02 1.0
3/1 - 3/15/3 90 0.457 1 0.045 0.975 1 0.02 1.0
3/16 - 3/31/3 89 0.462 1 0.046 0.978 1 0.021 2.0
4/1 - 4/15/3 88 0.471 1 0.048 0.98 1 0.022 2.0
4/16 - 4/19/3 87 0.477 1 0.049 0.981 1 0.023 0.8
4/20/3 corn;125bu 90day 30' cult;prim-swp 6-12 N
4/20 - 4/30/3 51 0.358 1 0.189 0.774 1 0.052 2.2
5/1 - 5/4/3 50 0.374 1 0.198 0.788 1 0.058 1.1
5/5/3 corn;125bu 90day 30' planter;db1.dsk.op N
5/5 - 5/14/3 43 0.371 0.983 0.241 0.815 1 0.072 2.7
5/15/3 corn;125bu 90day 30' anhyd applic; disk N
5/15/3 28 0.341 0.967 0.404 0.819 1 0.109 0.3
5/16 - 5/31/3 26 0.359 0.938 0.428 0.832 1 0.12 7.0
6/1 - 6/4/3 24 0.374 0.908 0.458 0.847 1 0.132 2.4
6/5/3 corn;125bu 90day 30' cult;row-mult sweepN
6/5 - 6/15/3 18 0.361 0.761 0.562 0.791 1 0.122 6.6
6/16 - 6/30/3 16 0.353 0.51 0.601 0.819 1 0.088 9.0
7/1 - 7/15/3 13 0.328 0.35 0.642 0.845 1 0.062 9.0
7/16 - 7/31/3 12 0.329 0.267 0.679 0.867 1 0.052 10.0
8/1 - 8/15/3 10 0.353 0.267 0.714 0.887 1 0.06 10.0
8/16 - 8/31/3 9 0.377 0.297 0.749 0.905 1 0.076 9.0
9/1 - 9/15/3 7 0.399 0.399 0.783 0.92 1 0.115 7.0
9/16 - 9/30/3 6 0.414 0.487 0.807 0.933 1 0.152 6.0

 

- Results By 15-Day Period - Page 2

% res.

cover plu * cc * sc * sr * sm = SLR %EI
10/15/3 corn;125bu 90day 30' harvest
10/15/3 93 0.431 1 0.041 0.945 1 0.017 0.2
10/16 - 10/31/3 93 0.44 1 0.042 0.948 1 0.017 2.0
11/1 - 11/15/3 91 0.456 1 0.043 0.954 1 0.019 2.0
11/16 - 11/30/3 91 0.465 1 0.044 0.959 1 0.02 1.0
12/1 - 12/15/3 90 0.468 1 0.045 0.963 1 0.02 1.0
12/16 - 12/31/3 90 0.467 1 0.045 0.967 1 0.02 0.0
1/1 — 1/15/4 90 0.465 1 0.045 0.969 1 0.02 0.0
1/16 - 1/31/4 90 0.463 1 0.045 0.971 1 0.02 1.0
2/1 - 2/15/4 90 0.46 1 0.045 0.972 1 0.02 1.0
2/16 - 2/28/4 90 0.458 1 0.045 0.974 1 0.02 1.0
3/1 - 3/15/4 90 0.457 1 0.045 0.975 1 0.02 1.0
3/16 - 3/31/4 89 0.462 1 0.046 0.978 1 0.021 2.0
4/1 - 4/15/4 88 0.471 1 0.048 0.98 1 0.022 2.0
4/16 - 4/19/4 87 0.476 1 0.049 0.981 1 0.023 0.8

 

 

 

121

RUSLE Version: SWCSl.05pre_d

Input File:

1CNB

corn -> Notill SOYbean
fall chisel, t
field cult 1x,

125 bu,

35 bu

zone 102b

- Inputs for C-Factor -

 

city code: 22002 GRAND RAPIDS MI
adjust for soil moisture depletion: NO
% surface covered by rock fragments: 0
surface cover function; B-value code: (1) normal conditions
1/2 crop: corn;125bu 90day 30' senescence code: NO
—--Date -------- Field Operation ----- Res. Add. (#/A) -------- New Growth Set ------
10/15/1 chisel;twisted pts.F
5/5/2 cult;secdry-sw6-12 F
5/5/2 p1anter;dbl.dsk.op F
5/15/2 anhyd applic; disk F
6/5/2 cult;row-mult sweepF
10/15/2 harvest 7000
2/2 crop: soybean; 30' 35bu mw senescence code: YES
---Date -------- Field Operation ----- Res. Add. (#/A) -------- New Growth Set ------
10/16/2 no operation
5/15/3 dri11;NT-f.res. fl F
10/10/3 harvest 2625
— Results By Crops -
C
crop start date end date %EI factor
corn;125bu 90day 30' 10/15/1 10/16/2 100.2 0.259
soybean; 30' 35bu mw 10/16/2 10/15/3 99.8 0.064
------------------------------------------- Rotation C Factor 0.162 -----
— Results By Operations -
crop # 1/2: corn;125bu 90day 30' previous crop: soybean; 30' 35bu mw
% res. cover op. date
operation after op. date next op. SLR %EI
chise1;twisted pts.F 27 10/15/1 5/5/2 0.261 18.3
cult;secdry—sw6—12 F 9 5/5/2 5/5/2 0 0.0
p1anter;dbl.dsk.op F 7 5/5/2 5/15/2 0.513 2.7
anhyd applic; disk F 3 5/15/2 6/5/2 0.564 9.7
cult;row-mult sweepF 1 6/5/2 10/15/2 0.206 69.4
harvest 93 10/15/2 10/16/2 0.021 0.2
------------ Rotation C Factor = 0.162 —-—---- Crop C Factor = 0.259 ---------
crop # 2/2: soybean; 30' 35bu mw previous crop: corn;125bu 90day 30'
% res. cover op. date
operation after op. date next op. SLR %EI
no operation 93 10/16/2 5/15/3 0.026 20.7
drill;NT-f.res. f1 F 41 5/15/3 10/10/3 0.075 78.1
harvest 73 10/10/3 10/15/3 0.048 1.0
------------ Rotation C Factor = 0.162 —-—-—-- Crop C Factor = 0 064 -———--—--

122

wisted

row cult 1x

 

- Results By 15-Day Period - Page 1

% res.

cover plu cc sc sr * sm SLR %EI
10/15/1 corn;125bu 90day 30' chisel;twisted pts.F
10/15/1 26 0.643 1 0.443 0.537 1 0.153 0.2
10/16 - 10/31/1 25 0.654 1 0.458 0.555 1 0.166 2.0
11/1 - 11/15/1 23 0.672 1 0.483 0.593 1 0.193 2.0
11/16 - 11/30/1 22 0.683 1 0.498 0.632 1 0.215 1.0
12/1 - 12/15/1 22 0.686 1 0.504 0.666 1 0.23 1.0
12/16 - 12/31/1 21 0.686 1 0.506 0.694 1 0.241 0.0
1/1 - 1/15/2 21 0.685 1 0.506 0.714 1 0.247 0.0
1/16 - 1/31/2 21 0.682 1 0.505 0.73 1 0.252 1.0
2/1 - 2/15/2 21 0.68 1 0.505 0.744 1 0.255 1.0
2/16 - 2/28/2 21 0.678 1 0.505 0.756 1 0.259 1.0
3/1 - 3/15/2 21 0.677 1 0.508 0.77 1 0.265 1.0
3/16 - 3/31/2 20 0.682 1 0.519 0.79 1 0.279 2.0
4/1 - 4/15/2 19 0.69 1 0.537 0.812 1 0.301 2.0
4/16 - 4/30/2 17 0.704 1 0.566 0.834 1 0.332 3.0
5/1 - 5/4/2 16 0.715 1 0.591 0.847 1 0.358 1.1
5/5/2 corn;125bu 90day 30' cult;secdry-sw6-12 F
5/5/2 corn;125bu 90day 30' planter;dbl.dsk.op F
5/5 - 5/14/2 7 0.739 0.983 0.801 0.882 1 0.513 2.7
5/15/2 corn;125bu 90day 30' anhyd applic; disk F
5/15/2 3 0.73 0.967 0.919 0.888 1 0.576 0.3
5/16 - 5/31/2 2 0.729 0.938 0.927 0.896 1 0.568 7.0
6/1 - 6/4/2 2 0.717 0.908 0.936 0.905 1 0.552 2.4
6/5/2 corn;125bu 90day 30' cult;row-mu1t sweepF
6/5 - 6/15/2 1 0.691 0.761 0.966 0.887 1 0.45 6.6
6/16 - 6/30/2 1 0.624 0.51 0.972 0.903 1 0.279 9.0
7/1 - 7/15/2 1 0.537 0.35 0.977 0.917 1 0.169 9.0
7/16 - 7/31/2 1 0.504 0.267 0.982 0.929 1 0.123 10.0
8/1 - 8/15/2 0 0.511 0.267 0.985 0.94 1 0.127 10.0
8/16 - 8/31/2 0 0.517 0.297 0.988 0.949 1 0.144 9.0
9/1 — 9/15/2 0 0.522 0.399 0.991 0.958 1 0.198 7.0
9/16 - 9/30/2 0 0.523 0.487 0.992 0.965 1 0.244 6.0
10/1 - 10/14/2 0 0.524 0.487 0.994 0.969 1 0.246 2.8
10/15/2 corn;125bu 90day 30' harvest
10/15/2 93 0.526 1 0.04 0.971 1 0.021 0.2
10/16/2 soybean; 30' 35bu mw no operation
10/16 - 10/31/2 92 0.534 1 0.041 0.972 1 0.021 2.0
11/1 - 11/15/2 91 0.548 1 0.043 0.975 1 0.023 2.0
11/16 - 11/30/2 90 0.555 1 0.044 0.978 1 0.024 1.0
12/1 - 12/15/2 90 0.557 1 0.044 0.98 1 0.024 1.0
12/16 — 12/31/2 89 0.556 1 0.045 0.982 1 0.024 0.0
1/1 - 1/15/3 89 0.553 1 0.045 0.983 1 0.024 0.0
1/16 - 1/31/3 89 0.55 1 0.045 0.984 1 0.024 1.0
2/1 - 2/15/3 89 0.547 1 0.045 0.985 1 0.024 1.0
2/16 - 2/28/3 89 0.544 1 0.045 0.985 1 0.024 1.0
3/1 - 3/15/3 89 0.542 1 0.045 0.986 1 0.024 1.0
3/16 - 3/31/3 88 0.546 1 0.046 0.987 1 0.025 2.0
4/1 - 4/15/3 87 0.552 1 0.048 0.989 1 0.026 2.0
4/16 - 4/30/3 85 0.565 1 0.051 0.99 1 0.029 3.0
5/1 - 5/14/3 82 0.581 1 0.056 0.991 1 0.032 3.7
5/15/3 soybean; 30' 35bu mw drill;NT-f.res. fl F
5/15/3 40 0.393 1 0.252 0.94 1 0.093 0 3

 

 

123

 

~ - Results By 15-Day Period — Page 2

% res

cover plu * cc * sc * sr * sm = SLR %EI
6/1 - 6/15/3 34 0.455 0.862 0.316 0.953 1 0.118 9.0
6/16 - 6/30/3 29 0.491 0.692 0.364 0.96 1 0.119 9.0
7/1 — 7/15/3 26 0.513 0.466 0.413 0.966 1 0.095 9.0
7/16 - 7/31/3 22 0.504 0.205 0.46 0.971 1 0.046 10.0
8/1 - 8/15/3 21 0.515 0.144 0.483 0.975 1 0.035 10.0
8/16 - 8/31/3 31 0.541 0.215 0.341 0.979 1 0.039 9.0
9/1 — 9/15/3 48 0.564 0.486 0.191 0.982 1 0.051 7.0
9/16 - 9/30/3 49 0.579 0.67 0.184 0.985 1 0.07 6.0
10/1 — 10/9/3 44 0.586 0.703 0.215 0.986 1 0.087 1.8
10/10/3 soybean; 30' 35bu mw harvest
10/10 — 10/14/3 72 0.595 1 0.081 0.987 1 0.048 1.0

 

 

 

124

RUSLE Version: SWCSl.05pre_d Input File: lNBNWH

SOYBEANS AFTER WHEAT
NO TILL SB AND WHEAT

35 bu soys, 50 bu wheat
zone 102b

— Inputs for C—Factor -

city code: 22002 GRAND RAPIDS MI
adjust for soil moisture depletion: NO
% surface covered by rock fragments: 0
surface cover function; B-value code: (1) normal conditions

1/2 crop: soybean; 30' 35bu mw senescence code: YES

 

 

---Date -------- Field Operation ----- Res. Add. (#lA) -------- New Growth Set ------
7/16/0 no operation

5/15/1 plantr;NT-fluted c N

10/10/1 harvest 2625

2/2 crop: wheat; winter 50bu senescence code: NO
—--Date -------- Field Operation ----- Res. Add. (#/A) -------- New Growth Set ------
10/15/1 drill;NT-f.res. fl F

7/15/2 harvest 5100

- Results By Crops -

C
crop start date end date %EI factor
soybean; 30' 35bu mw 7/16/0 10/15/1 144.8 0.017
wheat; winter 50bu 10/15/1 7/16/2 55.2 0.026

------------------------------------------- Rotation C Factor = 0.019 -----

- Results By Operations -

crop # 1/2: soybean; 30' 35bu mw previous crop: wheat; winter 50bu
% res. cover op. date
operation after op. date next op. SLR %EI
no operation 96 7/16/0 5/15/1 0.013 65.7
p1antr;NT-fluted c N 68 5/15/1 10/10/1 0.02 78.1
harvest 82 10/10/1 10/15/1 0.02 1.0
------------ Rotation C Factor = 0.019 ------- Crop C Factor = 0.017 -—-----——
crop # 2/2: wheat; winter 50bu previous crop: soybean; 30' 35bu mw
% res. cover op. date
operation after op. date next op. SLR %EI
dri11;NT-f.res. fl F 47 10/15/1 7/15/2 0.027 54.6
harvest 96 7/15/2 7/16/2 0.008 0.6
------------ Rotation C Factor = 0.019 —---—-- Crop C Factor = 0.026 -—-------

125

 

- Results By 15-Day Period - Page 1

 

 

% res.

cover plu * cc * sc * sr * sm = SLR %EI
7/16/0 soybean; 30' 35bu mw no operation
7/16 - 7/31/0 95 0.252 1 0.036 0.992 1 0 009 10 0
8/1 - 8/15/0 94 0.269 1 0.037 0.993 1 0 01 10 0
8/16 - 8/31/0 93 0.288 1 0.039 0.993 1 0 011 9 0
9/1 — 9/15/0 91 0.306 1 0.042 0.994 1 0.013 7 0
9/16 - 9/30/0 89 0.319 1 0.044 0.995 1 0 014 6 0
10/1 - 10/15/0 88 0.328 1 0.047 0.995 1 0 015 3 0
10/16 - 10/31/0 86 0.335 1 0.049 0.995 1 0.016 2 0
11/1 - 11/15/0 86 0.338 1 0.05 0.996 1 0.017 2 0
11/16 - 11/30/0 85 0.339 1 0.051 0.996 1 0 017 1 0
12/1 - 12/15/0 85 0.338 1 0.051 0.996 1 0 017 1 0
12/16 - 12/31/0 85 0.336 1 0.052 0.996 1 0 017 0 0
1/1 - 1/15/1 85 0.333 1 0.052 0.996 1 0 017 0 0
1/16 - 1/31/1 85 0.331 1 0.052 0.996 1 0 017 1 0
2/1 - 2/15/1 85 0.328 1 0.052 0.996 1 0 017 1 0
2/16 - 2/28/1 85 0.326 1 0.052 0.996 1 0.017 1 0
3/1 - 3/15/1 84 0.324 1 0.052 0.997 1 0 017 1 0
3/16 - 3/31/1 84 0.324 1 0.053 0.997 1 0 017 2 0
4/1 — 4/15/1 83 0.326 1 0.054 0.997 1 0 018 2 0
4/16 - 4/30/1 82 0.33 1 0.056 0.997 1 0 019 3 0
5/1 - 5/14/1 81 0.336 1 0.059 0.997 1 0 02 3 7
5/15/1 soybean; 30' 35bu mw plantr;NT-f1uted c N
5/15/1 68 0.334 1 0.094 0.985 1 0.031 0 3
5/16 — 5/31/1 67 0.339 0.97 0.099 0.986 1 0 032 7 0
6/1 — 6/15/1 64 0.347 0.862 0.109 0.988 1 0 032 9 0
6/16 - 6/30/1 61 0.353 0.692 0.121 0.99 1 0.029 9.0
7/1 - 7/15/1 58 0.35 0.466 0.134 0.991 1 0.022 9.0
7/16 — 7/31/1 55 0.329 0.205 0.147 0.992 1 0.01 10.0
8/1 — 8/15/1 53 0.326 0.144 0.157 0.993 1 0.007 10 0
8/16 - 8/31/1 58 0.335 0.215 0.132 0.994 1 0.009 9 0
9/1 - 9/15/1 67 0.343 0.486 0.096 0.994 1 0.016 7 0
9/16 - 9/30/1 67 0.346 0.67 0.096 0.995 1 0.022 6 0
10/1 - 10/9/1 64 0.347 0.703 0.108 0.995 1 0.026 1 8
10/10/1 soybean; 30' 35bu mw harvest
10/10 - 10/14/1 82 0.352 1 0.057 0.995 1 0.02 1 0
10/15/1 wheat; winter 50bu dri11;NT—f.res. fl F
10/15/1 . 47 0 339 1 0.199 0.943 1 0.063 0 2
10/16 - 10/31/1 46 0 344 0.97 0.208 0.946 1 0 066 2 0
11/1 - 11/15/1 44 0 345 0.862 0.224 0.952 1 0 063 2 0
11/16 - 11/30/1 42 0.323 0.716 0.235 0.957 1 0.052 1 0
12/1 - 12/15/1 42 0.309 0.657 0.239 0.962 1 0.047 1 0
12/16 - 12/31/1 41 0 308 0.657 0.242 0.965 1 0 047 0 0
1/1 - 1/15/2 41 0 306 0.657 0.242 0.968 1 0 047 0 0
1/16 - 1/31/2 41 0.304 0.657 0.242 0.97 1 0 047 1 0
2/1 - 2/15/2 41 0.301 0.657 0.242 0.971 1 0 047 1 0
2/16 - 2/28/2 41 0.3 0.657 0.243 0.973 1 0 047 1 0
3/1 — 3/15/2 41 0.299 0.657 0.246 0.975 1 0 047 1 0
3/16 — 3/31/2 40 0 302 0.657 0.254 0.977 1 0 049 2 0
4/1 - 4/15/2 38 0 307 0.632 0.267 0.979 1 0 051 2 0
4/16 - 4/30/2 36 0 309 0.48 0.288 0.982 1 0 042 3 0
5/1 - 5/15/2 33 0 287 0.249 0.318 0.984 1 0 022 4 0
5/16 - 5/31/2 30 0.247 0.151 0.358 0.985 1 0 013 7 0

126

- Results By 15-Day Period - Page 2

% res.

cover plu * cc * sc * sr * sm = SLR %EI
6/16 - 6/30/2 23 0.229 0.144 0.456 0.989 1 0.015 9.0
7/1 - 7/14/2 20 0.238 0.144 0.5 0.99 1 0.017 8.4
7/15/2 wheat; winter 50bu harvest
7/15/2 96 0.243 1 0.035 0.991 1 0.008 0 6

 

127

RUSLE Version: SWCSl.05pre_d Input File: lBC

Soybean -> Corn

- Inputs for C-Factor -

city code: 22002
adjust for soil moisture depletion: NO
% surface covered by rock fragments: 0
surface cover function; B-value code: (1) normal conditions

GRAND RAPIDS MI

1/2 crop: soybean; 30' 35bu mw senescence code: NO
---Date -------- Field Operation ----- Res. Add. (#/A) -------- New Growth Set ------
11/1/1 chisel;twisted pts.N
5/15/2 cult;secdry-sw6-l2 N
5/15/2 p1anter;db1.dsk.op N
6/15/2 cult;row-mult sweepN
10/10/2 harvest 2625
2/2 crop: corn;125bu 90day 30' senescence code: NO
---Date -------- Field Operation ----- Res. Add. (#/A) -------- New Growth Set ------
10/15/2 chisel;twisted pts.F
5/5/3 cult;secdry-sw6-l2 F
5/5/3 p1anter;db1.dsk.op F
5/15/3 anhyd applic; disk F
6/5/3 cult;row-mult sweepF
10/15/3 harvest 7000
- Results By Crops -
C
crop start date end date %EI factor
soybean; 30' 35bu mw 11/1/1 10/15/2 97.8 0.355
corn;125bu 90day 30' 10/15/2 11/1/4 202.2 0.184
------------------------------------------- Rotation C Factor = 0.24 -—---

- Results By Operations -

crop # 1/2: soybean; 30' 35bu mw previous crOp: corn;125bu 90day 30'
% res. cover op. date
operation after op. date next op. SLR %EI
chisel;twisted pts.N 17 11/1/1 5/15/2 0.447 18.7
cult;secdry—sw6-12 N 9 5/15/2 5/15/2 0 0.0
p1anter;db1.dsk.op N 8 5/15/2 6/15/2 0 587 15.7
cult;row-mu1t sweepN 5 6/15/2 10/10/2 0.274 62.4
harvest 79 10/10/2 10/15/2 0.05 1.0
------------ Rotation C Factor = 0.24 -—--—-- Crop C Factor = 0.355 ---—-—---
crop # 2/2: corn;125bu 90day 30' previous crop: soybean; 30' 35bu mw
% res. cover op. date
operation after op. date next op. SLR %EI
chisel;twisted pts.F 27 10/15/2 5/5/3 0.284 18.3
cult;secdry-sw6-12 F 9 5/5/3 5/5/3 0 0.0
p1anter;db1.dsk.op F 7 5/5/3 5/15/3 0.552 2.7
anhyd applic; disk F 2 5/15/3 6/5/3 0.603 9.7
cult;row-mult sweepF 1 6/5/3 10/15/3 0.215 69.4
harvest 93 10/15/3 11/1/4 0.096 102.2
———————————— Rotation C Factor = 0.24 ————-—- Crop C Factor = 0 184 ------—-—

128

— Results By 15-Day Period - Page 1

% res.
cover plu * cc * sc * sr * sm = SLR %EI

11/1/1 soybean; 30' 35bu mw chisel;twisted pts.N

11/1 - 11/15/1 17 0.863 1 0.583 0.729 1 0 367 2.0
11/16 - 11/30/1 16 0.865 1 0.593 0.757 1 0 388 1.0
12/1 - 12/15/1 16 0.865 1 0.596 0.782 1 0 403 1.0
12/16 - 12/31/1 16 0.864 1 0.597 0.802 1 0 413 0.0
1/1 - 1/15/2 16 0.861 1 0.596 0.816 1 0 419 0.0
1/16 - 1/31/2 16 0.859 1 0.596 0.827 1 0 423 1.0
2/1 - 2/15/2 16 0.856 1 0.596 0.836 1 0 426 1.0
2/16 - 2/28/2 16 0.853 1 0.596 0.844 1 0 429 1.0
3/1 - 3/15/2 16 0.85 1 0.597 0.853 1 0 433 1.0
3/16 - 3/31/2 15 0.849 1 0.604 0.867 1 0 444 2.0
4/1 - 4/15/2 15 0.848 1 0.615 0.881 1 0 46 2.0
4/16 - 4/30/2 14 0.849 1 0.633 0.895 1 0 481 3.0
5/1 - 5/14/2 12 0.851 1 0.657 0.908 1 O 507 3.7
5/15/2 soybean; 30' 35bu mw cult;secdry-sw6-12 N

5/15/2 soybean; 30' 35bu mw p1anter;db1.dsk.op N

5/15/2 8 0.875 1 0.762 0.921 1 0.614 0.3
5/16 - 5/31/2 7 0.875 0.97 0.777 0.927 1 0.611 7.0
6/1 - 6/14/2 6 0.871 0.862 0.803 0.937 1 0.565 8.4
6/15/2 soybean; 30' 35bu mw cult;row-mult sweepN

6/15/2 5 0.864 0.791 0.851 0.918 1 0.534 0.6
6/16 - 6/30/2 4 0.859 0.692 0.862 0.925 1 0 474 9.0
7/1 - 7/15/2 4 0.835 0.466 0.88 0.936 1 0 321 9.0
7/16 - 7/31/2 3 0.776 0.205 0.896 0.946 1 0 135 10.0
8/1 - 8/15/2 3 0.756 0.144 0.909 0.954 1 0 094 10.0
8/16 - 8/31/2 2 0.758 0.215 0.923 0.961 1 0 144 9.0
9/1 - 9/15/2 2 0.759 0.486 0.934 0.967 1 0 333 7.0
9/16 - 9/30/2 2 0.759 0.67 0.943 0.973 1 0 466 6.0
10/1 - 10/9/2 2 0.757 0.703 0.947 0.976 1 0 492 1.8
10/10/2 soybean; 30' 35bu mw harvest

10/10 - 10/14/2 78 0.762 1 0.068 0.977 1 0.05 1.0
10/15/2 corn;125bu 90day 30' chisel;twisted pts.F

10/15/2 27 0.698 1 0.434 0.564 1 0 171 0.2
10/16 - 10/31/2 26 0.708 1 0.449 0.583 1 0 185 2.0
11/1 - 11/15/2 24 0.726 1 0.476 0.619 1 0 214 2.0
11/16 - 11/30/2 22 0.735 1 0.492 0.656 1 0 237 1.0
12/1 - 12/15/2 22 0.738 1 0.498 0.689 1 0 253 1.0
12/16 - 12/31/2 22 0.738 1 0.5 0.715 1 0 264 0.0
1/1 - 1/15/3 22 0.736 1 0.5 0.734 1 0 27 0.0
1/16 - 1/31/3 22 0.734 1 0.5 0.749 1 0 275 1.0
2/1 - 2/15/3 22 0.731 1 0.499 0.762 1 0 278 1.0
2/16 - 2/28/3 21 0.729 1 0.5 0.773 1 0 282 1.0
3/1 - 3/15/3 21 0.727 1 0.503 0.786 1 0 288 1.0
3/16 - 3/31/3 20 0.731 1 0.514 0.805 1 0 303 2.0
4/1 - 4/15/3 19 0.739 1 0.534 0.826 1 0 326 2.0
4/16 - 4/30/3 17 0.751 1 0.564 0.846 1 0 358 3.0
5/1 - 5/4/3 16 0.76 1 0.591 0.858 1 0 385 1.1
5/5/3 corn;125bu 90day 30' cult;secdry-sw6-12 F

5/5/3 corn;125bu 90day 30' p1anter;db1.dsk.op F

5/5 - 5/14/3 7 0.784 0.983 0.803 0.893 1 0.552 7
5/15/3 corn;125bu 90day 30' anhyd applic; disk F

5/15/3 2 0.774 0.967 0.92 0.898 1 0.618 0.3

 

 

129

- Results By 15-Day Period - Page 2

% res

cover plu * cc * sc * sr * sm = SLR %EI
6/1 - 6/4/3 2 0.754 0.908 0.938 0.914 1 0.587 2.4
6/5/3 corn;125bu 90day 30' cult;row-mult sweepF
6/5 - 6/15/3 1 0.724 0.761 0.967 0.898 1 0.479 6.6
6/16 - 6/30/3 1 0.651 0.51 0.973 0.912 1 0 294 9.0
7/1 - 7/15/3 1 0.557 0.35 0.979 0.925 1 0.177 9.0
7/16 - 7/31/3 0 0.521 0.267 0.983 0.936 1 0.128 10.0
8/1 - 8/15/3 0 0.526 0.267 0.987 0.946 1 0.131 10.0
8/16 - 8/31/3 0 0.53 0.297 0.99 0.954 1 0.148 9.0
9/1 - 9/15/3 0 0.532 0.399 0.992 0.962 1 0.203 7.0
9/16 - 9/30/3 0 0.533 0.487 0.994 0.968 1 0.25 6.0
10/1 - 10/14/3 0 0.532 0.487 0.995 0.972 1 0.251 2.8
10/15/3 corn;125bu 90day 30' harvest
10/15/3 93 0.534 1 0.04 0.973 1 0.021 0.2
10/16 - 10/31/3 92 0.542 1 0.041 0.975 1 0.022 2.0
11/1 - 11/15/3 91 0.556 1 0.043 0.977 1 0.023 2.0
11/16 - 11/30/3 90 0.563 1 0.044 0.98 1 0.024 1.0
12/1 - 12/15/3 90 0.564 1 0.044 0.982 1 0.025 1.0
12/16 - 12/31/3 89 0.563 1 0.045 0.983 1 0.025 0.0
1/1 - 1/15/4 89 0.56 1 0.045 0.984 1 0.025 0.0
1/16 - 1/31/4 89 0.557 1 0.045 0.985 1 0.025 1.0
2/1 - 2/15/4 89 0.554 1 0.045 0.986 1 0.024 1.0
2/16 - 2/28/4 89 0.551 1 0.045 0.987 1 0.024 1.0
3/1 - 3/15/4 89 0.55 1 0.045 0.987 1 0.024 1.0
3/16 - 3/31/4 88 0.553 1 0.046 0.988 1 0.025 2.0
4/1 - 4/15/4 87 0.559 1 0.048 0.989 1 0.027 2.0
4/16 - 4/30/4 85 0.571 1 0.051 0.99 1 0.029 3.0
5/1 — 5/15/4 82 0.587 1 0.056 0.991 1 0.033 4.0
5/16 - 5/31/4 78 0.608 1 0.065 0.992 1 0.039 7.0
6/1 — 6/15/4 73 0.631 1 0.078 0.993 1 0.049 9.0
6/16 - 6/30/4 67 0.652 1 0.096 0.994 1 0.062 9.0
7/1 - 7/15/4 61 0.67 1 0.118 0.994 1 0.078 9.0
7/16 - 7/31/4 56 0.684 1 0.143 0.995 1 0.097 10.0
8/1 - 8/15/4 50 0.696 1 0.173 0.995 1 0.12 10.0
8/16 - 8/31/4 45 0.708 1 0.21 0.996 1 0.148 9.0
9/1 - 9/15/4 39 0.718 1 0.255 0.996 1 0 183 7.0
9/16 - 9/30/4 35 0.724 1 0.295 0.996 1 0 213 6.0
10/1 - 10/15/4 32 0.727 1 0.329 0.996 1 0 238 3.0
10/16 - 10/31/4 30 0.727 1 0.356 0.997 1 0 258 2.0

 

 

 

130

Appendix E Modeling outputs for the calculated soil erosion and sediment load from
SEDMOD and four soil erosion models

Data format:

<---yymmdd--->

soil erosion (kg)

sediment yield (kg), monitored sediment yeild (kg), A, A%

 

Modeling outputs from the modified USLE used in AGNPS

<--910327--->

1 1439.05567648

3961 .045283 163, 1280, 2681 .045283 163, 2.094566627471
<---910404-->

397 .78379361 15

137.6865029824, l, 136.6865029824, 136.6865029824
<---910409--->

13657.24359272

4727.236602181, 562, 4165.236602181, 7.411453028792
<---910415--->

3690.549639578

1277.424776283, 144, 1133.424776283, 7.871005390854
<—--910419--->

2856.456249893

1005.610212734, 292, 713.610212734, 2.443870591555
<---910423--->

3828.866885747

1347.945605986, 43.4, 1304.545605986, 30.05865451581
<---920424--->

13333.07687242

5102.232105448, 3940, l 162.232105448, 0.2949827678802
<---9207 l4—-->

13103.34612345

4885.011000413, 50, 4835.011000413, 96.70022000826
<---930419--->

31657.03399741

11961 .15303369, 13073, -11 11.84696631, 00850491062732
<---930504--->

15756.00579978

5962.04516176, 16, 5946.04516176, 371.62782261
<---930607--->

46516.29578424

17521.8477963, 138, 17383.8477963, 125.9699] 15674
<---930614-—->

16545.32598142

6232.325240302, 12, 6220325240302, 518.3604366918

131

<--930619--->

6257 1.43449568

23480.01069774, 5654, 17826.01069774, 3.152814060442
<---940412--->

1810.949202128

675.6269852903, 35, 640.6269852903, 18.30362815115
<--940430--->

548.2294178137

194.8990468269, 27, 167.8990468269, 6.218483215811
<--940720--->

5456.434383847

2007.442610518, 114, 1893.442610518, 16.6091457063
<--950426--->

1273.156196689

501.5537998959, 53.5, 448.0537998959, 8.374837381232
<--950704--->

16418.76396908

6314.389190491, 4.9, 6309.489190491, 1287.650855202
<--960510--->

131 10.29895763

4980.984184057, 647, 4333.984184057, 6.698584519408
<--960521-->

10461.81561377

4006.994025422, 179, 3827.994025422, 21.3854414828
<---96061 8--->

38700.51541314

15056.45406947, 10846, 4210.45406947, 0.3882033993611
<---970405--->

6751987597818

2759753927593, 50.5, 2254753927593, 4.46485926256
<--970505--->

1373.720587761

457.5880910959, 34.4, 423. 1880910959, 1230197939232
<--970519-->

2999.589094702

920.7470092525, 51.3, 869.4470092525, 1694828478075
<-—-970623--->

2830.47597479

878.5299220054, 32.3, 846.2299220054, 2619906879274

132

Modeling outputs from USLE which use EI substitutes R factor in the equation to
calculate the soil erosion for the single events

<---910327--->

12394.45481 192

4311.051069357, 1280, 3031.051069357, 2368008647935
<---910404-->

431.0204419598

149.8585100469, 1, 148.8585100469, 148.8585100469
<---910409--->

14798.36839036

5145142188963, 562, 4583.142188963, 8.155057275735
<---910415-->

3998.91 1853768

1390.353941733, 144, 1246353941733, 8.655235706479
<---910419-->

3097509980671

1098.773039628, 292, 806773039628, 2762921368589
<---910423-—->

415198145326

147282341607, 43.4, 1429.42341607, 32.93602341175
<--920424--->

14438.34597615

5597.346348065, 3940, 1657 .346348065, 0.4206462812348
<---9207 14—-->

1427596995005

5384.937040796, 50, 5334.937040796, 1066987408159
<--930419-->

34528.457 16662

13242.87461312, 13073, 169.87461312, 0.0129943098845
<---930504--->

17193.61909081

6605785895208, 16, 6589.785895208, 411.8616184505
<---930607--->

5075648625545

1940627336638, 138, 19268.27336638, 139.6251693216
<---930614--->

1805351431794

6902.594333566, 12, 6890.594333566, 574.2161944638
<---930619---> .

68239.70719787

2597853284957, 5654, 2032453284957, 3594717518495
<---9404 12--->

1958.655496504

738.1618954916, 35, 703.1618954916, 20.09033987119
<---940430--->

133

5921368510351

212.2461024822, 27, 185.2461024822, 68609667586
<---940720--->

5902507023853

2193.174006692, 114, 2079.174006692, 1823836847975
<--950426—-->

138283492977

551.2059321604, 53.5, 497.7059321604, 9302914619821
<--950704-->

17844.42345586

6936.717129256, 4.9, 6931.817129256, 1414.656556991
<---960510-->

1415432294194

5415044079757, 647, 4768044079757, 7.369465347383
<--960521-->

1 1309.2822362

4365.156190516, 179, 4186.156190516, 2338634743305
<--960618--->

4176957425904

163645521291, 10846, 55185521291, 05088098957311
<-—970405-->

730.2719118173

301 .3180090148, 50.5, 2508180090148, 4966693247818
<---970505--->

1472827916507

4928410132475, 34.4, 458.4410132475, 1332677364092
<--9705 19--->

3204. 167073403

984.8523571166, 51.3, 933.5523571166, 1819790169818
<--970623--->

3023.296773304

940.161430407, 32.3, 907861430407, 2810716502808

134

 

Modeling outputs from the modified USLE by Onstad and Foster (1974)

<---910327--->

6709.69988835

2573.069686516, 1280, 1293069686516, 1.010210692591
<---910404-->

230.6158326291

88.40266888092, 1, 87 .40266888092, 87.40266888092
<--910409--->

7715218012783

2957.497989202, 562, 2395497989202, 4.262451938082
<---910415--->

2244733540664

860.4805431 196, 144, 716.4805431196, 4975559327219
<--910419-->

2097.21123236

820.2196503933, 292, 5282196503933, 1808971405457
<--910423-—->

2192570728816

8575147573833, 43.4, 814.1147573833, 18758404548
<--920424--->

9936214459173

4246962303721, 3940, 306.962303721, 0.0779092141424
<---920714-->

7266360630277

3021931841241, 50, 2971.931841241, 59.43863682482
<--930419--->

3267733351274

13817.97546976, 13073, 744.97546976, 0.0569858081359
<---930504»->

866870586123 '
3672.012710608, 16, 3656012710608, 228500794413
<---930607--->

26511.41708352

11175.7389252, 138, 11037.7389252, 799836154
<---930614--->

9071568015766

3824.068547789, 12, 3812068547789, 317.6723789824
<---930619--->

37224.74264214

15624.34716169, 5654, 997034716169, 1.763414779216
<---940412--->

- 1246702628833

5180231354514, 35, 4830231354514, 138006610129
<---940430--->

314.7841627771

135

124.4008143085, 27, 97.4008143085, 3.607437566981
<---940720—-->

2953.180097887

1209815717702, 114, 1095815717702, 9.612418576333
<--950426--->

771.2201390432

3389335932869, 53.5, 2854335932869, 5.335207351157
<---950704--->

8947.212832221

3834.704874395, 4.9, 3829804874395, 7815928315092
<--960510-->

7586557489669

3200.002818362, 647, 2553.002818362, 3945908529153
<--960521--—>

5872273248573

2498.985365716, 179, 2319.985365716, 1296081209897
<--960618-->

2828129836904

12216.19993011, 10846, 1370. 1999301 1, 0.1263322819574
<--970405--->

7209320858499

3279650456622, 50.5, 277.4650456622, 5494357339846
<--970505--->

7753251074916

286.0430468115, 34.4, 251.6430468115, 7315204849172
<--970519-->

1715277021559

581.2767604599, 51.3, 5299767604599, 1033093100312
<--970623--->

1565875513989

5368730398674, 32.3, 5045730398674, 1562145634264

136

Modeling outputs from the modified RULSE

<--910327-->

1105981497918

384.6835361282, 1280, -895.3164638718, 06994659873998
<-~~910404--->

2455685068063

8538001213879, 1, 7538001213879, 7538001213879
<---910409--->

959.730942924

333.6822004465, 562, 2283177995535, 04062594298105
<--910415--->

383.4376978423

133.3147951838, 144, -10.6852048162, 00742028112236
<--910419--->

6161973782154

2185823666417, 292, -73.4l76333583, 02514302512271
<--910423--->

2608551558487

92.53258633428, 43.4, 4913258633428, 1.132087242725
<--920424--->

5650. 18729969

2190.420935845, 3940, -1749579064155, 04440556000393
<---9207 14-->

731.1081103026

275.7760880646, 50, 225.7760880646, 4515521761292
<--930419-->

4208 1 .6289687 3

16139.78087313, 13073, 306678087313, 0.2345889140312
<-—-930504--->

8387372333714

322.2427199467, 16, 306.2427199467, 19.14016999667
<---930607--->

317830559278

1215195768535, 138, 1077.195768535, 7805766438659
<--—930614--->

8687224534947

3321479989383, 12, 320.1479989383, 26.67899991153
<---930619--->

9739554836428

3707 .802302276, 5654, - l 946. 197 697724, 034421607671 1
<---940412--->

41 295429705

155630802694, 35, 120630802694, 3.446594362686
<---940430--->

3610882937206

137

 

12.94288359402, 27, -14.05711640598, 05206339409622
<---940720--->

2808633249569

1043594086209, 1 14, 06405913791, 00845665910447
<---950426—-->

104.5279387141

41 .66543574203, 53.5, -11.83456425797, 02212068085602
<---950704--->

8539362219818

331 .9532296641 , 4.9, 3270532296641 , 6674555707431
<---9605 10--->

1208675579283

4624051314673, 647, -1845948685327, 02853089158156
<---960521--->

7495007385397

2892922575565, 179, 1102922575565, 06161578634441
<---96061 8--->

2915603666817

11422.799712, 10846, 576.799712, 00531808696294
<---970405--->

3893946767272

160.6684118481, 50.5, 110.1684118481, 2181552709863
<---970505--->

8880017782228

29.71451663846, 34.4, 468548336154, 01362059116727
<---9705 19--->

247046273248

7593365087837, 51.3, 24.63365087837, 0480188126284
<—--970623--->

1753758050177

54.5370103492, 32.3, 22.2370103492, 06884523327926

138

 

BIBLIOGRAPHY

139

BIBLIOGRAPHY

Ahuja, L. R., A. N. Sharpley, and O. R. Lehman. 1982. Effect of Soil Slope and Rainfall
Characteristics on Phosphorus in Runoff. Journal of Environmental Quality. 11(1):9-12.

American Society of Civil Engineers (ASCE). 1975. Sedimentation Engineering, Chapter
4: Sediment Sources and Sediment Yields. p. 437-493. American Society of Civil
Engineers. New York, NY.

Arnold, J. G., J. R. Williams, R. Srinivasan, and K. W. King. 1996. SWAT - Soil and
Water Assessment Tool. SWAT User’s Manual. Texas A & M University. Temple, TX.

Arnold, J. G., J. R. Williams, A. D. Nicks, and N. B. Sammons. 1990. SWRRB - A Basin
Scale Simulation Model for Soil and Water Resources Management. Texas A & M
University Press, College Station.

Arnold, J .G., M.D.Birket, J.R.Williams, W.F.Smith, and H.N.McGill. 1987. Modeling
the effects of urbanization on basin water yield and reservoir sedimentation. Water
Resources Bulletin. 23(6):]101-1 107.

Beven, K. J ., and M. J. Kirkby. 1979. A Physically Based, Variable Contributing Area
Model of Basin Hydrology. Hydrological Sciences Bulletin. 24(1):43-69.

Bingner, R. L, and F. D. Theurer. 2001. Topographic Factors for RUSLE in the
Continuous-Simulation Watershed Model for Predicting Agricultural, Non-Point Source
Pollution (AnnAGNPS). In: Soil Erosion Research for the 21St Century - Proceedings of
the International Symposium. Pp. 619-622. January 3-5, 2001. Honolulu, HI.

Bone, S. W., R. Christman, L. M. Feusner, R. L. Goettemoller, and B. H. Noltre. 1979.
Ohio Erosion Control and Sediment Pollution Abatement Guide for Agricultural Land.
The Ohio State University Extension Bulletin 594.

Borah, D. K. 1989. Runoff Simulation Model for Small Watersheds. Transactions of the
ASAE. 32(2):881-886.

Borah, D. K., and M. S. Ashraf. 1990. Modeling Storm Runoff and Sediment With
Seasonal Variations. Transactions in Agriculture. 33(1): 105-1 10.

Brannan, K., and J. M. Harnlett. 1998. Using Geostatistics to Select Grid-Cell Layouts for
the AGNPS Model. Transactions of the ASAE. 41(4): 101 1-1018.

140

Brenner, F. J. and J. J. Mondok. 1995. Nonpoint Source Pollution Potential in an
Agricultural Watershed in Northwestern Pennsylvania. Water Resources Bulletin.
31(6): 1 101-1 105.

Brown, L. C., and G. R. Foster. 1987. Storm erosivity using idealized intensity
distributions. Transactions of ASAE. 30:379-386.

Clark, K. 1990. A Biological Investigation of Sycamore Creek and Tributaries Ingham
County, Michigan. Michigan Department of Natural Resources, Lansing, Michigan.

Delwiche D. L. L., and D. A. Haith. 1983. Loading Functions for Predicting Nutrient
Losses from Complex Watersheds. Water Resources Bulletin. 19(6):951-959.

Dingman, S. L. 1993. Chapter 9: Stream Response to Water-Input Events. Physical
Hydrology. Macmillan Publishing Company.

Duda, A. M., and R. A. Johnson. 1985. Cost-effective Targeting of Agricultural Nonpoint
Source Pollution Control. Journal of Soil and Water Conservation. 40: 108-111.

Edwards, C. L, R. D. Shannon, and A. R. Jerrett. 1999. Sedimentation Basin Retention
Efficiencies for Sediment, Nitrogen, and Phosphorus from Simulated Agricultural Runoff.
Transactions of the ASAE. 42(2):403-409.

Edwards, D. R., and T. C. Daniel. 1994. A Comparison of Runoff Quality Effects of
Organic and Inorganic Fertilizers Applied to Fescuegrass Plots. Water Resource Bulletin.
30(1):3541.

Engman, E. T. 1986. Roughness Coefficients for Routing Surface Runoff. Journal of
Irrigation and Drainage Engineering. 112(1):39-53.

Flanagan, D. C., and G. R. Foster. 1989. Storm Pattern Effect on Nitrogen and
Phosphorus Losses in Surface Runoff. Transactions of the ASAE. 32(2):535-544.

Fogel, M.M., L.H. Hekman, Jr., and L.Duckstein. 1976. A stochastic sediment yield
model using the modified universal soil loss equation. In: Erosion and solid matter
transport in inland waters. pp 226-233. IAHS-AISH publication No. 122.

Foster, G. R. 2000. RUSLE for single events. Personal Communications.

Foster, G. R., G. A. Weesies, K. G. Renard, D. C. Yoder, D. K. McCool, and J. P. Porter.
1997. Support Practice Factor (P). In: Predicting Soil Erosion by Water: A Guide to
Conservation Planning With the Revised Universal Soil Loss Equation (RUSLE). US.
Department of Agriculture Handbook No. 703.

Foster, G. R. 1986. Understanding Ephemeral Gully Erosion. In: Soil Conservation:

Assessing the National Resources Inventory. Volume 2. National Academy Press.
Washington, D. C.

141

 

Foster, G. R. 1982. Modeling the erosion process, ch. 8. In: C. T. Haan, H. D. Johnson,
and D. L. Brakensiek (eds.). Hydrologic Modeling of Small Watersheds. ASAE Monogr.
No. 5, American Society of Agricultural Engineering. St. Joseph, MI.

Foster, G. R., L. D. Meyer, and C. A. Onstad. 1977. A Runoff Erosivity Factor and
Variable Slope Length Exponents for Soil Loss Estimates. Transactions of the ASAE.
20:683-687.

Foster, G. R., L. D. Meyer, and C. A. Onstad. 1977. An Erosion Equation Derived from
Basic Erosion Principles. Transactions of the ASAE. 20:678-682.

Foster, G. R., and W. H. Wischmeier. 1974. Evaluating Irregular Slopes for Soil Loss
Prediction. Transactions of the ASAE. 17:305-309.

Foth, H. D., and B. G. Ellis. 1997. Soil Fertility. Second Edition. Lewis Publishers. New
York.

Fraser, R. H. 1999. SEDMOD: A GIS—Based Delivery Model for Diffuse Source
Pollutants. Ph.D. Dissertation. Yale University.

Fraser, R. H., P. K. Barten, and D. A. K. Pinney. 1998. Predicting Stream Pathogen
Loading from Livestock Using a Geographical Information System-Based Delivery
Model. Journal of Environmental Quality. 27:935-945.

Frere, M. H, J. D. Ross, and L. J. Lane. 1980. The nutrient submodel. In: Knisel, W. (ed.).
CREAMS, A Field Scale Model for Chemicals, Runoff, and Erosion From Agricultural
Management Systems. US. Department of Agriculture, Conservation Research Report 26,
Vol. 1, ch. 4, pp 65-86.

Gianessi, L. P., H.M.Peskin, P. Crosson, and C. Puffer, 1986. Nonpoint source pollution:
Are cropland controls the answer? Resources for the Future, Inc. Washington, DC.

Gilley, J. E., D. C. Kincaid, W. J. Elliot, and J. M. Laflen. 1992. Sediment Delivery on
Rill and Interrill Areas. Journal of Hydrology. 140(1992):313-341.

Green, I. R. A., and D. Stephenson. 1986. Criteria for Comparison of Single Event
Models. Hydrological Sciences Journal. 3 1(3):395-41 1.

Grigar, J ., and S. Davis. 1995. Water Erosion Prediction and Control. Technical Guide
Section I-C. State Office of Michigan, NRCS, USDA.

Haith, D. A., and L. J. Tubbs. 1981. Watershed Loading Functions for Nonpoint Sources.
Journal of the Environmental Engineering Division, ASCE, EE 1(107):121-137.

142

He, C., C. Shi, C. Yang, and B. P. Agosti. 2001. A Window-Based GIS-AGNPS Interface.
Journal of The American Water Research Association. 37(2): 1-12.

Hoover, K. A., M. G. Foley, P. G. Heasler, and E. W. Boyer. 1991. Sub-Grid-Scale
Characterization of Channel Lengths for Use in Catchment Modeling. Water Resources
Research. 27(1 1):2865-2873.

Inarndar, S. P., and T. A. Dillaha. 2000. Relationships Between Drainage Area, Slope
Length, and Slope Gradient for Riparian Slops in Virginia. Transactions of the ASAE.
43(4):861-866.

Izuno, F. T., A. B. Bottcher, F. J. Coale, C. A. Sanchez, and D. B. Jones. Agricultural
BMPs for Phosphorus Reduction in South Florida. Transactions of the ASAE. 38(3):735-
744.

Jenq T. R., M. L. Granstrom, S. F. Hsueh, and C. G. Uchrin. 1984. A Phosphorus
Management LP Model Case Study. Water Resources Bulletin. 20(4):511-519.

Kang, Y. T. 1998. A Watershed Based Optimization Approach for Agricultural Non-
Point Source Pollution Management. Ph.D. Dissertation. Michigan State University.

Khanbilvardi, R. M., and A. S. Rogowski. 1984. Quantitative Evaluation of Sediment
Delivery Ratios. Water Resources Bulletin. December 1984. pp 865-874.

Kinnell, P. I. A. 2001. Runoff, the Universal Soil Loss Equation and Event Erosion. In:
Soil Erosion Research for the 218t Century - Proceedings of the International Symposium.
Pp. 99-102 January 3-5, 2001. Honolulu, HI.

Lee, G. F. 1971. Role of Phosphorus in Eutrophication and Diffuse Source Control.
Water Research Pergamon Press 1973, Vol. 7. pp. 111-128.

Lemunyon, J. L., and R. G. Gilbert. 1993. The Concept and Need for a Phosphorus
Assessment Tool. Journal of Production Agriculture. 6(4):483-486.

Licciardello, F., N. Rizzo, F. Somma, and S. M. Zimbone. 2001. Runoff and Soil Erosion
Modeling in a Small Sicilian Watershed. In: Soil Erosion Research for the 21“t Century —
Proceedings of the International Symposium. Pp. 298-301. January 3-5, 2001. Honolulu,
HI.

Liew, M. W. V. 1997. Prediction of Sediment Yield on a Large Watershed in North
Central China. Transactions of the ASAE. 41(3):599-604.

Lovejoy, S. B., J. G. Lee, T. O. Randhir, and B. A. Engel. 1997. Research Needs for

Water Quality Management in the 21St Century. Journal of Soil and Water Conservation.
January-February 1997: 18-21.

143

Manrique, L. A. 1988. Land Erodibility Assessment Methodology. University of Hawaii.
Editorial & Publishing Shop, Honolulu, HI.

Maidment, D. R. (ed.) 1992. Chapter 12: Erosion and Sediment Transport. Handbook of
Hydrology. McGraw-Hill, Inc.

McCool, D. K., G. R. Foster and G. A. Weesies. 1997. Slope Length and Steepness
Factor (LS). In: Predicting Soil Erosion by Water: A Guide to Conservation Planning
With the Revised Universal Soil Loss Equation (RUSLE). US. Department of
Agriculture Handbook No. 703.

McElroy, A. D., S.Y. Chiu, J. W. Nebgan, and others. 1976. Loading functions for
assessment of water pollution from nonpoint sources. Environmental Protection
Technical Service, EPA 600/2-76-151. Washington, DC.

Mironenko, E. V., and Y. A. Pachepsky. 1998. Estimating Transport of Chemicals from
Soil to Ponding Water. Journal of Hydrology. 208(1998):53-61.

Mitas, L., and H. Mitasova. 1998. Distributed Soil Erosion Simulation for Effective
Erosion Prevention. Water Resources Research. 34(3): 505-516.

Mitasova, H, J. Hofierka, M. Zlocha, and L. R. Iverson. 1996. Modelling Topographic
Potential for Erosion and Deposition Using GIS. International Journal of Geographic
Information Systems. 10(5): 629-641.

Moore, 1. D., and J. P. Wilson. 1992. Length-slope Factors for the Revised Universal Soil
Equation: Simplified Method of Estimation. Journal of Soil and Water Conservation.
47(5):423-428

Moore, I.D., and RB. Grayson. 1991. Terrain-Based Catchment Partitioning and Runoff
Prediction Using Vector Elevation Data. Water Resources Research. 27(6):]177-1191.

Moore, 1. D., R. B. Grayson, R. B., and A. R. Landson. 1991. Digital Terrain Modelling:
a Review of Hydrological, Geomorphological, and Biological Applications, Hydrological
Processes. 5: 3-30.

Moore, I.D., and G]. Burch. 1986a. Sediment Transport Capacity of Sheet and Rill Flow:
Application of Unit Stream Power Theory. Water Resources Research. 22(8): 1350-1360.

Moore, LD., and G]. Burch. 1986b. Modelling Erosion and Deposition: Topographic
Effects. Transactions of the ASAE. 29: 1624-1630.

Moore, LD., and G]. Burch. 1985. Physical Basis of the Length-slope Factor in the
Universal Soil Loss Equation. Soil Science Society of America Journal. 50: 1294-1298.

144

Munoz-Carpena, R., J. E. Parsons, and J. W. Gillian. 1999. Modeling Hydrology and
Sediment Transport in Vegetative Filter Strips. Journal of Hydrology. 214(1999):111-129.

Nearing, M. A., G. Govers, and L. D. Norton. 1999. Variability in Soil Erosion Data from
Replicated Plots. Soil Sci. Soc. Am. J. 63:1829-1835.

Needham, S. E. 1994. Assessing A Geographic Information System and ANN-AGNPS
Water Quality MODEL: A Case Study. Ph.D. Dissertation. Michigan State University.

Nelson, D. W., and T. J. Logan. 1983. Chemical Processes and Transport of Phosphorus.
In: Schaller, F. W., and G. W. Bailey (eds). Agricultural Management and Water Quality.
Pp 65-91.

Newson, M. 1994. Hydrology and the River Environment. Clarendon Press. Oxford.

Ng, H. Y. F., T. Mayer, and J. Marsalek. 1993. Phosphorus Transport in Runoff from a
Small Agricultural Watershed. Wat. Sci. Tech. 28(3-5):451-460.

Novak, L. T., D. C. Adriano, G. A. Columan, and D. B. Shah. 1975. Phosphorus
Movement in Soils: Theoretical Aspects. Journal of Environmental Quality. 4(1):93-98.

Novotny, V., and G. Chesters. 1989. Delivery of Sediment and Pollutants from Nonpoint
Sources: A Water Quality Perspective. Journal of Soil and Water Conservation.
44(6):568-576.

Onstad, C. A., and G. R. Foster. 1974. Erosion Modeling on a Watershed. Transactions of
the ASAE. 18:288-292.

Ouyang, D., J. Bartholic. 2001. Web-based Application for Soil Erosion Prediction. In:
Soil Erosion Research for the 21St Century — Proceedings of the International Symposium.
Pp 260-263. January 3-5, 2001. Honolulu, HI.

Ouyang, D., and J. Bartholic. 1997. Predicting sediment delivery ratio in Saginaw Bay
Watershed. In: The 22nd National Association of Environmental Professionals
Conference Proceedings. pp 659-671. May 19-23, 1997. Orlando, FL.

Ouyang, D., Y.-T. Kang, and J. Bartholic. 1996. Agricultural phosphorus assessment in
the Great Lakes Basin: A case study. In: An Agricultural Profile of the Great Lakes
Basin: Characteristics and Trends in Production, Land-use and Environmental Impacts.
Great Lakes Commission. East Lansing, MI.

Perrone, J ., and C. A. Madramootoo. 1999. Sediment Yield Prediction using AGNPS.
Journal of Soil and Water Conservation. 54(1):415-419.

Petrie, J ., and P. Diplas. 2000. Statistical Approach to Sediment Sampling Accuracy.
Water Resources Research. 36(2):597-605.

145

 

 

Pierzynski, G. M., and T. J. Logan. 1993. Crop, Soil, and Management Effects on
Phosphorus Soil Test Levels. Journal of Production Agriculture. 6(4):513-520.

Rapp, J. F., V. L. Lopes, K. G. Renard. 2001. Comparing Soil Erosion Estimates from
RUSLE and USLE on Natural Runoff Plots. In: Soil Erosion Research for the 21“
Century — Proceedings of the International Symposium. Pp. 24-27. January 3-5, 2001.
Honolulu, HI.

Rast W., and G. F. Lee. 1983. Nutrient loading Estimates for Lakes. Journal of
Environmental Engineering. 109(2):502-517.

Renard, K. G., G. R. Foster, G. A. Weesies, D. K. McCool, and D. C. Yoder,
coordinators. 1997. Predicting Soil Erosion by Water: A Guide to Conservation Planning
With the Revised Universal Soil Loss Equation (RUSLE). US. Department of
Agriculture Handbook No. 703.

Renard, K. G., and G. R. Foster. 1983. Soil Conservation: Principles of Erosion by Water.
Dryland Agriculture. Agronomy Monograph N o. 23. American Society of Agronomy.
Madison, WI.

Richards, R. P., D. B. Baker, J. W. Kramer, and D. E. Ewing. 1996. Annual Loads of
Herbicides in Lake Erie Tributaries in Ohio and Michigan. Journal of Great Lakes
Research Vol. 22: 414-428.

Richards, R. P., and J. Holloway. 1987. Monte Carlo Strategies for Estimating Tributary
Loads. Water Resources Research. 23(10): 1939-1948.

Risse, L. M., M. A. Nearing, A. D. Nicks, and J. M. Laflen. 1993. Error Assessment in
the Universal Soil Loss Equation. Soil Sci. Soc. Am. J. 57:825-833.

Roehl, J. W., 1962. Sediment Source Areas, Delivery Ratios and Influencing
Morphological Factors. International Association of Scientific Hyrology Commission of
Land Erosion. Publication 59: 202-203.

Rudra, R. P., W. T. Dickinson, H. R. Whiteley. 2001. Lessons Learned from the Use of
Erosion and Non-Point Source Pollution Model. In: Soil Erosion Research for the 21St
Century — Proceedings of the International Symposium. Pp. 629-631. January 3-5, 2001.
Honolulu, HI.

Saleh, A., and D. W. Fryrear. 1999. Soil Roughness for the Revised Wind Erosion
Equation (QWEQ). Journal of Soil and Water Conservation. 2“d Quarter 1999. p. 473-476.

Selker, J. S., D. A. Haith, and J. E. Reynolds. 1990. Calibration and Testing of a Daily
Rainfall Erosivity Model. Transactions of the ASAE. 33(5): 1612-1618.

146

Seta, A. K., R. L. Blevins, W. W. Frye, and B. J. Barfield. 1993. Reducing Soil Erosion
and Agricultural Chemical losses with Conservation Tillage. Journal of Environmental
Quality. 22:661-665.

Shah, D. B., G. A. Coulman, L. T. Novak, and B. G. Ellis. 1975. A Mathematical Model
for Phosphorus Movement in Soils. Journal of Environmental Quality. 4(1):87-92.

Sharpley, A. N ., and S. J. Smith. 1990. Phosphorus Transport in Agricultural Runoff: The
Role of Soil Erosion. In: Boardman, J ., I. D. L. Foster, and J. A. Dearing (eds.). Soil
Erosion on Agricultural Land. Pp 351-366. John Wiley & Sons Ltd.

Sharpley, A. N., and M. Meyer. 1994. Minimizing Agricultural Nonpoint-Source Impacts:
A Symposium Overview. Journal of Environmental Quality. 23:1-3.

Sharpley, A. N., S. C. Chapra, R. Wedepohl, J. T. Sims, T. C. Daniel, and K. R. Reddy.
1994. Managing Agricultural Phosphorus for Protection of Surface Waters: Issues and
Options. Journal of Environmental Quality. 23:437-451.

Sharpley, A. N., S. J. Smith, and R. G. Menzel. 1982. Prediction of Phosphorus Losses in
Runoff from Southern Plains Watersheds. Journal of Environmental Quality. 11(2):247-
250.

Sharpley, A. N., and J. K. Syers. 1979. Phosphorus Inputs Into A Stream Draining An
Agricultural Watershed II: Amounts Contributing and Relative Significance of Runoff
Types. Water, Air, and Soil Pollution 11:417-428.

Stevens, R. G., T. M. Sobecki, and T. L. Spofford. 1993. Using the Phosphorus
Assessment Tool in the Field. Journal of Production Agriculture. 6(4):487-492.

Suppnick, J. 1999. Water Chemistry Trend Monitoring in Sycamore Creek and Haines
Drain, Ingham County, Michigan. Staff Report. Michigan Department of Environmental
Quality, Surface Water Quality Division.

Tim, U. S., and S. Mostaghimi. 1989. Modeling Phosphorus Movement and Distribution
in Vadose Zone. Transactions of the ASAE. 32(2):655-661.

Truman, C. C., G. J. Gascho, J. G. Davis, R. D. Wauchope. 1993. Seasonal Phosphorus
losses in Runoff From Coastal Plain Soil. Journal of Production Agriculture. 6(4):507-
513.

U.S. Department of Agriculture, National Soil Erosion Research Laboratory. 1995.
NSERL report No. 10. Water Erosion Prediction Project (WEPP) Technical
Documentation. West Lafayette, IN.

US. Department of Agriculture, Natural Resources and Conservation Service. 1992.
1992 National Resource Inventory. Michigan’s Land: Its Use and Condition.

147

US. Department of Agriculture, Soil Conservation Service. 1992. Soil Survey of Ingham
County, Michigan.

US. Department of Agriculture, Soil Conservation Service. 1990. Sycamore Creek
Watershed Water Quality Plan. Ingham County, Michigan.

US. Department of Agriculture, Soil Conservation Service. 1986. Technical Release 55 —
Urban Hydrology for Small Watersheds, US. Dept. of Agriculture, Soil Conservation
Service, Engineering Division, Washington, DC.

US. Department of Agriculture, Soil Conservation Service. 1972. National Engineering
Handbook, Hydrology Section 4, Chapters 4-10.

US. Environmental Protection Agency. 2000. Water Quality Conditions in the United
States - A profile from the 1998 National Water Quality Inventory Report to Congress.
EPA841-F-00-006. Office of Water, Washington, DC.

US. Environmental Protection Agency. 1999. Total Maximum Daily load. Office of
Water, Washington, DC. http://www.epa.gov/reg3wapd/tmdl/303d.htm.

US. Environmental Protection Agency. 1996 National Water Quality Inventory Report to
Congress. Office of Water, Washington, DC.

US. Environmental Protection Agency. 1994 National Water Quality Inventory Report to
Congress. Office of Water, Washington, DC.

US. Environmental Protection Agency. 1997. Compendium of Tools for Watershed
Assessment and TMDL Development. EPA841-B-97-006. Washington, DC.

US. Environmental Protection Agency. 1996. Section 319 National Monitoring Program
Projects. 1996 Summary Report. EPA-841-S-96-002. Washington, DC.

US. Environmental Protection Agency. 1994. Section 319 Success Stories. EPA-84l-S-
94-004. Washington, D. C.

US. Environmental Protection Agency. 1994. A Watershed Assessment Primer. EPA-
910/B-94-005. Region 10, Seattle, WA.

US. Environmental Protection Agency. 1992. TMDL Case Study: Sycamore Creek,
Michigan. EPA 841-F—92-012. Washington, D. C.

US. Environmental Protection Agency. 1992. Managing Nonpoint Source Pollution.
EPA-506/9-90. Washington, D. C.

148

US. Environmental Protection Agency. 1988. Sediment, Nutrient and Pesticide Transport
in Selected lower Great Lakes Tributaries. EPA-905/4-88-001. GLNPO Report No. 1.
Chicago, IL.

Wade, J. C., and E. O. Heady. 1976. A National Model of Sediment and Water Quality:
Various Impacts of American Agriculture. Center for Agricultural and Rural
Development, Iowa State University, Ames, IA.

Wagner, R. A., T. S. Tisdale, and J. Zhang. 1996. A Framework for Phosphorus
Transport Modeling in the Lake Okeechobee Watershed. Water Resources Bulletin.
32(1):57-73.

Walling, D. E. 1983. The Sediment Delivery Problem. Journal of Hydrology.
65(1983):209-237.

Walling, D. E. 1994. Measuring Sediment Yield from River Basins. In: R. Lal (ed.) Soil
Erosion Research Methods. Soil and Water Conservation Society.

Walpole, R. E., and R. H. Myers. 1985. Probability and Statistics for Engineers and
Scientists. Third Edition. Macmillan Publishing Company. New York.

Williams, J. R., J. G. Arold, C. A. Jones, V. W. Benson, and R. H. Griggs. 1994. Water
Quality Models for Developing Soil Management Practices. In: R. Lal, B. A. Stewart
(eds.). 1994. Soil Processes and Water Quality. Lewis Publishers.

Williams, J .R., C.A. Jones, and PT. Dyke. 1984. A modeling approach to determining
the relationship between erosion and soil productivity. Trans. ASAE 27: 129-144.

Williams, J. R. 1983. Hydrology and Sediment Transport in Small Watersheds. In:
Schaller, F. W., and G. W. Bailey (eds.). Agricultural Management and Water Quality.
Pp 141-151.

Williams, J. R., and R. W. Hann. 1978. Optimal Operation of Large Agricultural
Watersheds with Water Quality Constraints. Texas Water Resources Institute, Texas A &
M University, Technical Report No. 96.

Williams, J. R., and H. D. Bemdt. 1977. Sediment Yield Prediction Based on Watershed
Hydrology. Transactions of the ASAE. Pp 1100-1104.

Williams, J. R. 1975. Sediment Routing For Agricultural Watersheds. Water Resources
Bulletin. 1 1(5):965-974.

Wishmeier, W. H., and Smith, D. D. 1978. Predicting Rainfall Erosion losses - A Guide

To Conservation Planning. US. Department of Agriculture, Agriculture Handbook No.
537.

149

Wishmeier, W. H. 1976. Use and Misuse of the Universal Soil loss Equation. Journal of
Soil and Water Conservation. 31(1):5-9.

Wishmeier, W. H., C. B. Johnson, and B. V. Cross. 1971. A Soil Erodibility Nomograph
for Farmland and Construction Sites. Journal of Soil and Water Conservation. 26:189-
193.

Wishmeier, W. H., D. D. Smith, and R. E. Uhland. 1958. Evaluation of Factors in the
Soil-Loss Equation. Agricultural Engineering. 39:458-462, 474.

Yitayew, M., S. J. Pokrzywka, K. G. Renard. 1999. Using GIS For Facilitating Erosion
Estimation. Applied Engineering in Agriculture. 15(4):295-301.

Young, R. A., C. A. Onstad, D. D. Bosch, and W. P. Anderson. 1978. AGNPS,
Agricultural Non-Point-Source Pollution Model. A Watershed Analysis Tool. U. S.
Department of Agriculture, Conservation Research Report 35.

Yoder, D. C., J. P. Porter, J. M. Laflen, J. R. Simanton, K. G. Renard, D. K. McCool, and
G. R. Foster. 1997. Cover-Management Factor (C). In: Predicting Soil Erosion by Water:
A Guide to Conservation Planning With the Revised Universal Soil loss Equation
(RUSLE). US. Department of Agriculture Handbook No. 703.

Yu, B. 1999. A Comparison of the R-Factor in the Universal Soil Loss Equation and
Revised Universal Soil Loss Equation. Transactions of the ASAE. 42(6): 1615-1620.

Yuan, Y., and J. K. Mitchell. 1999. A Model to Evaluate Pollutant loads from Tile
Drains. Transactions of the ASAE. 42(5): 1313-1319.

Zeverbergen, L. W., and C. R. Thorne, 1987. Quantitative Analysis of Land Surface
Topography, Earth Surface Processes and Landforms. 12:47 - 56.

Zhang, J ., C. T. Haan, T. K. Tremwel, and G. A. Kiker. 1995. Evaluation of Phosphorus
loading Models for South Florida. Transactions of the ASAE. 38(3):767-773.

150

IIIIIIIIIIIIIIIIIIIII

1111111111111111111111111111111111111111111